Atmospheric particles have a remarkable impact on the global environment and climate change. The mineral dust, marine, polluted marine, absorbing, and other types of aerosols are important parts of the global biogeochemical cycles. The land-sea-wind circulation, different heights of boundary layers over sea and continents, the thermal and mechanical turbulence and the pollution emissions in the coastal zones have pronounced impact on the optical properties of the aerosols. In view of these, vertical resolved measurements of optical aerosol properties with calibrated and QA/QC checked lidar systems are necessary. Hence, evaluation and calibration of the data quality of observation equipment are needed urgently. The proposed project tasks are to intercalibrate the lidars from both partners by using EARLINET QA/QC procedures side by side. For this, the lidars from China are scheduled to be transported to Europe. The intercalibration and intercomparison will be conducted at TROPOS in Leipzig/Germany (http://www.tropos.de/) since often particle layers of dust, polluted marine aerosol and other types of aerosol had been observed at TROPOS, the technical infrastructure at TROPOS together with the running systems there is well established. Afterwards, the lidars will be theoretically and experimentally analyzed (including the determination of Müller Matrixes) to determine the contributions of the optical parts to the total system parameters and their uncertainties. With this system calibration and validation results, the optical particle parameters like the extinction coefficient, the backscatter coefficient, the lidar ratio, the aerosol optical thickness, the depolarization ratio, and the Ångström exponent will be measured at TROPOS during a following intensive measurement campaign of about 3 months. The mentioned intensive particle parameters will be used for aerosol type characterization from the observed data.

After this crosscheck, also the intercomparison of the measurement results from the ground-based lidars and from spaceborne lidars (carried by EARTHCARE, ADM-Aeolus, CALIPSO) will be conducted. Furthermore, the wind profiles, the turbulence, and the dynamic structure inside the atmospheric boundary layer will also be observed, which will support the research on the vertical mixing and lateral transport (including sea-land-wind) of aerosols. Through vertical wind speed detection, aerosol flux will be calculated, and thus the strength and deposition of aerosols can be estimated. After the transportation of the lidar systems to Changdao Island / China, a second joint intensive joint measurement campaign will be carried out in this project. This task will enhance the cognition of aerosols like polluted marine, polluted dust, dust, and other aerosol types. It is expected that the aerosols consist mainly of mixtures of mineral dust, pollution, and marine within the planetary boundary layer and in the lofted layers (above) at Changdao Island.

After its launch in autumn 2018, the spaceborne wind lidar ALADIN (Atmospheric LAser Doppler INstrument) on-board ESA’s Earth Explorer satellite Aeolus will allow for global observation of atmospheric wind profiles. Being the first ever satellite-borne Doppler wind lidar instrument, ALADIN will significantly contribute to the improvement in numerical weather prediction by providing one component of the wind vector along the instrument’s line-of-sight (LOS) from ground throughout the troposphere up to the lower stratosphere. The vertical resolution is 0.25 km to 2 km depending on altitude, while the precision in wind speed is envisaged to be between 1 m·s^-1 to 3 m·s^-1.

Over the past years, an airborne prototype of the Aeolus payload, the ALADIN Airborne Demonstrator (A2D), has been developed at DLR (German Aerospace Center) and deployed in several field experiments, aiming at pre-launch validation of the satellite instrument and at performing wind lidar observations under various atmospheric conditions. The A2D features a high degree of commonality with ALADIN in terms of laser source and Doppler lidar receiver design. Thus, it represents the key instrument for the planned calibration and validation activities during the Aeolus mission, as it allows validating the instrument concept, operating procedures as well as wind retrieval algorithms.

In autumn 2016, the A2D was engaged in the North Atlantic Waveguide and Downstream Impact Experiment (NAWDEX). Based in Keflavík, Iceland, this international field campaign had the overarching goal to investigate the influence of diabatic processes, related to clouds and radiation, on the evolution of the North Atlantic jet stream. Apart from providing accurate wind observations for quantifying effects of disturbances on the downstream propagation of the jet, the research flights performed during NAWDEX considerably extended the wind dataset obtained with the A2D as well as with the 2-µm coherent wind lidar on-board the same aircraft – the DLR-Falcon F20. Hence, NAWDEX was an ideal platform for assessing the performance of the two wind lidar systems in heterogeneous atmospheric scenes including strong wind shear and varying cloud conditions.

Besides the DLR-Falcon, three additional aircraft were involved in the campaign being equipped with diverse state-of-the art remote sensing instruments which enabled the observation of a large set of atmospheric parameters, while ground stations delivered a comprehensive suite of further measurements to complement the meteorological analysis. For the first time, coordinated flights were conducted involving the DLR-Falcon, the German HALO deploying an aerosol lidar, a cloud radar and dropsondes as well as the French Falcon SAFIRE with an on-board cloud radar and a UV Doppler lidar instrument. Comparative analysis of the wind data obtained during the collocated flight legs allowed quantifying the accuracy and the precision of the various instruments and demonstrated the complementarity of the different technologies for measuring wind speeds. This work will provide an overview of the NAWDEX campaign and present the results from the wind data analysis both from a meteorological and an instrument point-of-view.

The spaceborne Aerosol and Carbon dioxide Detection Lidar (ACDL) will measure the global column concentrations of carbon dioxide (CO₂) and aerosols profiles simultaneously . The column concentrations of carbon dioxide are measured by 1572 nm double-pulsed integrated path differential absorption (IPDA) lidar technique. The aerosols and clouds profiles are obtained by 532 nm high resolution spectrum lidar (HRSL) technique. Both techniques are combined in the ACDL lidar payload. The dedicated atmosphere and environment monitoring satellite will carry the ACDL lidar and is scheduled to launch in 2020. The spaceborne lidar prototype is being developed. An airborne Aerosol and Carbon dioxide Detection Lidar (AACDL) is developed and high altitude flight validation experiments are scheduled to implement in 2018.

For a double-pulsed 1.57-μm integrated path differential absorption (IPDA) lidar, the transmitted laser pulse energy is an important factor which can influence the uncertainty of the CO₂ Column concentrations measurement. Designing an 1.57μm double-pulsed laser energy monitor and to improve the accuracy of the normalized energy ratio of the transmitter pulse energies to returned echo pulse energies are presented. In the experiments, each pulse is divided into two parts .One is received by the detector directly and the other is delayed by the 200 m multimode fiber. Ground glass diffusers in front of the integrating sphere are used to reduce speckles generated by integrating sphere. Ground glass diffusers with different grits and the rotational speeds are compared. The results show that the rotated ground glass diffuser with 120 grits has the minimum standard deviation of the normalized energy ratio after a moving average. Compared to the situations without the ground glass diffuser or with static ground glass diffuser, the slopes of the Allan deviations of normalized energy ratio with rotated ground glass diffusers are more close to -0.5 in logarithmic coordinates.

In preparation of ESA’s upcoming Earth Explorer mission Aeolus which strives for the global observation of wind profiles from the ground to the lower stratosphere deploying the first-ever satellite-borne wind lidar system ALADIN, the ALADIN airborne demonstrator (A2D) has been developed at DLR (German Aerospace Center). Due to its representative design and operating principle, the A2D provides valuable information on the wind measurement strategies of the satellite instrument as well as on the optimization of the wind retrieval and related quality-control algorithms. Hence, it represents an essential testbed for the planned calibration and validation activities after the launch of Aeolus which is scheduled for end of August 2018.

The A2D was successfully employed for wind observations in the North Atlantic Waveguide and Downstream Impact Experiment (NAWDEX) conducted in Iceland in autumn 2016. Within the scope of the campaign, which aimed to study the influence of diabatic processes on the evolution of the North Atlantic jet stream, 14 research flights were performed extending the wind and calibration dataset of the A2D. In particular, the recording of very high wind speeds above 80 m·s^-1 and strong wind shear of 10 m·s^-1·km^-1 was obtained by sampling an intensified jet stream close to Scotland on 27 September 2016. Broad vertical and horizontal coverage across the troposphere was achieved thanks to the complementary design of the A2D receiver comprising a Rayleigh and Mie channel for analysing both molecular and particulate backscatter signals. Validation of the instrument performance and retrieval algorithms was conducted by comparison with DLR’s coherent wind lidar which was operated in parallel on-board the same aircraft. The systematic error of the A2D line-of-sight (LOS) wind speeds was determined to be less than 0.5 m·s^-1 for both receiver channels, while the random errors range from 1.5 m·s^-1 (Mie) to 2.7 m·s^-1 (Rayleigh). This work will present the operation principle of the A2D and demonstrate selected wind results obtained during NAWDEX.

Ground-based measurements were carried out during field campaigns in April–June of 2010, 2011 and 2012 over northwestern China at Minqin, the Semi-Arid Climate and Environment Observatory of Lanzhou University (SACOL) and Dunhuang, respectively. In this study, three dust cases were examined, and the statistical results of dust occurrence, along with physical and optical properties were analyzed. The results show that both lofted dust layers and near-surface dust layers were characterized by extinction coefficients of 0.25–1.05 km⁻¹ and high particle depolarization ratios (PDRs) of 0.25–0.40 at 527 nm wavelength. During the three campaigns, the frequencies of dust occurrence retrieved from the lidar observations were all higher than 88%, and the highest frequency was in April. The vertical distributions revealed that the maximum height of dust layers typically reached 7.8–9 km or higher. The high intensity of dust layers mostly occurred within the planetary boundary layer (PBL). The monthly averaged PDRs decreased from April to June, which implies a dust load reduction. Comparing the relationship between the aerosol optical depth at 500 nm (AOD₅₀₀) and the Angstrom exponent at 440–870 nm (AE_440–870) confirms that there is a more complex mixture of dust aerosols with other types of aerosols when the effects of human activities become significant.

“Wind Pump” is an important concept that has drawn significant attention in the recent years. Wind Pump is defined as a series of wind-driven processes that influence ocean currents and water movement, which subsequently affect marine ecological conditions. Wind Pump can change the transport of nutrients and promote the cycling of major elements in the ocean. It thus drives primary production and marine ecosystem and affects carbon fixation and global fishery resources (Tang, 2004). This presentation will introduce “Wind Pump” effects on marine systems and take some examples in the South China Sea.

Algal bloom is defined as a rapid increase or accumulation in biomass in an aquatic system. It not only can increase the primary production but also could result in negative ecological consequence, e.g., Harmful Algal Blooms (HABs). According to the two classical theories of algal blooms “critical depth” and “eutrophication”, oligotrophic waters are difficult to form a large area of algal blooms. Cruise observations were only able to capture sporadically the existence of algal blooms. Due to limitations of in-situ observational methods, most of previous studies investigated occasional or regional blooms along coastal eutrophic waters, without much success of understanding of main processes responsible in the offshore deep-ocean oligotrophic waters. Based on previous studies by taking a full advantage of remote sensing technology and multiple satellite data, we proposed the mechanism model of “Wind Pump effects”, which represent the oceanic dynamic mechanism of the bloom growth. Except for the classical coastal Ekman transport, the Wind Pumping effects explain that wind forcing affects the formation of algal bloom through a variety of mechanisms, including Ekman pumping, clip volume, stirring and mixing, and transport by wind and wind-induced surface currents.

The conventional way to study upwelling regions by remote sensing is to use infrared and optical sensors by which the sea surface temperature (SST) and the chlorophyll-a (Chl-a) concentration is measured. However, also synthetic aperture radars (SARs) are useful instruments to study upwelling regions. Upwelling regions are areas of high biological activity, where the marine beings (plankton and fish) secrete surface active substances which rise to the sea surface and damp there the short surface waves, which are responsible for the radar backscattering. Thus upwelling areas manifest themselves on SAR images often as areas of reduced normalized radar cross section (NRCS). However, not only biogenic slicks associated with upwelling regions cause a reduction of the NRCS, but also the change the stability of the air-sea interface (from neutrally-stable to stable) because in upwelling regions the SST is usually lower than over the adjacent areas. Biogenic slicks visible on SAR images as areas of reduced NRCS are often confounded with mineral oil films. Criteria for discriminating between both types of surface films are presented. Furthermore, the correlation between Chl-a distribution and biogenic slick coverage in upwelling areas, like in the South China Sea east of Hainan, the East China Sea north of Taiwan, the Atlantic Ocean west of South Africa, and the Agulhas Return Current in the Indian Ocean, is investigated. These upwelling events are studied by using Sentinel-1 SAR images, Modis SST and Chl-a maps and model data of geostrophic surface currents. It is shown that this synergism yields new insights into upwelling mechanisms.

Sea Surface Temperature (SST) is the most important parameter, which is widely applied for studying water masses, air-sea interaction, marine ecosystem and environment, and other subjects. With the development of half century, satellite remote sensing has become the dominant technique to detect the global SST. However, the satellite measured SST is more closely related to the skin temperature than the subsurface bulk temperature. It is not convictive to validate the satellite measured SST with the subsurface bulk temperature, which is generally measured at a depth of one meter or even deeper. In order to validate the satellite retrieved SST, it is necessary to measure skin temperature.

A new version of the Buoyant Equipment for Skin Temperature (BEST), has been recently manufactured. The new instrument consists of 1050 thermistors, which are integrated in one pole, and 840 thermistors are on the top part (505mm in length) of the pole at 0.6mm distance each and 210 thermistors are on the other part (1015mm in length) of the pole at distance about 5mm. The pole works with a liquid level meter, the liquid level meter uses the electrical capacitance sensors which were also arrayed at 0.6mm distance corresponding to the thermistors. The new instrument BEST was then calibrated in a thermal isolation calibration system, and totally 21 temperature points from temperature -4℃～45℃ were measured for the calibration. The calibration results show the accuracy of the BEST is 0.01K.

The new instrument was vertically floated in Haizhu lake, Guangzhou from January 30 to 31, 2018, continuously for 2 days when the weather is quite cold. It synchronically measures the temperatures of the bottom layer of the air, the skin layer and the subsurface layer of the water at every second and more than hundred thousand temperature profiles were measured. All the temperature profiles have similar distribution pattern. In the bottom of air, the closer to the water surface, the higher temperature. and under the water surface, there is a thin thermocline (or metalimnion) which is just several centimeters thick. In the thermocline the temperature increases with water depth quickly. The water generally increases in temperature by 0.65 degrees Celsius every centimeter. The thermocline has very strong intensity, which is thousand times stronger than normal thermocline occurs in the ocean columns.

Mapping the Yangtze River discharge and freshwater plume spreading is highly important for in the understanding of phytoplankton blooming and nutrient distribution and transportation from the estuary to the East China Sea. Satellite sensor synergy building on passive microwaves, imaging spectrometer and radars are explored together with in-situ observations and dynamic modeling. With new EO satellite data available, such as Chinese Gaofen-4 and the ESA Sentinel-1,2 and 3 there exist possibilities that the freshwater plume mix and transportation process on weekly to seasonal basis can be observed and modelled. Moreover, in this study the Yangtze River Plume transportation dynamics may also be studied by mapping the plumes over the past decades, which may link the variations with large damming in the catchment. We adapt some of the classical methods for retrieval of sea surface salinity distribution with optical remote sensing data by establishing relationships between colored dissolved organic matters (CDOM) and salinity. We will also opt for sea surface brightness temperature methods with which sea surface salinity is obtained by using K-S model, where the brightness temperature is derived from scattering coefficient of SAR data. A Debye Equation based synergic method for sea surface salinity inversion will be thoroughly explored, in which sea surface temperature is synergically derived from brightness temperature through high resolution optical images and sea surface emittance calculated from SAR data.

Large scale green tide (macroalgae blooms of Ulva prolifera) have ocurred in every summer in the Yellow Sea since 2007, causing serious damages on coastal ecological environment, aquaculture, tourism, transportation and so on. The green macroalgae of Ulva prolifera originate from the seaweed aquaculture zone in the Jiangsu shoal, and the blooms are mainly caused by the activity of recycling the seaweed aquaculture facilities. In this work, Gaofen (GF) optical images with high spatial resolution (16m) and high revisit frequency (4 days) and Sentinel-1 IW-GRD microwave data are used to monitor the seasonal changes in the seaweed aquaculture in Jiangsu Shoal (120.8–122°E, 31.9–33.5°N) in 2016 and 2017 with the aim of exploring the reasons on the changes in the magnitude of green tide in the Yellow Sea in the summer of 2016 and 2017.

Macroalgae have the similar spectral signature as that of green vegetation. The normalized differential vegetation index (NDVI) derived from the GF-1 reflectance spectra is used to extract the seaweed aquaculture zone. Considering the difficulty of detecting seaweed aquaculture zone under the ebb tide and bad weather conditions, Sentinel-1 IW-GRD images are used to determine whether it is seaweed aquaculture zone or not.

The result shows that the seaweed aquaculture facilities was recycled mainly in April and May. However, the area of the aquaculture zone was only 1.3 km2 on May 3rd, 2016 while it remained 137.4 km2 on May 7th, 2017. In 2017, the area of the aquaculture zone reduced to 0.7 km2 till June 9th, which shows that the completion time of recycling the seaweed aquaculture facilities in 2017 was about one month later than in 2016. We deduced that the lower magnitude of green tides in 2017 in the Yellow Sea than 2016 may be due to the delay of recycling the seaweed aquaculture facilities. In 2017, the late time of recycling the seaweed aquaculture facilities slowed down the speed of the green macroalgae into the sea, therefore, the scale of the Yellow Sea green tide decreased significantly due to the reduced release of green tide species.

It is well known that rain events leave fingerprints on synthetic aperture radar (SAR) images acquired over the ocean, but it is not always easy to identify them unambiguously, especially not on C-band SAR images. Rain becomes visible on SAR images acquired over the ocean via several mechanisms: 1) by variations of the sea surface roughness caused by downdraft winds associated with rain cells and by rain drops impinging onto the sea surface (surface scattering) generating ring waves, splash products (including stalks), and turbulence, and 2) by scattering and attenuation of the radar beam by rain drops in the atmosphere (volume scattering). Surface scattering is particularly intricate at C-band because the Bragg waves responsible for the radar backscattering at this radar frequency lie in the transition region, where the impinging raindrops can increase (usually) or decrease the backscattered radar power, and also because scattering at stalks generated by impinging rain drops can significantly enhance the radar backscattering. In addition, at very high rain rates, volume scattering and attenuation can also contribute.

In this paper we report about progress that has been made in our study of C-band radar signatures of rain over the ocean. Such studies are relevant also for retrieving sea surface wind fields from C-band scatterometer data. Rain is a main source of error in wind retrieval algorithms, especially when co- and cross-polarized scatterometer data are used, which will be the case in the future. In this study we have analyzed mainly Sentinel-1 SAR images acquired over the South China Sea and have compared them with rain data from the weather radar of the Hong Kong Observatory and from the Global Precipitation Measurement (GPM) mission. In contrast to previously analyzed ERS and Envisat SAR data, the Sentinel-1 SAR data are acquired at VV and VH polarization simultaneously, which allows investigating the role of scattering at stalks, consisting of small cylinders of water emanating from the sea surface, in more detail. Theoretical investigations show that coherent scattering at stalks is responsible for the large values of the normalized radar cross section (NRNCS) at VV and VH polarizations often observed in radar signatures of strong rain cells. This interpretation is supported also by data acquired by the Unmanned Aerial Vehicle Synthetic Aperture Radar (UAVSAR) of NASA/JPL over the Gulf of Mexico.

Yangtze Estuary is located in the margin of land, facing East China Sea. It is influenced by the interaction of land and ocean, developed special environmental characteristics. Riverine freshwater plumes appear in the estuarine area specially, which play an important role in the study of material transport and Yangtze River runoff. Salinity can directly reflect the distribution of freshwater plumes. Therefore, research on the spatial and time distribution and variation of Yangtze River salinity is significant to understanding the importance of freshwater plum and estuarine environment. Compared to the significance of salinity, the measurement of salinity cannot provide sufficient and timely dataset. Remote sensing as a new monitoring technique, is able to provide the real-time synchronous monitoring of large area fast and timely. Existing salinity satellite SOMS and Aquarius cannot apply to the estuarine area because of their low spatial and time resolution. Optical satellite like MODIS, has high spectral resolution, proved suitable to retrieve salinity in estuarine area. This study uses MODIS Terra/Aqua L1b data and field data from voyage and hydrometric station of year 2013 to 2017 to establish a half-experienced retrieval model of Yangtze Estuary. This study divides the study area into inside and outside the Yangtze river estuary. Statistical models are used to the salinity retrieval outside the estuary. A dynamic model is established to t retrieve the salinity inside the estuary, taking runoff volume and tide into consideration, because of the complex hydrological and dynamic environments. The salinity retrieval model is used to reconstruct the salinity distribution of Yangtze Estuary during recent 30 years and analyze the seasonal and spatial salinity variations.

Retrieval of ocean color information is one of the most important missiona of Sentinel-3 Ocean and Land Color Instrument (OLCI). As the successor to Medium Resolution Imaging Spectrometer (MERIS) aboard ENVISAT, OLCI shows significant superiority compared with MERIS as well as Moderate Resolution Imaging Spectroradiometer (MODIS) and Sea-Viewing Wide Field-of-View Sensor (SeaWiFS). The superiority shows in such aspects: the sensor has 21 bands, compared to 15 bands on MERIS, a design optimized to minimize sun-glint and a resolution of 300 m over all surfaces. In this study, we estimated the water quality in the Pearl River using Sentinel-3 OLCI. We appraise the precision of the water quality, including suspended sediment, Chlorophyll-a concentration, CDOM retrieved from OLCI, MERIS, MODIS and SeaWiFS. The results shows that the OLCI shows a good improvement in water quality detection in Pearl River Estuary. The additional bands enhance the ability to extract the information of coastal water quality.

Both the green tide caused by the outbreaks of Ulva prolifera and the golden tide caused by the outbreaks of Sargassum have appeared in the Yellow Sea and the East China Sea in recent years (Xing et al, 2017). The spectral characteristics of floating macroalgae are the basis for the remote detection by optical satellite remote sensing. A total of 10 samples of Ulva prolifera and Sargassum were collected from June 9, 2017 to June 19, 2017 in the Yellow Sea (33º37´～36º30´N, 120º00´～123º30´E). The spectral reflectance of them were measured by a hyperspectral spetroradiometer and a multi-spectral imager, respectively. The hyperspectral data was used to analyze spectral characteristics. The threshold method and neural network method based on the multi-spectral image were tested for the classifying of Ulva prolifera and Sargassum.
In this work, the Virtual-Baseline Floating macroAlage Height (VB-FAH) was calculated for extracting the floating macroalgae (Xing et al, 2016). The reasonable threshold was chosen for the classifying of the two macroalgae based on the trough depth (T-depth) and the Virtual-Baseline Floating macroAlage Height (VB-FAH) (Xing et al, 2013). And the pixel is regard as Ulva prolifera if the value of T-depth is larger than 0.30 or the value of VB-FAH is larger than 0.44. For the neural network method, we did 3 tests with different inputs: the 3-band reflectance image (Image_{_R}), the 3-band reflectance and the T-depth (Image_{_R+T-depth}), the 3-band reflectance and the VB-FAH (Image_{_R+VB-FAH}). The classification results of the above two methods were compared.

Xing Q G, Yu D F, Lou M J, et al, 2013. Using in-situ reflectance to monitor the Chlorophyll concentration in the surface layer of Tidal Flat. Spectroscopy and Spectral Analysis, 33(8): 2188—2191.

References

Xing Q G, Hu C, 2016. Mapping macrolagal blooms in the Yellow Sea and East China Sea using HJ – 1 and Landsat data: Application of a virtual baseline reflectance height technique. Remote Sensing of Environment, 178: 113—126.

Xing Q G, Guo R H, Wu L L, et al, 2017 . High-Resolution satellite observations of a new hazard of "Golden Tides" caused by floating Sargassum in Winter in the Yellow Sea. IEEE Geoscience and Remote Sensing Letters.

Locates at central Himalaya, Mt. Everest (Qomolangma) is the highest peak in the world. Famous glaciers such as Rongbuk glacier and Khumbu glacier were studied by for several long decades. Satellite geodetic observation provides important observation on glacier mass balance in the high-mountain area and plays an essential alternative to in-situ observations given the cold and harsh environment. In this research, we collected SRTM DEM observed in 2000, and bistatic TerraSAR-X/TanDEM-X SAR images observed in around 2013 and 2017. By referring SRTM as reference DEM, we obtained topographic changes between 2000 and 2013, also 2000 and 2017 by using an iterative D-InSAR method. Penetration depth differences between C- and X-band microwave on snow and ice were evaluated and corrected by comparing C- and X-band SRTM DEMs. Glacier mass balance between 2000 and 2013 was -0.38 ± 0.04 m w.e. (water equivalent) a-1, and was -0.75 ± 0.08 m w.e. a-1 between 2013 and 2017. The spatial pattern of the glacier mass loss was heterogeneous. The regional heterogeneity may possibly reflect debris-covering rates, terminating type, temperature rising rates and glacier flow rates. However, the spatial pattern in two periods kept constant. Glaciers without debris-cover at Chinese side present the slowest losing rate while lacustrine-terminating glaciers with heavy debris-covers show quickest lost rates.

Field observations and geodetic measurements suggest that glaciers in the Karakoram Range are either stable or have been expanding since 1990 and present positive or less negative mass changes. This situation is called the “Karakoram anomaly”. Previous studies found that the Central Karakoram has experienced a slight gain in glacier mass at the beginning of the 21st century. Glacier surface velocity is one of the key parameters of glacier dynamics and mass balance. The spatial-temporal characteristics of the glacier velocity in the Central Karakoram are essential to improve our understanding of glacier dynamics and the glacier responses to climate change and influences on regional water sources.

The inter-annual glacier velocity results during 1999-2003 are achieved by using a cross-correlation algorithm in the frequency domain with four pairs of Landsat-7 Enhanced Thematic Mapper Plus panchromatic images. The images were co-registered first, and the horizontal displacements were calculated with the COSI-Corr software package. Due to a lack of in situ measurements of the glacier velocity in the Central Karakoram, it was difficult to directly assess the results of the cross-correlation algorithm. Considering the stable properties of off-glacier areas that should not be displaced, the displacements of the off-glacier area have been widely used to evaluate the cross-correlation performance.

The results show that the variations in ice velocities during 1999–2003 are not obvious for most of the studied glaciers in the Central Karakoram. This indicates that the glacier velocities were quasi-stable during the study period. The uncertainty of the velocity results based on the off-glacier statistics was approximately 7 m/year in the four epochs of observation, which is less than one-half a pixel. We find that most of the glaciers on the southern slope flowed faster than those on the northern slope, which might be attributed to the differences in glacier sizes. From the transverse velocity profiles of seven typical glaciers, we infer that basal sliding is the predominant motion mechanism of the middle and upper glaciers, whereas internal deformation dominates closest to the glacier terminus.

By combusting surface vegetation and soil organic matter, wildfires can have a strong impact on tundra environment. Disturbed vegetation may need many years to recover to pre-fire phase or a mature stage. In this study, we quantified changes of C- and L-band SAR backscatter over 15 years (2002–2016) and used them to investigate vegetation regrowth affected by the Anaktuvuk River Fire in Arctic tundra environment. After the fire, C- and L-band backscatter coefficients increased by up to 5.5 and 4.4 dB in the severely burned areas compared to the unburned areas, respectively. Beyond 5 years after the fire, the C-band backscatter differences diminished between the burned and unburned areas, indicating that vegetation level in burned sites had recovered to the unburned level. This duration is longer than the 3-year recovery suggested by optical-based NDVI observations. Moreover, the L-band backscatter remained about 2 dB higher in the severely burned area than the unburned area after 10-year recovery. Such sustained differences are probably contributed by increased roughness of the surface. Our analysis indicates that long records of space-borne SAR backscatter can quantify post-fire vegetation recovery in arctic tundra environment and complement optical observations.

A single Interferometric Synthetic Aperture Radar (InSAR) interferogram provides a measurement of ground movement with centimetric accuracy, and therefore can only detect large ground motions such as those caused by co-seismic slip or volcano eruption. For detecting small amplitude and long term displacement such as post/inter seismic motion or ground subsidence, a time series of interferograms is needed to overcome the errors resulting from the atmosphere, DEM and orbit. In most of the currently available InSAR time series analysis packages, two fundamental assumptions are made, namely that (i) deformation signals are correlated in time, and (ii) atmospheric effects are correlated in space but not in time. Unfortunately, since atmospheric effects can be highly correlated with topography, the second assumption does not hold in most cases. The temporal correlation of atmospheric delays may completely mask or bias the geophysical signals and introduce unpredictable uncertainties on the velocity estimates.

To overcome this, we propose a strategy which (i) employs a generic InSAR atmospheric correction model for each interferogram by using tightly integrated HRES-ECMWF grid model output and GPS ZTD pointwise observations (global and all-time useable in near real-time); (ii) utilizes a series of model performance indicators to identify the date(s) with poor correction performance, including cross validation of ECMWF and GPS ZTD values, observed phase and modelled atmospheric delay correlations and phase standard deviations; (iii) uses an atmospheric phase screening (APS) model using partially corrected interferograms from step (i) to estimate atmospheric delays for each interferogram: higher performance of the correction model and reliable performance indicators will improve the estimation of APS; and (iv) applies the conventional time series analysis approach to extract the mean deformation rate as well as displacement time series. Our experiments with the proposed method suggest it is particularly beneficial for InSAR time series over mountain areas, as the residual atmospheric errors after correction are more likely to be randomly temporally distributed, which allows an easier minimization through time series analysis.

Landslide is one of the major and most frequently occurring geo-hazards around the world. After the 2008 Wenchuan Earthquake in China, a series of large-scale landslides were triggered. Unexpectedness and concealed nature of the landslides significantly increase the destruction degree and difficulty to prevent, exposing people’s livelihoods and infrastructure at risk.

Space borne radar remote sensing could realize macro dynamic monitoring of large-scale landslide hazards and provide an efficient way to obtain landslide surface deformation and spatio-temporal characteristics, hence contribute to early detection and early warning for local disaster risk management. This work shares several radar remote sensing applications in multiple landslide monitoring case studies in Sichuan since 2014 to till date. Long deformation evolutions of these landslides could be retrieved from time series InSAR processing with joint use of multi-platform InSAR observations. To fully investigate and validate the great potential of Sentinel-1 on landslide monitoring in complex terrain mountainous areas, and integrate the radar datasets from Sentinel-1 and TerraSAR-X, this work realized the landslide surface deformation acquisition with multi scales, short time intervals, and long time series, which also verify the great advantage of multi-platform spaceborne radar remote sensing on landslide monitoring. What’s more, combined with in situ measurements and other remote sensing observations for subsequent analysis and validation, space borne radar remote sensing applications could demonstrate great potentials to identify the spatio-temporal characteristics and investigate the failure mechanism for hazardous landslides.

This paper concludes that a comprehensive and effective Earth Observation (EO) based local landslides monitoring could avoid future human and infrastructure loses in the hill and High Mountain regions around the world.

Previous work with TerraSAR-X [1,2] indicated that landslides can be monitored on steep densely vegetated slopes in hilly terrain using sub-pixel offset tracking, sPOT over the Shuping area, Hubei, PR China. In this work, Cosmo-Skymed Spotlight data is employed at a later time period (27 June 2016 to 30 August 2016) to assess whether the mitigation measures employed to prevent further landslip have been effective using both dInSAR and sPOT processing. The results show good agreement between both methods over this 3 month time period with a small progressive motion towards the NNW of magnitude 10cm in azimuth and 5cm in slant-range. This is much smaller than the previous (accumulated) motion of up to 1m/year from February 2015-2016 using SBAS offset tracking [2] and from February 2009–April 2010 and January 2012–February 2013 using sub-pixel offset tracking [1], prior to the mitigation methods. Part of the reason for the success of dInSAR which was next to impossible to apply previously was that the mitigation measures resulted in a substantial portion of bare earth which had much higher phase coherence than the previously vegetated area. A comparison of the three methods are discussed alongside which one is best in different circumstances.

This work was partially supported by the CSC and UCL MAPS Dean prize through a PhD studentship at UCL-MSSL. We thank Space Catapult, Harwell space campus in general and Terri Freemantle, in particular, for arranging the provision of Cosmo-SkyMed data through the CORSAIR010 data grant.

[1] L. Sun and J.-P. Muller, “Evaluation of the Use of Sub-Pixel Offset Tracking Techniques to Monitor Landslides in Densely Vegetated Steeply Sloped Areas,” Remote Sensing, 8, 25, doi: 10.3390/rs8080659

[2] L. Sun, J.-P. Muller, and J. Chen, “Time Series Analysis of Very Slow Landslides in the Three Gorges Region through Small Baseline SAR Offset Tracking ,” Remote Sensing, 9, 1314. doi: 10.3390/rs9121314

3D SAR Tomography (TomoSAR) [1-4] and 4D SAR Differential Tomography (Diff-TomoSAR) [8-14] can be used to exploit multi-baseline SAR data stacks to create an important new innovation of SAR Interferometry, to sense complex scenes with multiple scatterers mapped into the same SAR cell. In addition to the 3-D shape reconstruction and deformation solution in complex urban/infrastructure areas [2,4], and recent cryospheric ice investigations [5], emerging tomographic remote sensing applications include forest scenarios [3,6,7], e.g. tree height and biomass estimation, sub-canopy topographic mapping, and even search, rescue and surveillance. However, these scenes are often characterized by temporal decorrelation of scatterers, orbital, tropospheric and ionospheric phase distortion and an open issue regarding possible height blurring and accuracy losses for TomoSAR applications particularly in densely vegetated mountainous rural areas. Thus, it is important to extend characterisations of temporal decorrelation, orbital, tropospheric and ionospheric phase distortion.

We report here on 3D imaging (especially of vertical layers) over densely vegetated mountainous rural areas using 3-D SAR imaging (SAR tomography) derived from data stacks of X-band COSMO-SkyMed Spotlight and L band ALOS-1 PALSAR data stacks over Dujiangyan Dam, Sichuan, China. The new TanDEM-X 12m DEM is being used to assist co - registration of all the data stacks first and has raised a number of unforeseen challenges, which will be described. Then, atmospheric correction is assessed using weather model data such as ERA-I and compared against GACOS in addition to ionospheric correction methods to remove ionospheric delay. Then the new TomoSAR method with the TanDEM-X 12m DEM is described to obtain the number of scatterers inside each pixel, the scattering amplitude and phase of each scatterer and finally extract tomograms (imaging), their 3D positions and motion parameters (deformation). A progress report will be shown on these different aspects.

This work is partially supported by the CSC and UCL MAPS Dean prize through a PhD studentship at UCL-MSSL.

[1] A. Reigber, A. Moreira, “First Demonstration of Airborne SAR Tomography using Multibaseline L-band Data,” IEEE TGARS, 38(5), pp.2142-2152, 2000.

[2] G. Fornaro, F. Serafino, F. Soldovieri, “Three Dimensional Focusing With Multipass SAR Data,” IEEE TGARS, 41(3), pp. 507-517, 2003.

[3] M. Nannini, R. Scheiber, R. Horn, “Imaging of Targets Beneath Foliage with SAR Tomography,” EUSAR’2008.

[4] F. Lombardini, F. Cai, D. Pasculli, “Spaceborne 3-D SAR Tomography for Analyzing Garbled Urban Scenarios: Single-look Superresolution Advances and Experiments," IEEE JSTARS, 6(2), pp.960-968, 2013.

[5] L. Ferro-Famil, C. Leconte, F. Boutet, X. Phan, M. Gay, Y. Durand, “PoSAR: A VHR Tomographic GB-SAR System Application to Snow Cover 3-D Imaging at X and Ku Bands,” EuRAD’12.

[6] F. Lombardini, F. Cai, “3D Tomographic and Differential Tomographic Response to Partially Coherent Scenes,” IGARSS’08.

[7] M. Pardini, K. Papathanassiou, “Robust Estimation of the Vertical Structure of Forest with Coherence Tomography,” ESA PolInSAR ’11 Workshop.

[8] F. Lombardini, F. Cai, “Evolutions of Diff-Tomo for Sensing Subcanopy Deformations and Height-varying Temporal Coherence,” ESA Fringe’11 Workshop.

[9] F. Lombardini, “Differential Tomography: A New Framework for SAR Interferometry”, IEEE TGARS, 43(1), pp.37-44, 2005.

[10] Xiang, Zhu Xiao, and Richard Bamler. "Compressive sensing for high resolution differential SAR tomography-the SL1MMER algorithm." In Geoscience and Remote Sensing Symposium (IGARSS), 2010 IEEE International, pp. 17-20. IEEE, 2010.

[11] F. Lombardini, M. Pardini, “Superresolution Differential Tomography: Experiments on Identification of Multiple Scatterers in Spaceborne SAR Data,” IEEE TGARS, 50(4), pp.1117-1129, 2012.

[12] F. Lombardini, F. Viviani, F. Cai, F. Dini, “Forest Temporal Decorrelation: 3D Analyses and Processing in the Diff-Tomo Framework,” IGARSS’13.

[13] Tebaldini, S., & Rocca, F. (2012). Multibaseline polarimetric SAR tomography of a boreal forest at P-and L-bands. IEEE Transactions on Geoscience and Remote Sensing, 50(1), 232-246.

[14] Huang, Y., Ferro-Famil, L., & Reigber, A. (2012). Under-foliage object imaging using SAR tomography and polarimetric spectral estimators. IEEE transactions on geoscience and remote sensing, 50(6), 2213-2225.

On 3 September 2017, North Korea (Democratic People's Republic of Korea, DPRK) claimed it has successfully tested a hydrogen bomb that could be loaded on to a long-range missile. Seismic readings of 6.3 indicated the test was bigger than any other that has been conducted. Punggye-ri Nuclear Test Site is the only known nuclear test site of North Korea. During the past 12 years, nuclear tests were conducted at the site in October 2006, May 2009, February 2013, January 2016, September 2016, and September 2017. Because of political and other complex factors, it is impossible to obtain any GPS, geology, and field surveying data for direct monitoring and research. InSAR provides a new inspiring research method for underground nuclear deformation monitoring. Here, we use multiple spaceborne SAR data that are ALOS-2, Sentinel-1 and TerraSAR-X to retrieve surface displacement caused by the latest 3 events. The results show that InSAR provides an independent tool to locate and retrieve surface displacement of nuclear tests in North Korea as a supplement of seismic and other methods.

Punggye-ri Nuclear Test Site is located in the northern part of DPRK with complicated land cover, high altitude and mountainous terrain. To achieve homogeneous and reliable measurements in the nuclear test site based on InSAR is really challenging. In mountainous regions, the atmospheric phase screen (APS) can cause serious problems in InSAR observation. From the images we have processed, it is obviously to distinguish atmospheric phase delay. Hence, we conduct APS correction based on WRF (Weather Research and Forecasting) and ECMWF (European Center for Medium range Weather Forecasting) to reduce the APS in D-InSAR processing. Second, the coherence of InSAR interferometric pairs is affected by many factors such as spatial-temporal baseline, wavelength and land cover. We selected multiple interferometric combinations and compared the performance of C-band Sentinel-1, L-band ALOS-2 and X-band TerraSAR-X in InSAR deformation monitoring. The results show that the L-band ALOS-2 data are generally more coherent therefore can provide effective information for surface deformation monitoring. Finally, due to the lack of external data to verify the reliability of InSAR results, we cross-validated the monitoring results of multi-source SAR data with different wavelengths, incident angles, and spatial resolutions aiming to get the robust and trustable result.

Key words：InSAR；Underground nuclear test；Surface deformations；Multiple SAR data；North Korea

Land cover classification is one of the important applications of polarimetric SAR (PolSAR) data. With the development of PolSAR techniques and the increasing demand for PolSAR data in applications, many SAR satellites with full-polarization mode have been successively launched, such as the widely used Japanese ALOS-2 PALSAR-2 (ALOS-2) and The Canadian RADARSAT-2 (R-2) data. China also successfully launched the first civilian SAR satellite with full-polarization in January 2017 - GF-3. However, due to the parameter differences in different SAR sensors, the resolution difference and difference in observation incidence, although in the same area there may be different land cover classification result obtained from different SAR images and the feature selection for classification may be different.

The aim of our study is to improve the land classification accuracy using GF-3, R-2 and ALOS-2 polarimetric SAR data. In this study, we used polarimetric decomposition results including Pauli decomposition H-α-A decomposition, and Yamaguchi decomposition as classification features and analyzed their distributions for different land cover types. After that, we selected the optimal combination of decomposition features as classification parameter for GF-3, R-2 and ALOS-2 respectively, and then carried out the experiments of land cover classification in Beijing. The results showed that for GF-3, using the components of Yamaguchi decomposition as feature parameters performs best, but for R-2 and ALOS-2, using the components of H-α-A decomposition as feature parameters performs best. Moreover, ALOS-2 has the highest classification accuracy (80%), but GF-3 and R-2 have similar classification accuracy (77%). Our study gives some references for the application of GF-3 PolSAR data.

Monitoring Anthropogenic Surface Deformation in Tibetan Plateau Using Sentinel-1 Data

Yunfeng Tian, Jingfa Zhang, Yi Luo, Yongsheng Li

Key Laboratory of Crustal Dynamics, Institute of Crustal Dynamics, China Earthquake Administration

1 Anningzhuang Road, Haidian, Beijing 100085, China

tel(O) 86-10-62842646 zhangjingfa@hotmail.com

1. INTRODUCTION

InSAR has been one of the key techniques for crustal deformation study. However, attentions should be paid to various nontectonic surface deformation which can also be captured by InSAR, for example, the ground subsidence due to extraction of underground water, which is common nowadays for densely populated urban regions. The presence of localized deformation arising from anthropogenic activities often obscures the movement of the Earth’s upper crust layer; and thus introduces bias in quantifying slip rates of active faults or motion of crustal blocks. In this work, we focus on the deformation related to human activities in Tibetan plateau, with the help the high-resolution Sentinel-1 C-band SAR data collected from late 2014 to early 2018, aiming to figure out various signals in the InSAR deformation map.

2. DATA & ANALYSIS

We used both ascending and descending orbital data of Sentinel-1 A/B satellites which serve as a validation of the signals we observed. The observation interval was 24 days from late 2014 to early 2017 and 12 days since middle 2017.

We processed the data using the GMTSAR software package (Sandwell et al., 2011). We first aligned all other acquisitions to the super master scene that we manually specified; and then generated interferograms for each acquisition pair. Strong decorrelation during the interferometric processing is rare due to the improved orbits of Senetinel-1 satellites and dry climate on the highland of Tibetan plateau, except for areas with strong seasonal frost deformation. The LOS displacement time series were generated using the coherence-based SBAS method which assigns small weights to pixels with lower coherence and produces a continuous deformation map, compared to traditional methods. Finally, the velocity was derived by fitting a straight line to the displacement time series.

3. RESULTS

(1). Ground subsidence due to mining

The Sentinel-1 data captured clearly the ground subsidence due to the mining activity at Zhaxikang (Figure 1), a town located right at the eastern fault trace of the Sangri-Cuona rift in southern Tibet. The maximum subsiding rate reaches ~10 mm/yr during the data period. Locations of construction sites and buildings were identified from the high-resolution multi-spectral images in Google Earth; and they were in good accordance with the distribution of the subsidence area in InSAR LOS rate map.

Figure 1. InSAR LOS rates (descending orbit) for Zhaxikang Mine in Sangri-Cuona Rift. (a) Location map. (b) Rate profile. The width of buffer zone is 5 km at both sides of the profile line. The color of symbol in profile plot represents the distance to the profile line.

(2). Ground uplift due to oil-drilling activity

There are several oil fields along the Mangya-Huatugou thrust fault zone in Qinghai province, China. The oil-drilling work usually involves injecting water down to the deep after extracting underground oil out, to maintain a certain level of pressure. We observed, using Sentinel-1 InSAR time series analysis, several localized uplifting areas in Qinghai province (Figure 2). The maximum uplifting rate can be > 10 mm/yr in the LOS direction.

Figure 2. InSAR LOS rates (descending orbit) for oil field north of Huatugou Town, Qinghai province, China. (a) Location map. (b) Rate profile.

(3) Other types of small-scale deformation or bias

The ground deformation can be also caused by other human activities, such as the extraction of underground water for agricultural irrigation or drinking in urban area. The cause of such subsidence can usually be investigated by checking the locations of villages or towns where high demand of water supply is often needed.

There are also some subsidence places where no obvious anthropogenic activities are presented. These regions often locate in the river basin or in valley between mountain peaks, and also along certain active fault zones. It is difficult to discern the cause of such deformation without help of other sources. Therefore, attentions should be paid when deriving the contemporary fault slip rate of such active fault.

Moreover, subsidence or uplift trend can also be fake deformation signal, especially in mountainous regions with high and steep topography. The situation might get worse, sometimes, in thrust faulting zone where both crustal uplifting and large topographic errors concur.

4. CONCLUSIONS

Our recent work using the latest spaceborne C-band SAR satellites (Sentinel-1 A/B) data demonstrated that InSAR technique nowadays is capable of measuring the crustal deformation at the millimeter level accuracy. Ground deformation related to anthropogenic activity, either subsidence or uplift, can be detected with sufficient confidence for the broad area in Tibetan plateau. However, there is also regional deformation whose origin is unknown or difficult to investigate. We should prefer to not make conclusions on geological issues before figuring out the origins of such observed deformation.

ACKNOWLEDGEMENTS

This work is supported jointly by National Science Foundation of China (41104001), China Earthquake Administration (Y201711), and Institute of Crustal Dynamics (ZDJ2017-29).

REFERENCES

Sandwell, D., R. Mellors, X. Tong, M. Wei, and P. Wessel (2011). Open radar interferometry software for mapping surface deformation, Eos Trans. AGU 92(28) 234, doi:10.1029/2011EO280002.

Landslides are one of the most damaging hazards for human beings and can be affected by multiple factors, including the natural environment and human activities. Since the Three Gorges Dam on the Yangtze River was completed in 2003, detecting and monitoring the landslides in the upstream area has become more important in order to protect human lives and properties. Compared to conventional in situ measurements, various remote sensing techniques have been carried out and found capable of monitoring landslides in difficult terrain over a large area.

This study focuses on monitoring landslides in the Three Gorges Region (TGR), which is characterised by the high humidity, dense vegetation, and steep slopes. Shuping with centre coordinates of 30.996◦N, 110.609◦E and Tanjiahe with centre coordinates of 31.030◦N, 110.509◦E are the two selected study sites. Synthetic aperture radar (SAR) techniques are applied to monitor landslides in these study areas and mitigation works performed to reduce the risks of landsldies in unstable areas. To assess the accuracy of digital elevation models (DEMs) derived from interferometric SAR data, TLS data was acquired by Zhang and co-workers and this is compared with the post-mitigation 6 m TDX CoSSC DEMs, SRTM and ASTER DEMs and DEMs derived from Cosmo-Skymed Spotlight data. The assessment of mitigation is also carried out by comparing two sets of Terrestrial Laser Scanning (TLS) data of the study sites before and after remediation.

The potential and limitations of using different SAR data, especially Sentinel-1 to identify unstable regions for follow-up acquisitions of TerraSAR-X Staring Spotlight and Cosmo-Skymed Spotlight data are described. The potential of TLS techniques which have not been widely used in previous studies will also be evaluated. Furthermore, the effect of mitigation in landslide area is also going to be assessed.

Acknowledgments: We thank Prof. J. Zhang, Dr. Q. Jiao, and Dr. T. Xue from the China Earthquake Administration for their support on our fieldwork.

Together with SAR interferometry (InSAR), Ground-Based Synthetic Aperture Radar (GBSAR) has proven to be a powerful field-based remote sensing tool for deformation monitoring. This work proposes two complete GBSAR data processing chains developed on the basis of advanced InSAR time series analysis algorithms including the Small Baseline Subset (SBAS) concept and the Persistent Scatterer Interferometry (PSI) for continuous deformation monitoring. The developed SBAS chain exploits redundant interferograms and processes consecutive GBSAR imagery unit by unit, which allows the opportunity to investigate temporarily coherent targets and reduces the requirement of computation memory. Contrarily, the PSI chain is more computationally sufficient and is developed to support early warning and rapid decision-making in urgent situations. Two practical applications are given in this work to demonstrate the feasibility of the developed GBSAR data processing chains for continuous deformation monitoring.

The magnitude 7.0 Jiuzhaigou earthquake occurred in August 8, 2017 resulted in a large number of landslides near the Jiuzhaigou panda sea, causing road congestion and seriously affecting the earthquake emergency rescue progress. The landslide caused by earthquake has the characteristics of wide distribution and large quantity. Because of the urgency of the disaster and high resolution of unmanned aerial vehicle (UAV) images the traditional artificial visual interpretation model cannot meet the needs of earthquake emergency response. Therefore, it is necessary to provide an automatic information identification method. Thus, the distribution range of landslide can be identified quickly and accurately.

Based on the deep analysis of the features of remote sensing images of landslide, an automatic information identification model for object oriented analysis is constructed. Firstly, the remote sensing images are segmented at different scales to obtain different levels of image objects according to different types and scales of land objects. Then, SEath algorithm is used to construct feature rule set automatically by comprehensive utilization of the information of spectrum, texture and shape of object at every level, and the distribution of earthquake-induced landslides is identified. After that, taking artificial visual interpretation as a reference, the recognition accuracy and efficiency are evaluated. Finally, the spatial distribution features of landslide body in topographic factor and fracture distribution layer are analyzed by statistical analysis. The overall accuracy is 94.8%, and the Kappa coefficient is 0.827. At the same time, on the basis of the same configuration of the computer, the present method is twice as efficient as that of the artificial visual interpretation method.

The paper also analyzes the earthquake-induced landslide distribution features in elevation, slope, aspect, fault distance and other factors. The correlation between landslide and topographic factors is found. It is concluded that the earthquake-induced landslide in the study area is mainly controlled by the Tazang fault. The spatial distribution rule can provide information support for landslide risk assessment, disaster investigation, prediction and prevention. There are obvious fault effects in the distribution of landslide.

The geologic observations presented above suggest that conjugate strike-slip faults are significant structures along the Bangong-Nujiang suture zone in central Tibet. However, some small fault zones located inside the Qinghai Xizang Plateau, especially in the secondary blocks, have not attracted enough attention. For example, a series of V-shaped conjugate strike slip fault systems between Lhasa block and Qiangtang block. The V-shaped conjugate strike slip fault zone is composed of a series of small fault zones with oblique lines. It is an important product of the neotectonic movement in the Qinghai Tibet Plateau. It plays an important role in the deformation of the East-West extensional tectonic deformation in the Qinghai Tibet Plateau. This study will use InSAR technology to obtain the surface deformation information of conjugate strike-slip faults(Bengco Fault and Dongqiao Fault). The two faults are nearly 300 km in length. Therefore, the wide range SAR data should be selected (for example, Sentinel-1 IW mode SAR width is 250km) and used to obtain the active fault deformation signal in the whole conjugate strike slip fault at one time, which will help the overall analysis of the fault distribution. We will analysis the whole motion characteristics of conjugate strike-slip faults，investigate the strain accumulation of tectonic deformation in time and space. It is helpful to understand the characteristics of a series of conjugate strike slip faults developed in the middle part of the Qinghai Tibet Plateau.

The tremendous development of InSAR missions (e.g., Sentinel-1A/1B, ALOS-2, TerraSAR-X/TanDEM-X, COSMO-SkyMED, RADARSAT-2, and Gaofen-3) in recent years facilitates the study of smaller amplitude geohazard deformation monitoring using longer time series and over greater spatial scale, and this trend is set to continue with Sentinel-1C/D, Gaofen-3B/C, RADARSAT Constellation planned for launch during 2018-2025. This poses more challenges for correcting interferograms for atmospheric (tropospheric) effects since the spatial and temporal variations of tropospheric delay may dominate over large scales and can cause errors comparable in magnitude to those associated with crustal deformation (e.g. landslides, city subsidence and so on). In previous attempts, observations from Global Navigation Satellite System (GNSS) and Numerical Weather Models (NWM) have been used to reduce atmospheric effects on InSAR measurements, but GNSS-based correction models are limited by the availability (and distribution) of GNSS stations, and for NWM-based correction models, there might be a time difference between NWM and radar observations.

To overcome this, we have developed a generic InSAR atmospheric correction model whose notable features comprise: (i) global coverage, (ii) all weather, all time useability, (iii) correction maps available in near real-time, and (iv) indicators to assess the correction performance and feasibility. The model integrates operational high resolution ECMWF data (0.125-degree grid, 137 vertical levels, 6-hour interval) and continuous GPS tropospheric delay estimates (every 5 minutes) using an iterative tropospheric decomposition model. The model’s performance for InSAR atmospheric correction was tested using globally-distributed interferograms, encompassing both flat and mountainous topographies, mid-latitude and near polar regions, monsoon and oceanic climate systems, achieving a phase precision and displacement accuracy of approximately 1 cm for the corrected interferograms. Indicators describing the model’s performance including (i) GPS network and ECMWF cross-RMS, (ii) phase-delay correlations, (iii) ECMWF time differences, and (iv) topography variations, were developed to provide quality control for subsequent automatic processing, and provide insights of the confidence level with which the generated atmospheric correction maps may be applied.

We have released the Generic Atmospheric Correction Online Service (GACOS) based on the proposed model (http://ceg-research.ncl.ac.uk/v2/gacos/). This service aims to provide InSAR atmospheric correction maps in a convenient way with all features discussed above. The website was released on 6th June 2017 and has received over 10 thousand requests from all over the world. Given the convenience and the real time availability, the website has rapidly responded to recent events such as the Maoxian Landslide (24 June 2017) and the Xinjiang Earthquake (8 August 2017) by providing the atmospheric corrections used in the generation of near real time deformation fields to identify surface damages and contribute to rescue and recovery operations, which have been reported and highlighted by over 20 social medias and organizations.

Seismic indirect economic loss assessment not only has a major impact on regional economic recovery policies, but also it is related to the economic assistance at the national level. However, due to the Cross-regional economic activities and the difficulty of obtaining data, the seismic indirect economic loss are often predicted based on the direct loss of buildings and life lines. Although this method takes into account the impact of production factor stock on economic flows, the effects of disasters on economic activity are neglected and the economic losses in the tertiary industry are seriously underestimated.

The Defense Meteorological Satellite Program (DMSP) provides global images of 4 periods which from morning to night. Since the Operational Linescan System of DMSP (DMSP-OLS) can observe the city night light, it was widely used in population distribution analysis, economic development monitoring and so on. This paper took Sichuan Province as an example to evaluate the impact of earthquake on economic activities on large spatial scale based on DMSP/OLS, and then estimated the recovery of the economy in the disaster area on the view of time and space by analyzing a series of data from pre-event 5 years to post-event 5 years. First, the county economic evaluation model is established. Upon image registration and correction, the nighttime light images are clipped by the county boundaries. Afterwards, counting the nightlight index of all counties, comparing with Sichuan Statistical yearbook, the corresponding relations between nightlight index and economic activities was finally established. Second, a seismic indirect loss Assessment method are presented. Through the analysis of the area and spatial distribution of night-time light around 2008, the spatial migration and change characteristics of economic activities were summarized, which were caused by Wenchuan earthquake. Then a functional relationship between seismic indirect economic loss and night-time light changes of post-earthquake was established. Third, the economic recovery of affected areas was evaluated. The economic recovery of Sichuan Province was evaluated in time and space by comparing with the cumulative growth of night-time light within the 5 years from 2009 to 2013 and the value of per-earthquake.

In this paper, more attention should be paid to the impact of earthquake on social economic activities. Especially in some areas dominated by the service industry, indirect economic losses can better reflect the impact of the disaster on the area. At the same time, it is also hoped that the application of night-time light data in the evaluation of earthquake disaster damage and restoration will also help the government to formulate a policy on regional economic assistance.

Abstract: In November 2017, a strong Mw7.3 earthquake occurred at the Iran Iraq border. The earthquake caused the surface to rise and settlement on both sides of the fault zone, and the maximum displacement of LOS was about 0.85m. The fault rupture begins in the northwest and continues along the fault to the southeast. The coseismic deformation field is retrieved based on ALOS-2 satellite data and Sentinel-1 satellite data. Using the two step inversion algorithm to do the seismic source mechanism inversion, the inversion results are compared with the USGS results and both of them have good coincidence degree, and the inversion of the seismic source mechanism is more fine. It can better analyze and describe the earthquake. The seismogenic structure laid the foundation for studying the fault structure in the area.

Keywords: Iran Iraq earthquake, D-InSAR, Seismic source mechanism inversion

1. research status

In November 12, 2017, a strong earthquake of magnitude Mw7.3 occurred on the Iraqi border in Iran. The epicenter was located at (34.886°N, 45.941° E) and the focal depth was 19km. The earthquake caused more than 500 deaths, thousands of injured, more than 7000 homeless and thousands of houses collapsing, causing huge economic losses and casualties to the local people.

The earthquake occurred at the front of the collision zone between the two large plates - the Arabia plate and the Eurasian plate, along the Iran and Iraq border in the northwest of the Zagros belt. The Zagros thrust belt is a long 1500km fold thrust belt which extends to the west of Iran and extends to northern Iraq. Although Iran and Iraq are earthquake prone areas, there has not been an earthquake above Mw5.0 for many years. The earthquake damage was relatively light on November 12 of 2017, because before the occurrence of the Mw7.3 earthquake, the region had 4.4 levels of pre-earthquake, and most of the people moved to the relatively safe area after the occurrence of the pre-earthquake.

After the earthquake, by collecting the SAR data before and after the earthquake, the coseismic deformation field can be analyzed and processed. Because the acquired SAR data can cover the focal area completely, so the differential interference measurement technique is used to deal with the very clear deformation field after the earthquake. Through the analysis of the coseismic deformation field, it can be seen that the earthquake caused a relative decline of the upper plate and uplifting of the footwall on both sides of the fault, and the maximum displacement of the satellite's flight direction is up to 0.85m.

The two step inversion algorithm is used to estimate the fracture set parameters and the slip distribution of the fault under the constraint of the InSAR result. Firstly, the fault is assumed to be a homogeneous fault model, and the geometric parameters of the fault are calculated. Then the distributed fault model is used to calculate the distributed slip on the fault surface. Using PSOKINV software to inverse the source parameters, the software uses an improved group cooperative stochastic search particle swarm optimization (Particle Swarm Optimization, PSO) algorithm, which mainly solves the optimal solution through a group of random solutions by iterative method.

2. research significance

The Iraq Iran border is located in the collision zone between the Arabia plate and the Eurasian continent plate. The energy of collisions is cumulative and released and then resulting the earthquakes. This area is a shallow source area at most time. Due to frequent devastating earthquake, the Iran government has formulated corresponding building regulations to ensure the safety of the lives and property of the residents. The earthquake magnitude is relatively large, but the casualties are relatively not very serious. It also indicates the necessity of the construction of earthquake resistant buildings and the study at the same time. The seismogenic background and fault structure of the area have important research significance for earthquake disaster prevention in this area.

One of the largest inland earthquakes in Taiwan happened on 21 September 1999, the Mw 7.6 Chi-Chi event. It struck the Taipei Basin, in the Central western part of the island, killing more than 2400 people and damaging 100 000 structures. The rupture was complex with several dislocations along the 100-km long Chelungpu thrust fault. An improved study of this earthquake will allow better understanding of regional fault properties.

Six ERS images from the descending track 232 and covering the period from 21 January 1999 to 25 May 2000 were processed to investigate the co-seismic deformation. The Interferometric Synthetic Aperture Radar (InSAR) technique was used and via the ESA open-source software SNAP. With InSAR, only the footwall can be analysed because the hanging-wall, which likely experiences the main deformation, is densely vegetated resulting in very low coherence in the interferograms. Co-seismic interferograms show about 10-11 fringes which is equivalent to a displacement variation of approximately 30 cm.

We used PSOKINV (Particle Swarm Optimization and Okada Inversion package), a geodetic inversion package, to determine the fault geometry and the slip distribution. First, the non-linear problem is to use the Particle Swarm Optimization (PSO) for geodetic modelling with the assumption of a uniform slip on a rectangular fault. Second, a joint inversion of InSAR and geodetic data (GNSS and levelling) is realised. The GNSS enables us to get information about the hanging-wall of the fault and to improve the modelling. The slip distribution is determined as a linear problem, optimally-smoothed parameters are obtained.

Sub-pixel offset tracking has been used in various applications, including measurements of volcanic activities, glacier movement, earthquakes, landslides, etc., as a complementary method to time series InSAR. In this work, we explore the use of a small baseline subset (SBAS) Offset Tracking approach to monitor very slow landslides with centimetre-level annual displacement rate, and in challenging
areas characterized by high humidity, dense vegetation cover, and steep slopes. This approach, herein referred to as SBAS Offset Tracking, is used to minimize temporal and spatial de-correlation in offset pairs, in order to achieve high density of reliable measurements. This approach is applied to a case study of the Tanjiahe landslide in the Three Gorges Region. Using the TerraSAR-X Staring Spotlight (TSX-ST) data. With sufficient point density, we estimate the precision of the SBAS offset
tracking approach to be 2–3 cm on average. The results are demonstrated accord well with corresponding GPS measurements.

The key point of earthquake emergency is to quickly grasp the disaster, that is, earthquake damage assessment, in which the seismic hazard assessment of buildings is closely related to human life and property, which is the main content of seismic hazard assessment. Bad weather will generally follow the earthquake, the polarization of synthetic aperture radar (PolSAR) which is an active microwave radiation source, can penetrate many materials such as the rain,clouds,fogs,etc, thus it can imaging for the disaster areas in all weather and in all time, withal, the acquisition of target polarization scattering characteristic is relevant to the shape and physical property of the ground target, which benefits to ground-object identification, therefore PolSAR is widely applied in earthquake emergency. Compared with early single-polarization and multi-polarization SAR, full polarization SAR obtain the best effect of observation through flexible change of polarization state, it gets more complete polarization information, more abundant measurement information data, stronger performance for features classification. Earthquake damage buildings extraction can be divided into two kinds of methods: using multi-temporal change detection method and single phase post-earthquake image extraction method. The former one does polarization target classification firstly, then constructs seismic difference map to extract the earthquake damage buildings. Its core is to construct the difference graph, common methods such as establishing the polarization likelihood ratio model, defining polarization difference degree through combining scattering difference and power difference, Whishart distance change detection method etc. There is a difference of scattering mechanism between the collapsed buildings and intact buildings in the fully polarimetric SAR image after the earthquake, which is the theoretical basis for the single phase post-earthquake image extraction.
The current methods include: Polarization classification combined with the minimum heterogeneity criteria aggregation of hierarchical clustering algorithm, and template matching based on feature of image retrieval, the introduction of polarization orientation Angle compensation mechanism to improve and complete the structure of the the collapsed buildings and intact buildings.

On 24^th June 2017, a catastrophic landslide suddenly buried the Xinmo village (in Sichuan province, south-western China), resulting in heavy causalities. After the failure, the disaster assessment was in urgent need for the rescue and relief work. Except the field observation or UAV, spaceborn SAR data could provide valuable information to the disaster assessment.

In this study, we proposed a method that used the SAR interferometry coherence map to identify the landslide boundary and source area. With use of Sentinel-1 SAR images acquired on 12^th, June 2017 and 24^th June, 2017 (13 hours after the failure), the landslide boundary and source area were mapped by this method. It was revealed that the source area of this landslide was not at the top of the mountain. Compared with the UAV image acquired on 26^th June 2017, the location of the landslide boundary and source area were consistent.

This results show that, this first Sentinel-1 interferogram, together with its corresponding coherence and amplitude maps, not only helped us identify the source area of this massive landslide, but also assisted with mapping the landslide boundary. Spaceborn SAR data could help the disaster assessment to some degree.

Abstract: The new developments of satellite series in China and Europe are bringing new opportunities to advance the agricultural monitoring with abundant satellite data. The Sentinel and GF are both quite similar high resolution satellite series onboard European and Chinese satellites, respectively. The Proba-V and FY3-MERSI both have quite similar channels and their own advantages in the medium to low resolution satellite.

This project is going to focus on the crop mapping, crop condition monitoring and crop drought monitoring with both satellite data. The Ningxia Hui autonomous region, one of small size provinces in China, was selected as the study area for the crop mapping study with GF and Sentinel optical satellite data. The field survey was conducted in June, 2016 and 2017 as well as in June/July this year. Sen2Agri, an open source system has been developed and demonstrated in various continents and is now considered as an operational system enabling the delivery in near real time of four products for any region in the world. The GF satellite data were also collected as much as possible for the coverage of Ningxia in the growing season. The processing method of GF data is now developing in order to automatically ingest large volume data. Based on the Sen2Agri system, the 2017 cropland product is already quite promising, with an overall accuracy of 86%. The compatibility of GF data need to be evaluated and combined with Sentinel-2 data in order to improve the classification accuracy.

Another two major agricultural production areas in China, North China Plain and Northeast China Plain were also selected for the crop monitoring and crop mapping with both medium to low resolution satellite data. The field surveys were conducted in summer 2016 and spring in 2017. The relevant Proba-V satellite data have been downloaded and a processing code was developed to extract the Proba-V data for the area of interesting. The FY-MERSI process chain has been developed in the past. The classification approach was integrated with Radom Forest, Support Vector Machine and Neural Net. Hopefully the preliminary results may be reported at the symposium.

Keywords: Crop Mapping; Classification; GF; Sentinel, Sen2Agri

Sentinel-2 for Agriculture system for crop mapping along the season in the Ningxia Hui Autonomous region

Mathilde De Vroey¹, Jinlong Fan², Nicolas Bellemans¹, Xiaoyu Zhang³ ,Lei Zhang³, Qi Xu^2,4,QiLiang Li^2,4, Hao Gao², Sophie Bontemps¹ and Pierre Defourny¹

¹ Earth and Life Institue, Université catholique de Louvain, Louvain-la-Neuve, Belgium

² National Satellite Meteorological Center, China

³ Ningxia Meteorological Science Institute, China

⁴ Shanxi Agricultural University, China

Abstract: Crop area extent estimates and crop type maps provide crucial information for agricultural monitoring and management. Remote sensing imagery in general and, more specifically, high temporal and high spatial resolution data, such as Sentinel-2A and B, constitute a major asset for this kind of application. The flows of observation data provided by these new sensors introduced new conceptual and processing challenges. The development of the Sentinel-2 for Agriculture system (www.esa-sen2agri.org) was supported by the European Space Agency to facilitate the Sentinel-2 and Landsat-8 time series exploitation for agriculture monitoring in most agricultural systems across the globe.

This open source system has been developed and demonstrated in various continents. Sen2Agri is now considered as an operational system enabling the delivery in near real time of four products for any region in the world, namely (1) a monthly cloud free surface reflectance composite at 10-20m, (2) a binary map identifying annually cultivated land at 10m updated every month, (3) a crop type map at 10m (provided twice along the season) for the main regional crops type and (4) an NDVI and LAI maps at 10m describing the vegetative development of crops on a 5 to 10 day basis. This Sen2Agri system can still be improved and further research is needed to optimize the use of the available processing chains and adapt them to the diversity of agricultural landscapes and biophysical environments.

In the context of a Dragon 4 project, this research aims to validate the system for the Ningxia Hui Autonomous Region in China and to evaluate the precision and accuracy of the crop mask and crop type products (L4A and L4B respectively) obtained from Sentinel-2A and Sentinel_2B. In 2017 a field campaign allowed collecting calibration and validation for the whole irrigated floodplain. The ground dataset have been complemented by delineating additional non cropland samples to cover the whole range of the landscape diversity. The whole study site covers an area of 66500 km² corresponding to 6 Sentinel-2 tiles. The Sentinel-2 images of the same season have been downloaded and pre-processed automatically by Sen2Agri system. The Sentinel-2 surface reflectance time series was then processed to generate a crop mask and then a crop type map from the ground truth data provided by a field campaign in 2017. The 2017 cropland product is already quite promising, with an overall accuracy of 86%. Secondly, the Sen2Agri system generates using a random forest classifier a very accurate and precise classification for the main crop types of the region. Nevertheless, several issues were brought to light. Firstly, Sen2Agri tends to neglect the marginal classes, which are much less represented in the training dataset. Secondly, the crop mask which should be generated without any in situ data, i.e. using ESA’s CCI Land Cover 2010 as default base map, needs be improved either by using the ESA’s CCI Land Cover 2015 or by alternative processing strategies. Based on the crop calendars, the timeliness of the products is still to be discussed to understand how long before harvesting an accurate crop type classification can be obtained.

In addition, this study aims to evaluate the potential contribution of GF images to crop mapping in combination with Sentinel-2. First of all the compatibility of GF data need to be evaluated and combined with Sentinel-2 data. Then the complementarity of both data sources will be assessed in terms of accuracy and timeliness.

Keywords: Sen2Agri; Crop Mapping; Classification; GF; Sentinel

Abstract: Identifying crop type with remotely sensed image is the fundamental step for calculating crop area and monitoring crop growth as well as estimating crop yield in the context of agricultural remote sensing. At present, the method of identifying single or two crops among the major staple crops, such as corn, rice and wheat, was well investigated by researchers, however, the identification of all crop types at the same image is very difficult and needs to be further improved.This study intends to use three kinds of classifiers, such as RF, SVM and NN with the Chinese High Resolution satellite (GF) data, to map the crop types in Ningxia. The crop types are recognized as rice, corn, wheat, clover, grapes, alfalfa, vegetables and greenhouse which are planted in the crop land. The Chinese High Resolution satellite (GF) data in 16m spatial resolution covering the entire Ningxia within the growing season was collected as much as possible. Around 1700 ground truth sample data were also collected In June 2017.

The main steps of the study are as (1) randomly dividing all field sample points into 70% training samples and 30% validation samples; further training more samples with the support of Google Earth image taking the crop phenology into account; adding more samples for non-crop area (Water, Built-up, Bareland, Forest, SolarPanel), and finally the best training sample datasets were obtained after the preliminary classification, self-test, and correction of training samples;(2) three classifiers are tuned to get the optimal classification model. The optimal NN activation function is Hyperbolic; The SVM optimal function is Polynomial with the Degree of Kernel Polynomial and Probability Threshold of 6,0.2 respectively; Number of trees and Number of features for RF were set as 1000 and 4 respectively;(3) the classification accuracy and the efficiency of the three classifiers were compared and evaluated. The accuracy evaluation indexes include Overall Accuracy, Producer accuracy, user accuracy, Kappa and F1 Score. The classification results show that NN>RF>SVM for the efficiency, RF>SVM>NN for the classification accuracy;(4) finally, the crop type map was created. The parameters for the Classifiers applied in this study were tuned specially with the training samples. It needs to be further investigated if those parameters may be extended to other areas and training samples.

Keywords: Classification; Crop type mapping; GF; RF;SVM;NN

Abstract: Crop and land cover classification requires a large amount of ground sample data with the location information in support of the supervised classification of remote sensing images and the accuracy evaluation. Due to the limitation of operating efficiency and cost, the traditional sampling method is not sufficient to support the crop classification at large scale. This study proposed an approach of retrieving the ground truth data from GPS photos taken as the vehicle is moving. The key technical aspects in the study include checking and restoring the photo location information; determining the observing azimuth; shifting the photo taken location to the object location; and interpreting the photos and outputting the data set with the crop type, code and the position information. (1) Checking and restoring the photo location information; Due to the failure connecting to the GPS signal, the GPS camera sometimes was not able to record the position information in the photo file. Another set of GPS recorder may be used to record the position as a complementary. The photos without GPS position may be added the position information later on. The photo and GPS records may be matched by the time but the time difference of two sets of equipment should be taken into account. The time difference may be calculated using the photos with the position information. In case that all photos do not have the position information, a few of typical photos should be checked and identified the position with the Google Earth image and then matched with GPS recorder data. An averaged time difference was further calculated and used as an offset to match both photos and the GPS recorder data. (2) Determining the observing azimuth. Many GPS cameras cannot record the observing azimuth. The observing azimuth may be 0-360 degree for one single sample point. When there are two sample points, the moving direction can be determined by the positions of two points. Adding the angle between the moving direction and the observing direction (close to 90 degree) to the azimuth of moving, the observing azimuth is available. The observation direction, left or right should be recorded as well. (3) Shifting the photo taken location to the object location; the position of the photo file is recorded as the position of photo taken and not the position of the object in the photo. The difference of the position should be compensated when the ground truth data is retrieving. The observing azimuth is available after the previous steps, and then the offset may be calculated with an estimated the distance between the photo taken and the position of the object in the photo. (4) Interpreting the photos and outputting the data set with the crop type, code and the position information. The software was developed to display the photo and select the preset crop types and the crop code. And finally, a text file with all these information was output as the ground truth data set. This approach and the software has been demonstrated for several case studies.

Keywords: Sample; GPS photo; GPS

In this paper, we review the main research work and results in the first phase of our dragon-4 project from kick-off to the mid-term. The contents of this paper include the following three parts: 1) multi-source altimetry data fusion and marine application, 2) sea ice freeboard retrieval by Cryosat-2, 3) Sea surface salinity algorithm based on combined active/passive microwave imagers.

In altimetry marine application, a multi-source satellite crossover data comparison of Sentinel-3 SRAL, HY-2A RA and Jason-2 altimeter were conducted, and the accuracy of the sea surface height of Sentinel-3 SRAL altimeter was analyzed. For the capabilities of the new satellite altimeter data to detect mesoscale eddies, the data fusion of multi-source satellite altimetry including Sentinel-3 and Jason-2/3 and the comparison of mesoscale eddies detection using these fusion data are carried out for the different satellite combinations. The mesoscale eddies observation abilities of Sentinel-3 SRAL were summarized.

In sea ice freeboard retrieval, a new method called Bézier curve fitting (BCF) that can simulate the CryoSat-2 SAR-mode waveform is developed for the retrieval of surface elevation of both sea ice and leads. We apply this method for optimizing the retracking procedure. The results of the retracking procedure for the algorithm was validated using data of the Operation IceBridge (OIB) airborne mission. The mean absolute differences between freeboard values retrieved from CS-2 and OIB data were 9.5 and 13.8 cm when using the proposed method. This suggests that the sea ice freeboard data obtained from our proposed BCF method has a high accuracy.

In the study of SSS retrieval, based on the combined active/passive observations of the L-band microwave radiometer and scatterometer onboard Aquarius, a method to retrieval the sea surface salinity under the rainy conditions is developed and validated. The L-band GMFs (Geophysical Model Functions) are developed and the radiation feature of the rough sea surface is analyzed. The dependence of the sea surface emissivity (sensitive to both roughness and freshening) on the backscatter (only sensitive to roughness) is obtained and the rain-induced roughness is corrected. The method is applied to the salinity retrieval under rain. The retrieval results (SSS_rc) are compared with HYCOM data corrected by RIM (Rain Impact Model). The standard deviation of SSS_rc is about 0.5 psu and the bias of SSS_rc shows no clear dependence on the rain rate.

Ocean mesoscale eddies transport properties such as heat, salt and nutrients around the ocean with typical horizontal scales of less than 100 km and timescales on the order of a month. Eddies are important in supplying nutrients to coastal zones and the surface ocean where plankton blooms may result.

Mesoscale eddies can be detected through satellite altimetry technique due to depressions formed as they spin. Traditionally, those measurements have been retrieved through satellite with the Low Resolution Mode (LRM) which allowed a limited resolution and distance to the coast. Now thanks to the constant advance, those limitations have been reduced, allowing a better resolution and consequently obtaining data where before it was not possible, thanks to the new satellites generation (Cryosat-2, Sentinel-3 and Sentinel-6) with Synthetic Aperture Radar (SAR) mode.

This presentation will show the results obtained with the in-house isardSAT SAR ocean retracker [RD-1 and RD-2], using CryoSat-2 L1B data over Bohai Sea region. To do so, an analysis of the precision has been carried out on the geophysical retrievals (Sea Surface Height, Sea Wave Height and sigma0) obtained against the ones on ESA L2. This retracker is able to fit ocean-like surfaces as well as more specular-like responses, expected when getting close to the coast, thanks to an additional fitting parameter related to the surface roughness (Mean-Squared Slopes). Some pre-processing stage is required to choose the proper portion of the waveform related to the surface beneath the track, especially when getting close to the coast due to land contamination. DEM/geoid supported retracking operation is exploited in this case.

On further stages the same analysis will be repeated with Sentinel-3 data since the 1.5 years data only became available recently.

References:

[RD-1] E. Makhoul, M. Roca, C. Ray, R. Escolà, and A. Garcia-Mondéjar, “Evaluation of the precision of different Delay-Doppler Processor (DDP) algorithms using CryoSat-2 data over open ocean”, accepted for publication in Advances in Space Research.

[RD-2] Q. Gao, E. Makhoul, M. J. Escorihuela, M. Zribi, and P. Quintana-Segui, “Comparision of Retrackers’ performances over inland water bodies”, in Geophysical Research Abstracts, vol. 20, EGU2018-14298, 2018, EGU General Assembly 2018.

Oceanic mesoscale eddy is an important mesoscale dynamic process in the global ocean, and it is one of the research hotspots in physical oceanography. Mesoscale eddies play an important role in ocean circulation, material and energy transport and other marine dynamics and marine biochemical processes in the global ocean. Mesoscale eddies usually have a spatial scale of tens to hundreds of kilometers and a time scale of more than ten days to several months. Conventional in situ observations make it difficult to achieve complete observations of mesoscale eddies. Satellite altimetry is the important means of mesoscale eddies detection. Multi-source satellite altimetry data fusion provides abundant data for the global mesoscale eddies detection. ESA launched sentinel-3 satellites equipped with Synthetic Aperture Radar Altimeters (SRAL) on February 16, 2016, which provides new data sources for the detection of mesoscale eddies in global ocean.

In this study, the northwestern Pacific Ocean of Kuroshio region is selected as the experimental area and the mesoscale eddies observation abilities of Sentinel-3 SRAL are analyzed, including the independent detection abilities of Sentinel-3 SRAL and the improvement of mesoscale eddies detection abilities by data fusion with other satellite altimetry data. Firstly, Jason-2 altimeter is taken as the reference and Sentinel-3 SRAL data are compared with Jason-2 at the crossover of each other. Then the data of Sentinel-3 SRAL are corrected and uniformed based on their comparisons at the crossovers. The uniformed Sentinel-3 SRAL data are mapped by the spatial-temporal objective analysis method to the sea level anomaly grid data. The mapping errors are analyzed by the comparisons between the grid data and the Jason-2 along track data. The independent detection abilities of Sentinel-3 SRAL are analyzed by the comparison between the grid data and the AVISO MSLA data. On the other hand, through the multi-satellite data fusion of different combinations of Sentinel-3 altimeter and other satellite altimeter such as Jason-2/3, the mesoscale eddies detection was performed based on the merged sea level anomaly data, and the addition of Sentinel-3 SRAL data for the improvements of mesoscale eddies detection abilities by multi-satellite altimeters are concluded. Based on the above analysis, the mesoscale eddies observation abilities of Sentinel-3 SRAL are summarized.

During the last decade the radar altimetry has entered in its golden age as demonstrated by the different number of missions (Jason-2/-3, CryoSat-2, Saral/Altika, Sentinel-3) currently operating and the forthcoming ones (Sentinel-6). The relatively new operational synthetic aperture radar (SAR) mode in CryoSat-2 and Sentinel-3 missions, opens a new paradigm in the capabilities that can offer an altimetric radar mission. In this line, a scientific proposal within the DRAGON-4 tries to exploit the lessons learned from classical 2-D SAR focusing to evaluate the imaging-like capability of delay-Doppler altimetric radar mounted on Sentinel-3 over coastal areas. In this way, the altimetric product gets closer to the conventional SAR imaging data, but in the altimeter case a “strip-like” image is obtained compared to the classical 2D SAR image.

Conventional delay-Doppler processor (DDP) coherently integrates a series of pulses to provide specific Doppler beams focused to a specific location, which after being correctly aligned (compensating for the slant-range variation, among others) provide several looks that can be incoherently averaged, increasing the performance in terms of geophysical retrieval (increasing the signal-to-noise ratio-SNR). The fully focused DDP moves one step ahead and intends to coherently integrate such information to get an even higher along-track resolution with an improved SNR and the available number of beams.

In order to achieve such imaging capability, the azimuth or along-track phase modulation needs to be compensated for. The relative movement between the scene and the satellite creates a chirp-like modulation in the along-track dimension (quadratic phase response), and so an azimuth compression needs to be performed (once range migration has been compensated) to obtain a fully focused SAR strip, analogous to the well-known range compression (where a specific chirp pulse is compressed).

The main objective of the scientific proposal within the DRAGON-4 is to evaluate the potential capabilities offered by the state-of-the-art Sentinel-3 operational synthetic aperture radar (SAR) mode, when extending the delay-Doppler processing (DDP) to a fully focused DDP (FF-DDP) altimetric operation. This will confer the SAR altimetric product a very high resolution (in the order of 0.6 m) of great interest for Coastal Altimetry (being able to get closer to the coastline), providing much higher number of looks that can be averaged to improve the altimetric performance as anticipated by Raney in [RD-1]:

• Development of an efficient fully focused SAR altimetric processor

• Validation of the processor’s chains using point-like target (transponder)

• Evaluation of the capabilities of the fully focused SAR over coastal regions in Chinese seas

The core of this presentation is to show the preliminary investigations carried out in the development of such innovative processor (FF-DDP), pointing out the specificities of such processing compared to the conventional DDP. The initial implemented processing chain will be described, showing preliminary tests on simulated point-targets. ESA Sentinel-6 simulated data will be exploited as testbed, since the flexibility of the Sentinel-6 interleaved mode allows to emulate different acquisition configurations (potentially simulating a closed burst operation, similar to Sentinel-3 or CryoSat-2 modes) and how this may impact the final results.

References:

[RD- 1] Curlander, John C., and Robert N. McDonough. Synthetic aperture radar. New York, NY, USA: John Wiley & Sons, 1991.

This paper presents two work we developed in the past two years. The first is sea ice freeboard retrieval by Cryosat-2 data; and the second is sea ice drift detection by Sentinel-1 SAR data.

For sea ice freeboard retrieval, a new method called Bézier curve fitting (BCF) that can simulate the CryoSat-2 (CS-2) SAR-mode waveform is developed for the retrieval of surface elevation of both sea ice and leads. We apply this method for optimizing the retracking procedure. Retracking points are fixed on positions at which the rise reaches 70% of the Bézier curve peak in case of leads, and 50% in case of sea ice. In order to evaluate the proposed retracker algorithm we compare it to other methods currently reported in the literature, namely the Threshold-First-Maximum-Retracker-Algorithm and the ESA CS-2 L2I. The results of the retracking procedure for the different algorithms are validated using data of the Operation IceBridge (OIB) airborne mission. For two OIB campaign periods in March 2015 and April 2016, the mean absolute differences between freeboard values retrieved from CS-2 and OIB data were 9.5 and 13.8 cm when using the BCF method, 11.4 cm and 15.6 cm for TFMRA, and 14.5 cm and 15.5 cm for L2I. This suggests that the sea ice freeboard data obtained from our proposed BCF method has a high accuracy.

For sea ice drift detection, in order to solve the problem of high error rate of sea ice drift retrieval that caused by SAR sea ice images have similarities in many areas. And for the purpose of improving the computational efficiency of SAR sea ice drift detection method, multi-scale fast sea ice drift detection method based on principal direction constraint was proposed. Firstly, a pair of full low-resolution SAR image pairs is divided into several sub-image pairs using SAR sea ice image segmentation method based on image matching, and then the main direction of sea ice drift is extracted. Finally, the main direction is used to limit the matching search area of the feature point of SURF algorithm to more accurately extract sea ice drift information of the original resolution SAR. To verify the performance of the fast SURF algorithm based on the main direction constraint. The method is compared with the classic sea ice drift retrieval method. The measured data results show that compared with the traditional SURF algorithm, the matching ratio of feature points is improved by about 10 times, and the calculation efficiency can be increased by about 1 times. Compared with the NCC algorithm, the computational efficiency of this method is dozens of times faster than NCC method, and the image matching accuracy is still higher than that of the NCC method.

Optical remote sensing is one of the most important methods for large-scale observation of ocean internal wave, which has the advantages of wide width and high temporal resolution. However, the optical remote sensing image is affected by cloud, sea condition and imaging angle, which brings difficulty to extract and retrieve ocean internal wave information from the optical remote sensing image. Currently, parameter inversion of internal solitary wave on optical remote sensing image is still based on the inversion model of SAR image. Therefore, a new approach is proposed to establish an experimental system of optical remote sensing to detect internal solitary wave in the laboratory, which aims to explore the response characteristics of optical remote sensing images caused by internal solitary waves. An experimental platform for detecting internal solitary wave by optical remote sensing is constructed by a 3D internal wave flume, a LED light source, CCD cameras and an air blower. The imaging principle of internal waves on optical remote sensing images is quasi-mirror reflection, and LED simulates the parallel incident of sunlight. The method of gravity collapse is used to generate internal waves in the flume of two-layer water. Internal solitary waves with different amplitudes are generated by different collapse heights. Two CCD cameras are used to synchronously observe the surface optical remote sensing images and vertical internal wave images caused by the propagation of internal solitary waves in the same field of view. The mechanism of the internal solitary waves detected on optical remote sensing is compared and analyzed by changing the parameters such as the collapse height, the zenith angle of the sun and the receiving angle of CCD in turn. The experimental results show that the higher the collapse height brings the larger the amplitude of the internal solitary wave. To be more precise, the amplitude is proportional to the collapse height in a certain range. During the process of internal solitary wave propagation, the surface mirror elements are inclined, and the response of the optical remote sensing image corresponds to the vertical displacement of the internal solitary wave one by one. At the same time, stripes are detected on the surface of water by optical remote sensing, which result in the change of gray scale. The relative gray value difference is positively correlated with the amplitude of the internal solitary wave. The larger the amplitude of the internal solitary wave leads to the larger slope of the surface, and finally the greater the change of the light intensity is received by the optical sensor. The research provides a useful reference for quantitative inversion of internal wave parameters on optical remote sensing image.

Keywords: optical remote sensing, internal solitary wave, surface response, relative gray value difference

Mesoscale eddies are rotating coherent structures of ocean currents, which generally refer to ocean signals with spatial scales from tens to hundreds of kilometers and time scales from days to months. Eddies can be found nearly everywhere in the world ocean, and dominate the ocean’s kinetic energy. Over the recent decades, with the advancements in remote sensing satellites and the abundance of in-situ observations data, people find that mesoscale eddies can transport water, heat, salt, and energy as they propagate in the ocean. By combining satellite altimetry and Argo profiling float data, the analysis of eddy three-dimensional structure becomes an important part of studying the oceanic eddy.

The Bay of Bengal, the largest bay in the world, forms the northeastern part of the Indian Ocean. It connects with the South China Sea through the Andaman Sea and the Strait of Malacca. The bathymetric contour of the Bay of Bengal is oriented east-west and the bay presents “n” pattern. As these bathymetric constraints, the local ocean dynamics is complex, with a broad spectrum of processes, from a seasonal reversing monsoon, cyclonic storms, small-scale river plumes, instabilities generated near the continental slope, eddies and large-scale circulation. The Bay of Bengal is a region abundant of mesoscale eddies. In this paper, we analyzed statistical characteristics of mesoscale eddies in the Bay of Bengal based on merged satellite altimetry data as well as Argo profile data.

Firstly, based on satellite altimeter data, the automatic identification method was used to extract the position and shape information of the mesoscale vortices. A series of statistical analysis methods were used to study the statistical characteristics of the mesoscale eddies in the region, e.g., eddy number and lifetime, geographical distribution of eddies, and evolution of eddy properties. Then, based on Argo profile data and climatology data, the eddy synthesis method was used to construct the three-dimensional temperature and salt structure of the eddy in this area.

The Bohai Sea is the southernmost frozen sea in the Northern Hemisphere. The sea ice is a major marine disaster to the Bohai Sea in the winter, which seriously impacts the marine transportation, oil and gas exploitation etc., leading to the great loss to the economical circle surrounding the Bohai Sea. It is very important to evaluate the damaging effects of the sea ice on the marine transportation and offshore constructions (e.g. the oil platform) quantitatively, which has not been studied and analyzed systematically using long-term data so far. In this paper, the quantitative evaluation of the sea-ice disaster in the Bohai Sea will be studied based on the GOCI and Sentinel-1 data. GOCI (Geostationary Ocean Color Imager), to be a payload of COMS satellite launched in Korea in 2010, is the first geostationary sensor in the world, which covers the whole Bohai Sea completely with a spatial resolution of about 500 m of 8 images for one daytime. The Sentinel-1 consists of two satellites (AB) loading C band SAR, which provides single- and dual-polarization data.

The different sea-ice-disaster indexes should be defined for different disaster-bearing bodies. For the marine transportation, its sea-ice-disaster index is equal to multiplying the sea-ice concentration (C_i) by the sea-ice thickness (H_i), which is represented by I₁, that is I₁= C_i × H_i (unit: %∙cm), indicating the sea-ice mass per unit area in physics, and a bigger value means harder breaking ice and less navigable; For the offshore constructions (e.g. the oil platform), its sea-ice-disaster index is equal to multiplying I₁ by the sea-ice velocity (V_i), which is represented by I₂ , that is I₂= I₁ × V_i = C_i × H_i × V_i (unit: %∙cm²∙s⁻¹), indicating the sea-ice momentum per unit area in physics, and a bigger value means a higher extruding pressure and impulse force imposed by the sea ice. In the paper, based on the GOCI and Sentinel-1 data, the sea ice and the sea water are recognized through combining the sea-ice optical and microwave features, which is used to calculate the sea-ice concentration; the sea-ice thickness is retrieved using the sea-ice optical information of GOCI; the sea-ice velocity is extracted through the GOCI geostationary characteristics and the maximum cross correlation method (MCC); based on the sea-ice parameters of the sea-ice concentration, thickness, and velocity, the two types of the sea-ice-disaster indexes I₁ and I₂ can be calculated, which are used to evaluate quantitatively the spatial distribution features and the interannual variations of the sea-ice disaster in the Bohai Sea in the period from 2011 to 2018. The research results will quantitatively shows that the period from 2011 to 2018 is conventional ice condition, which is relatively heavy in 2011 and 2013. The sea-ice-disaster indexes I₁ and I₂ will quantitatively illustrate the space-time distribution features of the sea-ice disaster for the marine transportation and the offshore construction, which can satisfy the request of the sea-ice disaster prevention and reduction and provide the reference of the monitoring and research on the sea-ice disaster.

Bohai sea is located in the northern latitude 37 ° 07 '- 41 ° 0', eastern longitude 117 ° 35 '-121 ° 10', the Bohai sea and its surrounding their rich oil and gas resources, there are a number of important large fields. However, due to the Accumulated ice that drift ice accumulates and accumulates will cause various degrees of impacts on shipping traffic, marine structures, and fishery production in Bohai. It may even cause serious disasters and bring incalculable losses to China's economy. There is an urgent need for studies related to sea ice drift monitoring. The daily drift of sea ice in Bohai sea is changing rapidly, The daily drift of sea ice in Bohai sea is changing rapidly, and the revisit period of microwave scatterometer, microwave radiometer and SAR is longer, and it cannot meet the demand for sea ice drift monitoring in Bohai Sea. The "GF-4" satellite is China's first high resolution geostationary optical remote sensing satellite. It has the unique advantages of short imaging time interval (20s) and high resolution (50m), and is more suitable for sea ice drift tracking. However, the effect of GF4 satellite image product's own error on sea ice drift is rarely researched at home and abroad. Therefore, it is necessary to carry out error analysis of sea ice drift tracking of GF4 satellite imagery.

This paper mainly uses GF4 satellite imagery to carry out the sea ice drift monitoring error analysis with time intervals of 1 minute, 3 hours, 4 hours, and 24 hours. Firstly, the orthorectification of the 28 image data available from August 2016 to March 2018 in the Bohai Sea area was carried out. Then we select the sea-land edge points as control points, and registration of two images which have the same time interval. Next, we recorded the marked same name points which searched from the bottom of Liaodong bay, east of Liaodong bay and west of Liaodong bay respectlly. Statistics the direction and frequency of land point offset sub-regionally and created the rose plots. And maked histogram of the offset and offset angle of land point.

The results show that, when the time interval is 4 hours and 24 hours, the dominant migration direction in the three regions in Liaodong bay is east; when the time interval is 1 minute, the dominant migration direction in Liaodong Bay bottom and Liaodong Bay west coast land is south, Followed by east and southeast respectively; the dominant migration in Liaodong Bay East Coast is north, followed by east; When the time interval is 3 hours, the dominant migration direction in west of Liaodong Bay, bottom of Liaodong Bay and east of Liaodong bay are east, west and south respectively, followed by southeast, east, southeast respectively. The land offset in three regions is major centralized distribution in a range which is from 60m to 80m. That is to say, the offset of land is basically equal to 1.2 times of pixels, and the maximum land offset is less than 2 times of pixels. Through statistical analysis, it can be seen that with the increase of time interval, the land offset will not change much. This study also paves the way for the study of the drift of sea ice.

Iceberg is a potential threat to maritime transport, drilling platforms and shore facilities in high latitude. In existing research iceberg is mainly detected by Constant False Alarm Ratio(CFAR) according to brightness variation between icebergs and background in Synthetic Aperture Radar image. The performance of iceberg detection strongly depends on the accurate statistical modeling of local background clutter measurements, which is also focused on in existing research. In order to accurately detecting iceberg especially iceberg edge, an iceberg detection method combining image segmentation and CFAR algorithm is proposed in this paper. The image is firstly segmented by watershed algorithm which can accurately determine edge of iceberg，the segmentation areas (aggregation of similar pixels) are used for subsequent processing instead of pixels to reduce speckle noise and improve operational efficiency. The statistical characterization of local background including sea ice and water is modeled accurately and the iceberg is finally detected by CFAR.

The Bohai Rim Region is an important economic circle in China. In winter the freezing of sea ice in the Bohai Sea has caused serious impacts on sea shipping and sea-related production, resulting in accidents such as channel blockage, ship damage, and oil platform collapse. The monitoring of sea ice in Bohai Sea is of great significance and has now become the routine operational work of the marine management department. The first geostationary orbit satellite launched by China on December 29, 2015—the High Resolution Four Satellite (GF-4), with an orbit altitude of 36,000 kilometers, equipped with a visible light sensor with 50 m resolution, 400 M-resolution mid-infrared sensors, and gaze cameras with a width greater than 400 km. It can perform a wide range of observations on about one-third of the Earth's surface and can obtain multiple observations within a day. The spatial resolution of the GF-4 is an order of magnitude better than that of the existing geostationary-satellite GOCI. At the same time, it has the characteristics of high temporal resolution of geostationary satellites. It is very advantageous to detect changes in the ice conditions of sea ice in the Bohai Sea. Within one hour, the drift and change of sea ice in the Bohai Sea are relatively fast. Therefore, the better spatial resolution of GF-4 is very suitable for Bohai Sea ice monitoring, play an important role in monitoring and forecasting sea ice conditions in the Bohai Sea.

This article based on the GF-4 Bohai Sea ice imagery studied the optimal wavebands for the identification of sea ice and seawater: use the 29 pictures remote sensing images of the Bohai Sea between 2017 and 2018 obtained from the GF-4 to extracted 377 samples of sea ice and seawater samples respectively, and normalize the spectral values of the five bands of sea ice and seawater samples respectively; There are a total of 57 species band combinations that single band, two bands combination(adding, subtracting, dividing) ,Band 2 (B2) and band 4 (B4) and band 5 (B5) three bands combination(only analysis of 208 sea ice and sea water samples in 2017: the recognition of sea ice and seawater in single band is relatively good, with B2, B4 and B5). Using graphic method and feature distance method to analyze the ability of these band combinations to identify sea ice and seawater. The graphic method is to display the spectral values of sea ice and seawater corresponding to each band combination in a scatter plot, by visual interpretation of scatter plots, qualitative analysis of sea ice and seawater aliasing (total number of mixed sea ice and seawater samples/samples total 377) is less than 10%, think this band can identify of sea ice and seawater; The feature distance method selects the Bahman distance and the Euclidean distance for quantitative analysis of the ability of each band combination to identify sea ice and seawater . Research results show that, In the graphic method, the B2/B4/B5 has the lowest rate of aliasing, which is 5.31%; In the feature distance method, the feature distance of B2/B4/B5 has the largest calculation result, the Euclidean distance calculation result is 8.89336, and the Bahrain distance calculation result is 91.84793; Shows that the analysis results of the two methods are consistent ,The conclusion is that the qualitative and quantitative analysis of the band results is consistent, B2/B4/B5 is the optimal band combination for GF-4 sea ice and seawater identification. The conclusions obtained in this paper have important significance and reference value for GF-4 sea ice monitoring.

Snow depth derived from passive microwave (PMW) with spatial resolution of 25 km is difficult to describe the snow condition and has been generally overestimated in the Qinghai-Tibetan plateau (QTP) which is characterized by patchy snow cover. The main reason for such overestimation is the contribution of low-temperate snow-free ground to the brightness temperature difference between K and Ka bands (TBD). Therefore, in this study, a new method combining MODIS snow cover fraction (SCF) and land surface temperature (LST) and PMW brightness temperature data is developed to derive snow depth at cell size of 0.005^o. MODIS’S SCF is used to identify the snow cover portion of a PMW pixel, and its LST is applied to calculate the TBD contributed from snow-free portion of the PMW pixel. Result shows that after removing such contribution, the TBD value of the PMW pixel is more reasonable and the derived snow depth exhibits relatively smaller errors. The bias and RMSE are 23.4% and 37.3%, respectively, as compared with the 48.5% and 60.5% before such contribution was removed, when using meteorological station observations (2003-2010) as reference. They are 22.5% and 76.1%, respectively, compared with 54.9% and 107.0%, when using field observations (March 2014) as reference. The remaining bias (i.e., overestimation) is mostly due to the TBD contribution (up to 10K) from the low temperature of frozen ground underlying the thin and patchy snow cover (or area). This phenomenon also exists in other cold areas, such as eastern Russia, although not as obvious as on the QTP, because the overall thin and patchy snow cover in QTP could not shield the underling soil from the impact of low air temperature.

Meltwater from rock glaciers could provide a relevant contribution to water supply especially in dry regions. Moreover, rock glaciers could have serious hazard potentials when located at or above steep slopes or when damming lakes. Existing investigations about rock glaciers in High Mountain Asia indicate that the landforms are abundant but information is rare for the Tibetan Plateau and the northern slopes of the Himalaya.

We compiled a rock glacier inventory for the Poiqu region (28° 17´N, 85°58´E) – Central Himalaya/Tibet. The mapping was mainly based on optical images from Sentinel 2 and Google Earth. In addition, we used a hillshade calculated from the new 8 m High Mountain Asia DEM where we filled existing gaps with the 12 m TanDEM-X DEM. Rock glaciers were identified based on their characteristic shape and surface structure. Additional information on the occurrence and activity of the rock glaciers was provided by the InSAR technique using ALOS-1 data.

The preliminary results of the inventory reveal 362 rock glaciers covering an area of about 42 km². The largest one has an area of 2 km² and four have an area of around 1 km². The rock glaciers are located between ~4100 m and ~ 5700 m with a mean altitude of ~5040 m a.s.l.. The mean slope of all rock glaciers is close to 20° (min. 8°, max. 35°). Most of the rock glaciers face towards the Northeast (19%) and West (18.5%). Our study indicates that 158 rock glaciers can be classified as active. We also found rock glaciers damming lakes and infrastructure (streets), which could be threatened by the instability from rock glaciers above.

Future work will concentrate on additional datasets like Sentinel 1 for the improvement of the rock glacier inventory in the Poiqu region.

A group of tongue-shaped periglacial landforms near Jingxian Valley (35°40´N, 94° 00´E) in eastern Kunlun Shan have been reported 20 years ago and classified as “Kunlun-type” rock glaciers due to their unique morphology and slow creeping rates. However, the nature of the northern slopes has remained contentious and later been interpreted as gelifluction deposits. Moreover, the kinematic features of the landform had not been fully investigated or understood. Here, we use satellite SAR interferometry to quantitatively characterize the spatial and temporal changes of the surface movement of these landforms. Five ALOS-1 PALSAR images acquired between 2008 to 2009 over eastern Kunlun Shan area have been used to generate three interferograms to measure the surface movement velocities of the landform. One interferogram records the kinematic information during winter/early spring and the other two are averaged to represent the surface movement during summer. We observe that: (1) the eastern slope is also active with a summer velocity of 20­–60 cm/yr (in the satellite line-of-sight direction, same for all the velocities reported here); (2) the northern lobes moved at 20 to 50 cm/yr in summer, which are much larger than the field measured velocities of less than 3 cm/yr near the front as reported in a previous study conducted from 1980 to 1982; and (3) both the northern lobes and eastern slope are inactive during winter.

The seasonal acceleration in movement of rock glaciers during summer have been observed and, in some cases, no movement can be detected in winter. Gelifluction processes can also trigger seasonal velocity variations. However, creeping rates during summer are typically smaller than 20 cm/yr in cold and dry climate conditions such as Jingxian Valley. Several key pieces of evidence, such as (1) the widespread and relatively fast movement and (2) the large-scale tongue-shaped morphology, suggest that the northern lobes are rockglaciers. The lack of oversteepened fronts presumably results from gelifluction processes of the fine-grained deposits covering the slopes, which smooths out the surface of the landform. The eastern slope shows a similar pattern of seasonal surface kinematic variations to the northern lobes. However, the different morphologic characteristics of the two groups of targets indicate different types of periglacial landforms. With a relatively high surface moving speed and large geometry scale, the northern lobes are unique parts of the alpine permafrost in Eastern Kunlun Shan, representing a mixed type of rock glaciers and gelifluction deposits.

There are approximately 1200 lakes whose area is greater than 1 km² on the Tibetan Plateau (TP), the highest plateau of the world. These lakes are important indicators of environment change because they integrate the basin-wide variations of climate, cryosphere and ecosystems. Previous work on lake changes on the TP during the last several decades have focused on surface area because volume variations need information about lake levels — either in-situ or by satellite altimetry data. However, in-situ measurements are very limited and altimetry data such as ICESat-1 and CryoSat-2 are available at a short term. Here, we present an innovative and robust method that combines digital elevation data and multispectral images to estimate water volume changes for the two largest lakes on the TP, Selin Co and Nam Co from the 1970s to 2015. In addition, the contribution of glacier mass changes to lake volume change between 2000 and 2015 is examined at lake-basin scale using existing estimates based on ICESat and ASTER DEM data. The lake storage changes for Selin Co and Nam Co between 1970s and 2015 are 18.8 and 7.0 Gt. Combining with previous studies of glacier mass balances, the lake volume increase from glacier contribution for two largest lakes, Selin Co and Nam Co, are approximately 28% and 8%, respectively. The future research will extend the estimates of glacier contribution to early 1970s combining declassified satellite data, SRTM and TanDEM-X DTMs and other data sources.

Rock glaciers are important periglacial phenomena in high mountain regions. The Yarlung Tsangpo River basin in the Tibet Autonomous Region of China, the distribution of rock glaciers and their hydrological and environmental effects are poorly understood in the basin. We have produced the first comprehensive inventory of rock glaciers in the Yarlung Tsangpo River basin through the fine spatial resolution satellite data that is freely available on Google Earth, we identified 372 rock glaciers based on their morphological features. We then generated attributes of these rock glaciers including the average length, width, slope, orientation, average elevations of the upper and lower limits, their average elevation and median elevation, as well as hypsometry of each glacier. Through statistical analysis, we show that rock glaciers are situated between 4307 and 5814m a.s.l, with the mean minimum elevation at the front estimated to be 4427 m a.s.l, and the mean maximum elevation at the front estimated to be 5731 m a.s.l. The majority (53%) were found to have a northerly aspect (NE, N, and NW).It provided an important basis for our further understanding of the rock glacier in the Yarlung Tsangpo River basin.

Glacier mass balance, as a direct indicator of climate change, attracted increasing attention in the field of cryosphere. Measuring the region-wide glacier mass balance plays a significant role in understanding the response of glaciers to climate change and their influence on water resources and glacial hazards. In this paper, we derived the mass changes of glaciers according to the geodetic method based on three DEMs representing status of glaciers in different years. These DEMs were generated from declassified Hexagon images (1973-1980), SRTM DEM with 30 m resolution (2000) and TerraSAR-X/TanDEM-X data (2012). All DEMs were co-registered by eliminating errors resulted from horizontal difference and removal of the elevation anomalies. We also took into account errors in the glacier boundary delineation, the seasonal fluctuation in surface elevation, snow and ice density and penetration depth of radar beam. Our expected result is that glaciers mass budgets are negative in the Mt. Xixiabangma during the past period.

This study is aimed to studying electromagnetic anomalies before main shock and aftershocks of Jiuzhaigou earthquake (M=7.0, August 8, 2017) and comparing the phenomana observed by the ground-based CSELF network and by the SWARM satellites. The Jiuzhaigou earthquake (M=7.0, 13:19:46, August 8, 2017, UTC) occurred in the Sichuan province. The CSELF network consists of 30 stations across two main seismic belts in China, in which 15 stations are located in Sichuan and Yunnan provinces. Each station records five alternate EM filed components (Ex, Ey, Hx, Hy, Hz) in a frequency band of 0.001-1000Hz. The data have been recording for about 3 years using the network. In the study on the EM anomalies before earthquakes, the following steps are involved. The first step is to choose the quality data from huge amount of the observed data. Secondly Top-Down Level analysis is carried out for identifying and catching anomalies in the data based on the different time and different frequencies either for Network data or for SWARM data. The final step is to investigate the relationship of anomalies with earthquake events.

Through analysis on the huge amount of Network data, the time series from August 6^th to 12^th is good meaning on obvious disturbance noise existing in the data. But some anomalous phenomena appeared before main shock and successive 18 mid-strong aftershocks. Except for three aftershocks the anomalies are featured as (1) anomalous pulsating clustering of EM fields appeared simultaneously at several stations, e.g., at the station of JianGe in Sichuan, at the LiJiang and JingGu stations in Yunnan with 205km, 770km and 1110km distances to the epicenter, respectively. The smooth variation of EM fields appeared between adjacent clusterings. (2) The pulsating clustering started at about 12-13 minutes before the earthquake and lasted for about 10 minutes and recovered at about 3 minutes before the shock. (3) Individual pulse in the clustering has a period of about 60-80s. (4) The amplitude of maximum pulse in the clustering is about 70% higher than the background value of corresponding EM component. The anomalous pulses seem to be decreased with the distance to epicenter. The clustering form is similar to those of the Pc3-Pc5 pulse clustering, but the observed anomalies by SWARM appeared in the different time section. The clustering is also not caused by co-seismic waves (P and S waves). It is postulated that the anomalies before each shocks may be caused by the shocks during the process of earthquake generation.

Acknowledgement: Tang J, Chen X, Zhan Y, Xiao Q, etc. from IGCEA joined the CSELF observation. The study is supported by NDICC (15212Z0000001) and NSFC (41374077）.

Yaxin Bi¹, Vyron Christodoulou¹, George Wilkie¹, Zhao Guoze², Ming Huang¹ and Han Bing²

1) Faculty of Computing, Engineering and the Built Environment, University of Ulster, Co Antrim, United Kingdom

2) Institute of Geology, China Earthquake Administration, Beijing, China

Email: y.bi@ulster.ac.uk

Electromagnetic (EM) field is sensitive to the stress of plate tectonics, and changes in the duration of earthquake preparation, the changes would cause electromagnetic emission to transmit into ionosphere, which could be observed by satellites. There were a number of studies conducted on DEMETER satellite data, the results shown that precursory phenomena were captured before earthquakes. Piša D, et al. (2013) carried out a rigorous statistic analysis on the 8400 earthquakes that have a magnitude of 5 or greater than 5 and electromagnetic perturbations within 440 kilometers of the earthquake epicenters, the results revealed that the probability of electromagnetic attenuation was very high before 0-4 hours of the events. Le et al. (2015) conducted a survey on studies of ionospheric abnormal behaviors before some great earthquakes and reported ionospheric disturbance to different extent.

This study reports the progress of development of anomaly detection algorithms and their application to analysing the SWARM satellites data and discovering precursory phenomena before large earthquakes. The study selects three earthquakes, i.e. the Ludian earthquake with a magnitude 6.2 occurred on 3 August 2014 in Yunnan, China, the Peloponnese earthquake with a 5.7 magnitude occurred in southern Greece on 29 August 2014 and the Eketahuna earthquake with a 6.8 magnitude occurred in Peru on 20 January 2014. For each earthquake, a 1000kmx1000km study area is defined and divided into 9 grids. For each grid a time series data is generated, as a result each area has 9 sets of time series data. The duration of the selected data is from 25th March 2014 to 24 January 2015, which were recorded by the Vector Field Magnetometer (VFM).

Four different methods are used to generate time series data, i.e. first day, middle, predefined and average points in order to investigate artificial anomalies introduced when generating time series data. The three detection algorithms of CUSUM-EWMA, Fuzzy-inspired and Hot-SAX are specifically selected to address the unknown nature of the EM signals with respect to their duration, their amplitude and frequency changes, they are applied to analyse 27 sets of time series data in order to detect anomalous phenomena before these three earthquakes. The detected results show various phenomena, and no specific patterns can be discovered, which are closely related to the times of occurrence of these earthquakes. From this studying results, the interesting points are observed as follows:

• the algorithms are capable of detecting anomalies, the CUSUM-EWMA provides good anomaly detection, but it struggles in different anomaly cases.
• the satellites observe the whole earth, their revisit time and orbit reveal a serious constraint in generating sufficient and high quality time series data for earthquakes.
• difficulties appear in selecting the fuzzy membership functions (MF) that depend a lot on the form of the input signals

References:

1. Piša D, Němec F, Santolik O, et al. 2013. Additional attenuation of natural VLF electromagnetic waves observed by the DEMETER spacecraft resulting from preseismic activity. J Geophys Res, 118: 5286–5295.
2. Huijun Le, Jing Liu, Biqiang Zhao, Libo Liu. Recent progress in ionospheric earthquake precursor study in China: A brief review, Journal of Asian Earth Sciences. Volume 114, Part 2, Pages 420-430.

In this work we report the development of a system pipeline for the analysis of the Swam satellite electromagnetic data. Our objective is to provide a streamlined functional tool for analyzing electromagnetic data over regions and investigate the relationship of precursory electromagnetic signals to seismic events. The process of the system pipeline consists of three stages of data extraction, data pre-processing and anomaly detection. The first stage provides an interactive interface, allowing users to define study regions and periods of seismic events, and then extract data from the Swarm CDF data archive. The second stage consists of four different pre-processing methods, including the first arrival sampling within regions, middle points and average value, which address the data sparsity problem and the cause of artificial anomalies in a defined region. The last stage offers a range anomaly detection functions underpinned with a variant of the basic CUSUM-EWMA statistical algorithm, fuzzy-logics inspired method, and HOT-SAX method, etc. To demonstrate the potentials of the tool in applying different kinds of algorithms under an anomaly detection scope of electromagnetic sequential time series data, we select a seismic event under scrutiny is in Ludian, China and occurred on 03/08/2014, and present the usefulness of our approach and pinpoint some critical problems regarding satellite data that were identified.

With the support the Wireless Electro-Magnetic Method (WEM) project, we built the first Control Source Extremely Low Frequency (CSELF) continuous observation network which include 30 electromagnetic stations in Beijing Capital Area (BCA) and Southern Section of North-South Seismic Belt in China for the artificial and nature source singles recording. The instruments collect the data 16 seconds every ten minutes with sample rate of 256Hz and then the whole day’s data was analyzed with the method of Flourier transformation and the FFT length was set as 4096. After that we can get the spectrum with the frequency range from 3Hz to 48Hz and the Schumann resonance and six harmonic frequencies can be observed clearly, however, the peak frequency of Schumann resonance are slightly different due to the stations’ location and other factors.

By comparing the long-term observation data of the same station, we can see that 1.The annual variation of the spectrum in Schumann resonance frequency is basically the same as that of other frequency bands. the intensity of the magnetic field is strong in summer, low in winter and the law of long term change conforms to the half cycle sine wave form. From January to July, the power spectral density is increasing, while from July to December, the spectral density of the vibration amplitude decreases.2. The power spectrum of Schumann resonance frequency is smaller than that of surrounding frequency, that is, its variation is more concentrated. 3.For one station the peak frequency of Schumann resonance shift during time. Take Lijiang as an example, and the peak frequency of the first Schumann resonance frequency of the north to south magnetic field component in one year is between 7.5Hz and 7.9Hz, and tends to low frequency in winter and summer, and to high frequency in spring and autumn.

Accurate and recursive maps of crops at the field scale is of great interest for the farmers to optimize the agronomical practices by minimizing the intra-field yield variability. Sentinel 2 and Venμs free available multispectral satellite imagery, with a spectral configuration optimized for vegetation and a revisit time less than 5 days, opens up new perspectives in the framework of precision agriculture. These satellite data can lead to the development of higher level products both at the farm and field scale such as yield estimation and prediction maps, crop nitrogen (N) balance assessment, weed patch detection and bare soil properties estimation (e.g. soil texture and organic matter).

The objective of the study, which was conducted in the framework of the Topic1 of the Dragon4 #32275 program, is to carry out a systematic work to explore the optimal configurations and possible alternative set-ups of algorithms allowing to exploit the full potential of S-2 and Venμs sensors in terms of their spectral and spatial resolutions.

To this aim, the Maccarese farm located in Central Italy, which is the second largest Italian private farm with about 3500 ha of agricultural fields (typically 10 ha or larger) was selected as study area. This because the farmers were equipped of yield maps machinery and in 2018 Venμs new generation satellite started programmed acquisitions on this study area (ADEPAMAC project).

Freely available toolboxes such as BV-Net (Baret et al., 2007), ARTMO or SNAP (ESA) were used for semi-automated retrieval of biophysical parameters through radiative transfer model inversion, i.e. by optimizing LUT-based inversions.

The biophysical canopy variables, expressing the crop ability to intercept and convert solar radiation also reflecting the vigour of the plant canopy, were retrieved using both S2 and Venms sensors when near acquisitions occurred. In particular, LAI and Chl were retrieved and compared in terms of accuracy with respect to the ground truths (LAI measured with LAI2000 and Chlorophyll with Dualex) acquired during 4 different field campaigns in the winter wheat growing season in the study area.

Moreover, these analyses were coupled with the analysis of different spectral indexes/procedures in order to assess the capability of the red-edge bands in retrieving leaf/plant pigments (i.e. chlorophyll, carotenoids) and the Leaf Area Index (LAI).

The results show that both S2 and Venμs satellite sensors are able to retrieve with a good accuracy crop biophysical variables such as LAI and chlorophyll, both by using retrieving algorithms from RT codes or spectral indexes/procedures. Moreover, the few available experimental results suggest that the use of multi-temporal remote sensing data can significantly improve estimation of canopy biophysical variables.

Introduction

Yellow rust (Puccinia striiformis) is one of the most severe epidemic diseases for winter wheat in China, annual affected area of yellow rust on winter wheat is greater than 6.7 million ha during 2000-2016. Pathologically, the development of yellow rust comprises five spore stages, including uredospores, appressorium, basidiospores, spermatia, and aeciospores, and the foliar biophysical variations are critical indicators for tracking the progressive host-pathogen interactions at different stage. The interaction of electromagnetic radiation with plant leaves is governed by their biophysical constituents, and response to infestations. However, current researches for agricultural pests and diseases monitoring generally are premised on a given infestation stage. Hyperspectral- and multispectral- continuum observations permit the acquiring of the host-pathogen processes within entire epidemic stages of rust on wheat. Tracking the progressive infestation is complicated by the following aspects: 1) the pre-existing VIs are not disease-specific, 2) these VIs nonlinearly varying as the increase of pathogen attack hard to express progressive spectral variations caused by the infestation process, 3) spatial and spectral redundancy have to be taken into account. The continuous wavelet transformation (CWT) have been proven to be a promising tool to capture subtle spectral absorption characteristics in detection of foliar constituents. The CWT-derived wavelet features are capable of decomposing raw spectral data into different amplitudes and scales (frequencies) in order to facilitate the recognition of subtle variation (or signals) and held the potential on retrieving foliar constituents.

Objective

The contributions of this paper are: 1) to identify a wavelet-based rust sensitive feature set (WRSFs) for characterizing the spectral changes caused by rust infestation at different stages, 2) to provide insight of the proposed WRSFs into specific leaf biophysical variations in the rust development progress, 3) to evaluate the performance of the proposed WRSFs as input feature space for tracking rust progress and retrieving rust severities on hyperspectral and multispectral images, such as sentinel-2. These continuous goals depend on a multi-temporal hyperspectral observation which covered entire circle of rust infestation.

Study Area

A series of in-situ observations were conducted at the Scientific Research and Experimental Station of Chinese Academy of Agricultural Science (39°30’40’’N, 116°36’20’’E) in Langfang, Hebei province, China, from jointing season (20^th April) to milk-ripe season (25^th May) of winter wheat in the 2017. We selected a cultivar, ‘Mingxian 169’, due to their susceptibility to yellow rust infestation, which were inoculated with yellow rust by spore inoculation in 13^th April. The concentration levels of 9 mg 100^-1 ml^-1 spores solution was implemented to naturally generate infestation levels (all treatments applied 200 kg ha^-1 nitrogen and 450 m³ ha^-1 water). Each treatment and repeat occupied 220m² of field campaigns. The makeup of topsoil nutrients (0 ~ 30 cm deep) in the experiment sites were as follows: soil organic matter 1.41~1.47%, nitrogen 0.07~0.11%, available phosphorus content 20.5~55.8 mg kg^–1, and rapidly available potassium 116.6~128.1 mg kg^–1.

Methodology

A wavelet-based technique for extracting the shape-based reflectance spectral feature was proposed based on the implementation of continuous wavelet transform (CWT), which provides a powerful method for detecting and analyzing weak signals at various scales and resolutions, and for analyzing multidimensional hyperspectral signals across a continuum of scales

A total of 9 hyperspectral VIs that have been reported as the rust-related proxies in relevant researches were selected and compared with the extracted WRSFs for disease detection. These adopted VIs have proved to (1) sensitive to crop growth: modified simple ratio (MSR); (2) pigment variation: structural independent pigment index (SIPI), normalized pigment chlorophyll index (NPCI), anthocyanin reflectance index (ARI), and modified chlorophyll absorption reflectance index (MCARI); (3) water and nitrogen content: Ratio Vegetation Structure Index (RVSI), (4) photosynthetic activity: photosynthetic radiation index (PRI), physiological reflectance index (PHRI); and (5) crop disease: yellow rust-index (YRI), aphid index (AI), and powdery mildew-index (PMI),

In the past, various supervised classification frames have been developed to detect plant stresses from remotely sensed observation, such as Artificial Neural Network (ANN), Decision Trees (DT), and Support Vector Machines (SVM). In this section, linear discrimination analysis (LDA) model and SVM model were used as the example frames for testing and comparison of the performance of WRSFs and VIs on detecting the progressive rust development under the linear and non-linear conditions, respectively.

Conclusion

This study proposed a new shape-based WRSFs from the wavelet transformed reflectance spectra of winter wheat leaves inoculated with yellow rust. The identified wavelet features in WRSFs is capable of capturing and tracking rust related biophysical indices (CHL, ANTH, NBI, and PDM) in progressive host-pathogen interaction. The performance of WRSFs as input feature space for DR estimation and lesions detection of rust was evaluated and compared with traditional VIs that sensitive to disease infestation. Our findings suggest that the WRSFs-PLSR model provide insight into specific host-pathogen interaction during rust development progress, which is more effective than VIs-PLSR model in DR estimation. For the rust lesion detection, the WRSFs-based feature space performed best for both LDA and SVM classification frame. Unlike the traditional techniques, the CWT based technique for WRSFs extraction is simple and straightforward to reflectance spectral signals. No predetermination of wavelength delimitation or other parameterization is needed. The practical WRSFs has great robustness for better understanding the pathological progress in tracking the rust development with hyperspectral data from various sensors. This method may be even applicable to others plan-pathogen systems.

Wheat powdery mildew is one of the serious crop diseases which affect the food safety of China. Integrating multi-source information (Earth Observation-EO, meteorological, etc.) to support decision making in the sustainable management of wheat powdery mildew in agriculture is demanded. With the development of satellite and sensor, the amount of available remote sensing data has increased dramatically. However, the high cost of filed survey data in regional level causes the inconsistency between the number of filed survey samples and the amount of remote sensing data, thus affecting the accuracy of crop monitoring model. In this study, a framework of transfer learning, TrAdaBoost, was used to monitoring the distribution of wheat powdery mildew in study area using the auxiliary field data from another region. This study was carried out in western Guanzhong Plain, Shaanxi province and the auxiliary field survey samples were acquired from south-central part of Hebei province. The Landsat-8 OLI images were used to extract vegetation indices which could indicate the growth status of wheat and meteorological data including Climate Hazards Group InfraRed Precipitation with Station data and the MODIS/Terra Land Surface Temperature and Emissivity (LST/E) product were used to describe the environmental conditions of wheat from booting stage to grain filling stage. With these features, TrAdaBoost with weak learner of Support Vector Machines was used to develop the wheat powdery mildew monitoring model. To evaluate the effect of auxiliary data, a referenced model which only used the samples available in study area was developed using Support Vector Machine. The experimental results suggested that two models provided similar disease distribution patterns over the study area while TrAdaBoost had significant higher accuracy than Support Vector Machine when too few samples available in study area and it could give better or comparative performance with the increase of available samples. When all the samples became available, TrAdaBoost had a higher overall accuracy (80%) and kappa coefficient (0.66) than Support Vector Machine (overall accuracy was 75% and kappa coefficient was 0.59). All these results reveal that transfer learning could be used to monitor the occurrence of wheat powdery mildew.

Key words: wheat powdery mildew; transfer learning; TrAdaBoost; disease monitoring;

Biophysical variables such as Leaf Area Index (LAI) and Leaf Chlorophyll Content (LCC) are of crucial importance for a range of agricultural, forestry and ecological applications. Many approaches have been developed to extract these variables from satellite images. Broadly, these methods can be categorized as Statistical (Parametric and Non-Parametric), based on Radiative Transfer Physical Models and Hybrid approach. Recent studies show that hybrid methods, which combine Radiative transfer modeling (RTM) with intelligent Machine Learning (ML) algorithms may overcome many of the disadvantages of the other methods, for example, by being fast, more robust and having higher generalization capabilities. Different ML methods have been used with data simulated from RTM to extract LAI and LCC, but only Neural Networks (NN) have been successful to reach to the operational use. Although, the retrieval of biophysical variables using NN has been widely applied, the algorithm has some drawbacks: 1) low accuracy at higher values of LAI due to the saturation effect in RTM simulation and NN inversion algorithm and 2) unpredictable results, if training and test data are deviating from each other, even slightly. In the recent years, a suite of kernel-based algorithms have been explored to estimate LAI and LCC from satellite images, and have been shown to be a valid alternative to NN. For example, Kernel Ridge Regression (KRR) algorithm is simple for training and it provides competitive accuracy as compared to NN. Gaussian Processes Regression (GPR), performs well in terms of computational costs and speed, it provides higher accuracies, and uncertainty intervals. However, if trained on large simulated datasets from RTM, these algorithms are computationally expensive and this limits their operational use. Active Learning (AL) techniques have been proposed to reduce the size of the input training data generated from RTM, as they only selects the most informative cases from a large dataset, based on either the uncertainty or diversity of the data points.

In this work a study is presented on the comparison of different hybrid approaches for the retrieval of biophysical variables from Sentinel-2 data, based on the training of kernel-based ML algorithms with simulations from RTM. As a benchmark, the results obtained from these methods are compared to those from the biophysical processor implemented into the ESA Sentinel Application Platform (SNAP) software, which relies on the training of NN with PROSAIL simulations. The same simulated training set, sampled and optimized using active learning techniques is tested with different kernel based machine learning algorithms for the retrieval of biophysical variables from Sentinel-2 images acquired over European and Chinese test sites. Ground data measurement campaigns, on the wheat crop, have been carried out in Maccarese (Italy) and Shunyi (Beijing, China) in correspondence with Sentinel-2 acquisitions, to verify the accuracy of the algorithms. The results of this comparison study allow to obtain useful information in terms of quantitative statistical assessment, as well as of the practicality, computational time and cost of emerging hybrid approaches.

The productivity of wheat, including grain yield and quality, directly determines its market price and related agriculture policies. Currently, most prediction models of wheat yield and grain protein content (GPC), one parameter of grain quality, by remote sensing are a little mechanism and difficult to expand at interannual and regional scales. The objective of this study is to use Hierarchical Linear Model (HLM) integrating hyperspectral data at anthesis and environmental data to achieve yield and GPC prediction at interannual scales. Eight experiments during seven growing seasons, during 2008/2009, 2010/2011, and 2012-2017, were carried out. Fifteen spectral indices from hyperspectral data correlated with GPC at anthesis were calculated, and environmental information including daily radiation, maximum and minimum temperature, and rainfall was mean counted one month before anthesis at each growing season. Results suggested that Standardized leaf area index determining index (sLAIDI) and spectral polygon vegetation index (SPVI) showed the best correlation with yield (r = 0.77) and GPC (r = 0.38), respectively. The estimation of yield and GPC based HLM model considering environmental variations showed higher accuracy (Yield: R² = 0.75 and RMSE = 0.96; GPC: R² = 0.58 and RMSE = 1.21%) than the simple linear models (Yield: R² = 0.60 and RMSE = 0.97; GPC: R² = 0.13 and RMSE = 1.73%). A high consistency between the predicted values and the measured values with HLM method was shown at different years. Overall, these results in this study have demonstrated the potential applicability of HLM model for yield and GPC prediction at various years.

This study estimated wheat aboveground biomass (AGB) based on GF-3 synthetic aperture radar (SAR) data. In the Gaocheng research area, We collected ground. Including: biomass data (aboveground fresh biomass, aboveground dry biomass, fresh ear biomass, dry ear biomass) and soil moisture data. The collected ground samples data and the corresponding SAR data were used to establish biomass estimated models, that were water cloud model and polarized water cloud model. Finally, the effects of different biomass types, ROI window sizes and location accuracy about the biomass estimation result were analyzed. The result shows that the water cloud model is the best wheat biomass estimation model. However, the polarized water cloud model can replace the water cloud model for wheat biomass estimation when there is no soil moisture data. The final results provide a reference for estimating wheat biomass based on GF-3 data.

Introduction

In this paper an approach towards the automatic olive tree extraction from satellite imagery is presented. Automatic olive trees detection at a large geographical scale and their health status evaluation are necessary in order to provide an inventory map that may help in a better planning of the management activities and for predicting the olive production. The olive planted areas can be determined more precisely and in a short time through high resolution satellite images at low cost.

In a previous paper the possibility to detect olive groves affected by xylella was demonstrated by considering several fields and observing the behavior of the annual variation of the NDVI. Therefore, unchanged and changed (eradicated) olive groves show a distinctive behavior due to the change in the presence of olive trees as consequence of the trees eradication requested to stop the spread of the disease. In the plot where the olive trees have been eradicated the NDVI standard deviation (STD) increases significantly due to the reduced importance of the evergreen olive trees in determining the behavior of NDVI with respect to the background characterized by the presence of grass or shrub. However, this analysis was carried out on several plots selected taking into account the sites where the presence of the disease was evaluated with different results. In some cases it was not necessary to remove the plants, in other cases the olive plants were eradicated. The delineation of the plots was carried out manually by visual inspection of the image. In fact, the use of polygons of olive groves taken from the 2012 Corine Land Cover map (CLC 223) was not effective due to the variability of the surface cover types within such polygons. Thus, we decided to devote this paper to develop a technique suitable to identify olive groves with higher accuracy than CLC. Since olive groves are characterized by a significant variability in terms of tree density our classification introduce also a way to assess the olive trees surface cover fraction. A vegetation classification method based on plant biochemical composition and phenological development through the year is presented.

Objectives

The enhanced classification of olive groves has been achieved by utilizing EO data, developing new algorithms, and combining new and existing data from multi-source EO sensors to produce high spatial and temporal land surface information. Concerning this last point. The research activity follows two main approaches:

- Improving the classification of the agricultural areas devoted to olive trees, starting from what has been made available from the Corine Land Cover initiative;

- Developing an approach suitable to be automated for counting trees by using very high spatial resolution images in areas at high risk of infection.

The analysis starts from the following observations and hypothesis:

- the CLC polygons, corresponding to the class 223, outline areas containing fields characterized by different distribution density of olive trees;

- an accurate classification of the olive groves is required for applying further analysis aiming at the assessment of the plant tress status potentially due to diseases;

- the assessment of the olive groves status is based on the analysis of a temporal series of NDVI (Normalized Difference Vegetation Index) taking into account that olive trees exhibit values almost constant of NDVI during the year.

Data and study Area

The study area corresponds to the Province of Lecce, located in the Southern part of the Apulia Region. 77 Sentinel-2 (MSI) cloud free images of the area of interest covering the period February 2015 – July 2017, were found in the ESA database (tale T33XE). The research activity covers the Province of Lecce, that is the Italian area most affected by the Xylella fastidiosa disease causing a rapid decline in olive plantations, the so-called olive quick decline syndrome (OQDS, in Italian: complesso del disseccamento rapido dell'olivo). By the beginning of 2015 it had infected up to a million trees in the southern region of Apulia (Lecce Province).

Methodology

The tree density in the olive groves varies significantly and significantly influences our ability to detect them. The lower is the density, the greater the contribution of the underlying and surrounding vegetation to the detected spectral signature. The changes of leaf Carotenoid (Car) content and their proportion to Chlorophyll (Chl) are widely used for monitoring the physiological state of plants during development, senescence, acclimation and adaptation to different environments and stresses.

Then we developed an automatic olive tree detection technique based on tracking the NDVI and CRI2 (Carotenoid Reflectance Index 2) indexes development during the year. The chlorophyll/carotenoid index CRI2 was specifically implemented to help detecting sparsely populated olive orchards. Decisional rules selected on the basis of NDVI and CRI2 characteristics, retrieved over different test, sites were implemented to carry out the classification. The first objective was to clean the CLC 223 polygons removing the areas that shows no olive coverage at all; the second objectives consisted of adding new olive areas not previously classified to the CLC 223 polygons.

Then, the segmentation of the classified areas has been carried out by using NDVI maps of the area of interest and a mathematical morphology approach. The processing procedure has been implemented in Matlab. The procedure to estimate the fractional cover of olive trees within the previously classified areas foresees the following steps:

1. apply a segmentation function (gradient filter) to each CLC polygon;
2. Compute foreground markers (morphologic operators). These are connected blobs of pixels within each of the objects.
3. Compute max, min, average and standard deviation of NDVI in each polygon;
4. Estimate FCOV (Fraction of trees in each field).

Based on the above described procedure, the accurate olive tree distribution is retrieved for the area of interest. An automatic and continuous olive groves monitoring system is currently under development; this system will be capable of tracking the olive groves status, detecting and evaluating the presence and effects of stressing factors such as pests infestation, specific disease affecting olive plants, and general adverse environmental conditions due to climate changes.

Abstract: Yellow rust is one of the most destructive diseases for winter wheat and has led to a significant decrease in winter wheat quality and yield. Identifying and monitoring yellow rust is of great importance for guiding agricultural production over large areas. Compared with traditional crop disease discrimination methods, remote sensing technology has proven to be a useful tool for accomplishing such a task at large scale. This study explores the potential of the Sentinel-2 Multispectral Instrument (MSI), a newly launched satellite with refined spatial resolution and three red-edge bands, for discriminating between yellow rust infection severities (i.e., healthy, slight, and severe) in winter wheat. The corresponding simulative multispectral bands for the Sentinel-2 sensor were calculated by the sensor’s relative spectral response (RSR) function based on the in situ hyperspectral data acquired at the canopy level. Three Sentinel-2 spectral bands, including B4 (Red), B5 (Re1), and B7 (Re3), were found to be sensitive bands using the random forest (RF) method. A new multispectral index, the Red Edge Disease Stress Index (REDSI), which consists of these sensitive bands, was proposed to detect yellow rust infection at different severity levels. The overall identification accuracy for REDSI was 84.1% and the kappa coefficient was 0.76. Moreover, REDSI performed better than other commonly used disease spectral indexes for yellow rust discrimination at the canopy scale. The optimal threshold method was adopted for mapping yellow rust infection at regional scales based on realistic Sentinel-2 multispectral image data to further assess REDSI’s ability for yellow rust detection. The overall accuracy was 85.2% and kappa coefficient was 0.67, which was found through validation against a set of field survey data. The combination of REDSI and the optimized thresholding method proved to be a powerful method for detecting YR infection in winter wheat at regional scales. This study suggests that the Sentinel-2 MSI has the potential for yellow rust discrimination, and the newly proposed REDSI has great robustness and generalized ability for yellow rust detection at canopy and regional scales. Furthermore, our results suggest that the above remote sensing technology can be used to provide scientific guidance for monitoring and precise management of crop diseases and pests.

Keywords: yellow rust; Sentinel-2 MSI; red edge disease stress index (REDSI); winter wheat; detection

Objective

The aims of this study were to: (1) select the most sensitive bands of multispectral data (Sentinel-2) for identifying healthy wheat and both slight and severe yellow rust infection in winter wheat; (2) propose a new red-edge multispectral vegetation index for discriminating yellow-rust-infected winter wheat from healthy wheat; and (3) map yellow rust infection using realistic Sentinel-2 satellite imagery at regional scales.

Data and study Area

A series of in-situ canopy hyperspectral observations were conducted at the Scientific Research and Experimental Station of Chinese Academy of Agricultural Science (39°30’40’’N, 116°36’20’’E) in Langfang, Hebei province, China, at grain filling stage on 15, 18, and 25 May 2017. The winter wheat cultivar known as ‘Mingxian 169’ was selected, the yellow rust pathogens infected the winter wheat through an inoculation process (spore solution concentration of 9 mg 100⁻¹ mL⁻¹) according to the National Plant Protection Standard (NPPS) on 13 April 2017.

Field surveys of wheat yellow rust infection were conducted in Chuzhou and Hefei, Anhui Province, China (32°6.36′–32°38.02′ N, 117°6.09′–117°49.10′ E) on 9 May 2017 at grain filling stage, where winter wheat is considered to be one of the area’s major crops. Two simultaneous Sentinel-2 multispectral images were acquired on 12 May 2017, from https://scihub.copernicus.eu/, and the full coverage image was mosaicked by two images acquired simultaneously.

Methodology

We integrated the field canopy hyperspectral data based on the sensor’s RSR function to simulate the multispectral reflectance of Sentinel-2 to assess its potential for winter wheat yellow rust monitoring and detection. B4 (red), B7 (Re3), and B5 (Re1) were the most sensitive bands for identifying winter wheat infected with yellow rust through random forest model. Consisting of these three sensitive bands, Red Edge Disease Stress Index (REDSI) was proposed.

REDSI and nine SVIs (NDVI, VARI_green, EVI, RGR, NDVIre1, NREDI1, NREDI2, NREDI3, and PSRI) were selected for testing and comparison of their performances on detecting the wheat yellow rust at the canopy and the regional scales, respectively. The overall accuracy (OA) and kappa coefficient were used to evaluate the classification and discrimination performance of FLDA.

Conclusion

In this study, we developed a new index, REDSI (consisting of Red, Re1, and Re3 bands), for detecting and monitoring yellow rust infection of winter wheat at the canopy and regional scale. Compared with other common spectral vegetation indexes, REDSI has excellent performance in detecting and monitoring yellow rust in winter wheat at the canopy and regional scale, with the overall accuracy of 84.1% and 85.2%, respectively. Furthermore, the index had to be continually validated with other diseases and other cultivars to guide agriculture precision management.

Introduction

Powdery mildew (Blumeria graminis) is one of the most destructive foliar diseases of winter wheat and occurs in areas with cool or maritime climates. The infection of this disease results in a reduction of yield and quality of wheat. According to the statistics of National Agricultural Technology Extension and Service Center (NATESC) of China, the average outbreak area of powdery mildew was recorded to be as high as 10 million ha in the last 17 years. Powdery mildew can infect winter wheat in the whole growth period. Generally, powdery mildew hypha recovers growth in the first decade of February, the beginning development period of the disease is in March and the disease occurs generally in April and greatly in May. However, the current studies on crop diseases were mostly based on one single growth phase image in late stage of disease development, did not consider the temporal change characteristics of diseased crops. Otherwise, remote sensing-based time series were successfully used for crop phenology detection and, crop classification, crop area estimation, etc.

Objective

the objectives of this study were: (1) to analyze the relationship between normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) time series and winter wheat powdery mildew, (2) to monitor the occurrence severity of winter wheat powdery mildew through NDVI and EVI time series, (3) to map the spatial distribution of winter wheat powdery mildew occurrence severity, and (4) to assess the performance of the proposed disease monitor models.

Data and study Area

A total of 42 field survey points were collected in 10th May 2014 in western Guanzhong plain in Shaanxi Province, China, which area is a commonly occurred area of winter wheat powdery mildew. In order to field investigation of diseases occurrence match with the spatial resolution of the remotely sensed image, five 1m×1m representative ranges were relatively uniformly selected in a 30m×30m spatial extent. The central latitude and longitude of each point were recorded by sub-meter differential GPS. The specific survey included wheat growth condition, height and occurrence severity. The occurrence severity was reclassified there levels which include normal, slight and severe to reduce the difficulty of monitoring.

Methodology

A monitoring model for monitoring of powdery mildew occurrence severity based on the NDVI and EVI time series was established. The model almost contained all the critical disease infected information in whole growth period of winter wheat.

Totally, 18 remote sensing images were acquired, for the period from 16^th November 2013 to 9^th April 2014. In order to reduce the impact of cloud cover, three sensors’ data (include WFV sensor data of Gaofen-1 satellite, CCD sensor data of the environment and disaster reduction small satellites and the OLI sensor data of Landsat-8) were chosen. The NDVI and EVI which sensitive to green vegetation and is often used to calculate the quantity and viability of surface vegetation and adverse effects of environmental factors such as atmospheric conditions and soil background were selected to develop time series for disease monitoring, and compared the performance of models with NDVI and EVI time series, respectively.

A significant level of noise was present in the temporal signatures due to clouds, aerosols and snow, etc. Hence, in order to ensure the quality, the NDVI and EVI time series need to be smoothed by discrete wavelet transformation (DWT) before being used, which is an orthogonal function which can be applied to finite group of data and has been widely used in the fields of signal processing and image compression. Support vector machines (SVM) exhibits many unique advantages in solving small sample, non-linear and high-dimensional pattern recognition problems and largely overcomes the problems of dimensionality disaster and over-study. And SVM has been widely used in text recognition, face recognition, gene classification, time series prediction, risk assessment, image classification, etc. In this study, SVM was used to construct monitor model with NDVI and EVI time series, and a leave-one-out cross validation method was used to testing and evaluate the performance of NDVI and EVI time series on monitoring the disease occurrence severity due to the small total sample size.

Conclusion

This study developed a monitoring model of disease occurrence severity based on NDVI and EVI time series features. The difference between NDVI and EVI time series curves of winter wheat infected with different disease severities was obvious. The NDVI and EVI time series were both able to discriminate the disease severities. Both the accuracies of the NDVI and EVI time series models suggested that the NDVI and EVI time series preformed good in quantifying disease severity. Compared the NDVI time series models, the EVI time series achieved a higher monitor accuracy for powdery mildew occurrence severity on winter wheat. Furthermore, the monitoring models with NDVI and EVI time series de-noised by DWT outperformed the models with original NDVI and EVI time series. These results reveal that the disease severity monitoring models based on satellite image time series can be a reference for field disease management.

Crop counting is of great importance for harvest yields estimating, to detect crop stress emergencies, to locate plants and tree species, among others. At the same time, in case of commercial orchards the tree identification is essential for the subsidies given by the European Union. Lately, and thanks the easy accessibility to remote sensing datasets provided by an increasing number of Earth Observation satellites (as Landsat, Sentinel, Planet, or GeoFen constellations) and Unmanned Aerial Vehicles (UAVs), the possibility to include these datasets to retrieve value added information, reducing time and simultaneously covering wide areas is a reality.

In particular forest management applications, and those related to tree identification, give an essential input in order to increase the efficiency of orchards management and the potential detection of pest outbreaks scenarios.

During last years, several projects have been carried out focused on forest monitoring, and some of them in particular related to analyze and control the diffusion of pests, as the case of Xyllela Fastidiosa (Xf), which is one of the most dangerous plant bacteria worldwide, causing a variety of diseases, with huge economic impact for forestry and the environment. In particular in Italy, more than 30,000 trees are under monitoring, and almost 2% of these resulted positive for Xf. And in this context the development of methodologies for tree detection, and a rapid anomaly detection is one of the main challenges, whose outcomes will support further surveys and inspections.

Among the most important international initiatives recently carried out, we could mention the Xf-Actors and POnTE projects, funded by the European Union within the Horizon 2020 EU Framework. Also in this context, the AMEOS project, sponsored by an ESA-Dragon agreement, aims to bring together cutting edge research to provide pest and disease monitoring and forecast information, integrating multi-source information (Earth Observation-EO, meteorological, entomological and plant pathological, etc.) to support decision making in the sustainable management of insect pests and diseases in agriculture. In particular the project team also explores the possibility of using remote sensing images to assess the evolution of diseases on permanent crops (olive groves, vineyards).

For the tests carried out in the present work the area of interest is located in Puglia region, Italy, widely infected with Xf. The region has been divided by the regional authorities in Infected, Containment and Bearing areas. The surveillance of big areas requires the assistance of a remote sensing approach, that has proved its effectiveness in detecting infected trees. For each of these areas, in this work a comprehensive imagery dataset taking into account different spatial and spectral resolutions have been processed. The algorithm has been tested with a set of satellite images (Landsat8-OLI, Sentinel-2, QuickBird, Planet, Gaofen-1), and imagery acquired with the MicaSense-RedEdge sensor on board a UAV SkyRobotics-VTOL-SF6 platform.

Crop counting complexity depends on the quality and resolution of the image, the spacing between trees and the algorithm implemented. In this case, FX (Feature Extraction) algorithm performance has been tested in a wide range of scenarios ingesting the procedures with images of spatial resolutions from 30 m. to less than 5 cm, and spacing trees from 4m. to 10m. The algorithm can be summarized in the following main subtasks: Image calibration, Morphological Filtering, Binary thresholding, Rule based Segmentation, Regionalization, Crop Counting and Geodatabase ingestion.

While with Landsat-8 and Sentinel-2 imagery feature extraction (FX) algorithms are able to detect, extract and count the number of trees in most of aged well point-distributed orchards, FX algorithms applied on QuickBird and UAV imagery are capable to achieve the main goal with a high level of effectiveness, and in case of UAV imagery even in recently planted fields, where the dimension of the objects is within the centimeter scale.

An orchard database automatically enriched with the geo-location of detected trees, will be a valuable resource to update existing orchards monitoring systems, essential to detect unexpected anomalies with the assistance of information extracted from other sources (i.e. on-field sensors, meteorological station, or plant-water-transport sensors.).

This study estimated wheat Aboveground biomass (AGB) based on GF-3 synthetic aperture radar (SAR) data. In the Gaocheng research area, and 40 ground samples data were collected. Including: biomass data (aboveground fresh biomass, aboveground dry biomass, fresh ear biomass, dry ear biomass) and soil moisture data. The collected ground samples data and the corresponding SAR data were used to establish biomass estimated models in the Gaocheng Research Area, that were water cloud model and polarized water cloud model. Finally, the effects of different biomass types, ROI window sizes and location accuracy on the biomass estimation result were analyzed.
The result shows that the water cloud model is the best wheat biomass estimation model. However, the polarized water cloud model can replace the water cloud model for wheat biomass estimation under the situation of without soil moisture data. The final results provide a reference for estimating wheat biomass based on GF-3 satellites.