Conference Agenda

The Poyang Lake and Dongting Lake are Chinas two largest freshwater lakes. Located in the middle reaches of the Yangtze River catchment these large floodplain lakes are home to four Ramsar sites initiated by the UNESCO, to foster wetland conservation. The wetlands in and around the lakes provide numerous important ecosystem services for human well-being, e.g. freshwater resources, retention area for Yangtze River floods, buffer for drought events, natural habitats for millions of endemic and migratory birds, etc. However, anthropogenic influences and related degradation of the lakes wetlands have dramatically increased.

In the project presentation current advances and first results of EO data analyses will be exhibited. EO data, especially from the new European Sentinel-family available since 2014, provide high-resolution information about the Earth surface. Multi-sensor satellite data are used to generate time series for the assessment of water surface and wetland dynamics. Qualitative and quantitative analyses will provide insights into the changes on the Earth’s surface in and around the Yangtze flood plain lakes that have taken place due to anthropogenic influence during the last 10 to 15 years. The results show that: Both lakes have shown a shrinking trend in water surface extent. Wetlands are increasingly used for natural resources exploitation, e.g. sand mining on lake ground and the cultivation of economically important plants at lake shores.

Within the project duration the potential of newly available Sentinel-data in combination with historic satellite data (Landsat, Envisat, etc.) and Chinese satellite data will be demonstrated for the generation of longer and preferably consistent Earth observation time series.

The spatial and temporal variation characteristics of the inundation extent for a lake have a great influnce on landscape structure and function of wetland ecosystem. Poyang Lake is the largest freshwater lake in

China. Poyang Lake wetland is a very important wetland in the world for biodiversity conservation. It is very important to measure accurately the water surface dynamics of the Poyang Lake. In this paper, the scatterplot between lake level and MODIS-derived inundation area of the Poyang Lake was used as a starting point, and the uncertainty of inundation areas with lake water levels was discussed. The results showed that: (1) though a significant linear relationship was found for the water level and inundation area for the Poyang Lake, the water surface area shows uncertainty; (2) Water level in Poyang Lake is higher in the south and lower in the north when lake water level is low, and the spatial heterogeneity of water level was decreasing with lake level increasing; (3) Water level slope in Poyang Lake in the time period of water withdrawal was greater than that in flood period; (4) Affected by fishery practice by levees of dish-shaped sublakes, the water surface area of the sublakes in water withdrawal period is greater than that in flood period. Sand mining also changed the water extent around the sandpits when in low lake level. Therefore the inundation extent for Poyang Lake was affected by topography of lake basin, water discharge from the tributaries in Poyang Lake watershed, and the backwater effect of the Yangtze River, as well as with human activities such as sand mining and fishing practice with levees.

Ground remote sensing by GNSS has been a research topic for the Institute of Geodesy and Photogrammetry (IGP) for some years. In the context of the Hydrology & Cryosphere research of Dragon 4 programme we focus on the determination of the snow water equivalent and of the snow depth. Beside these experimental observations we have been carrying out measurements in order to better assess the propagation of GNSS signals in water. A further interesting, however, not yet really elaborated topic is the measurement of the soil moisture in the frame of landslide hazard monitoring.

In the case of snow water determination we devised a method where GNSS receivers buried in the snow cover are used. From differential measurements between 'snow-free' and 'snow-covered' receivers the water content can be inferred. However, to this end a special refractivity model has been developed and accounted for in the GNSS data treatment. For the determination of the snow height we are mesuring reflected signals in a geometric mode or as applied by different authors, we use the SNR data directly. The propagation through water has been investigated on a dedicated experiment, where we could show the extinction of the signal at roughly 3 cm penetration depth. The water depth can be determined by the geometric analysis of the GNSS signal. Permafrost ground might be a hazard especially if it is located on steep slopes. This is the case in the Alps where Rock Glacier are commonly encountered. The main hazard is due to warm up and to a partial de-freezing. This might be reflected in changing soil/ground moisture. Therefore, the information on this parameter could be one piece of the puzzle in natural hazard assement in alpine areas. We showed, that the temperature and the movements are highly correlated and might even allow for an inversion of the data to determine the sliding horizon.

Soil moisture, the amount of water contained in the soil, is an environmental descriptor that integrates much of the land surface hydrology. From agriculture production to flood and drought prediction, soil moisture plays an important role in human production and human life. Soil moisture monitoring based on GNSS-R technique has been proposed to measure soil moisture using GNSS signals reflected from land surface. It is economical, flexible and able to work all weather and all day. It is a good supplement to current soil moisture observation methods. GNSS signal reflection process over bare land and reflected signal characteristics are explained theoretically. Then retrieval theory is introduced briefly. It can be seen that satellite elevation and azimuth have direct relevance to key parameters involved in soil moisture monitoring, especially surface roughness and antenna gain, which adds complications to the practical application of this technique. Fortunately, BDS GEO satellites have fixed positions relative to earth surface, resulting in constant elevation and azimuth and thus a constant influnce of parameters above. Finally, soil moisture monitoring method using reflected signals of BDS GEO satellites is proposed and its signal processing flow is presented as well. A ground-based experiment was carried out to validate the retrieval method. A RHCP antenna was set up to sky to receive the direct signal while an LHCP antenna was set pointing to the field covered with wheat at initial growth stage. We collected the GPS L1 and BDS B1 signals using a multi-channel GNSS intermediate frequency signal sampler. And the true value of soil moisture was collected by oven drying method once per hour during the experiment period. Then data processing was conducted according to the signal processing flow. Experiment results show an agreement with the true values. Forther more, the results of BDS GEO satellites present a better performance in both accuracy and temporal continuity owing to the fixed position of GEO satellites. We introduce the soil moisture retrieval principle, propose a retrieval method using BDS GEO signals and present the experiment results dedicated to validate method performance. However, priori information of soil composition, surface roughness and antenna gain is needed to obtain a better performance. As result simplified and valid models needs to be further studied. Machine learning algorithms can be applied to this technique, for example.

Extensive amount of water stored in snow covers has a high impact on flood development during snow melting periods. Early assessment of these parameters in mountain environments enhance early-warning and thus prevention of major impacts. Sub-snow GNSS techniques are lately suggested to determine liquid water content, snow water equivalent or considered for avalanche rescue. GNSS antennas are submerged into soil to derive soil moisture. This technique is affordable, flexible, and provides accurate and continuous observations independent on weather conditions. However, the characteristics of GNSS observations for applications within a snow-pack or submerged into water still need to be further investigataed.

The magnitude of the main interaction processes involved for the GPS wavelength propagating through different layers of snow, ice or water is examined theoretically. Liquid water exerts the largest influence on GPS signal propagation through a snow-pack. Therefore, we focus on determining the characteristics of GNSS observables under water.

An experiment was set-up to investigate the characteristics and limitations of submerged GPS observations using a pool, a level control by communicating pipes, a geodetic and a low-cost GPS antenna, and a water level sensor. The GPS antennas were placed into the water. The water level was increased daily by a step of two millimeters up to thirty millimeters above the antenna. Based on this experiment, the signal penetration depth, satellite availability, the attenuation of signal strength and the quality of solutions are analysed. Our experimental results show an agreement with the theoretically derived attenuation parameter and signal penetration depth.

The assumption of water as the limiting parameter for GPS observations within a snow-pack can be confirmed. Higher wetness in a snow-pack leads to less transmission, higher refraction, higher attenuation and thus a decreased penetration depth as well as a reduced quality of the solutions.

In consequence, GPS applications within a snow-pack are heavily impacted by wetness which is even more pronounced during melting period.

Placing the antenna in a fresh water layer as for soil moisture retrievel, a high attenuation of signal strength leads to a signal penetration up to 3.5 centimeter.

In this poster, we present a short introduction to the principle, explain the developed algorithms and show results of experiments dedicated to the signal propagation in water.

Glaciers on the Tibetan Plateau are the main reservoirs of water in the region and much attention is being given to monitor the response of these glaciers to climate variability. Climate forcing varies across the Plateau with the westerlies from inner Eurasia and monsoons from India and East Asia. The surface properties of glaciers characterize the interaction between the atmospheric boundary layer and glaciers. Traditionally, the surface properties of glaciers are observed in situ, but on the Plateau the spatial and temporal variability of such properties cannot be captured by in – situ experiments only. Our main goal is to characterize the surface energy balance of glaciers using a suite of measurements by space – borne observing systems.

Building upon the work done in the previous Dragon investigations, during this 1^st Dragon 4 year we focused on the following aspects:

• Temporal and spatial variability in the albedo of glaciers and their close surroundings
• Monitoring glacier surface flow velocity over an extended period of time
• Estimating and mapping the aerodynamic roughness of glaciers
• Discriminating ice cover types using observations of spectral reflectance and albedo at high spatial resolution
• Relating the albedo of glaciers to depth and duration of snow cover
• Parameterizing glacier albedo in the WRF model

Time series of data products on albedo and net radiation were analyzed to assess the variability in space and time of albedo in glacial areas and to document the corresponding impact on net radiation. Snowfall leads to sudden and large increase in albedo, which may be as high as 0.8 with deep, fresh now. This leads to a change in net radiation from about 800 Wm^-2 to 200 Wm^-2, i.e. an extremely large reduction in radiative load, slowing down snow-melting. The same analysis documented a large spatial variability on both minimum, i.e. glacier, and maximum albedo. Such variability is due to a combination of surficial deposit (ice cover) and terrain, which affects irradiance.

The response of glaciers to climate variability was evaluated in different ways, including the retrieval of the glacier surface velocity using image correlation of paired high spatial resolution images (Landsat TM and OLI). This study focused on the Nyainqêntanglha Range and the period from 1993 to 2015. We analyzed wintertime images at intervals of about one year. The analysis of these time series of ice surface displacement, revealed that the observed signals were a combination of a linear trend and a multi-annual component with variable amplitude from place to place.

Surface features, namely roughness, slope and elevation were retrieved with a combination of ICESat/GLAS and ASTER GDEM data to estimate and map the aerodynamic roughness of glaciers. This property modulates fluxes of latent heat (evaporation, sublimation) and sensible heat. The response of the laser waveform to morphology was studied in detail, showing that roughness and slope of the surface can contribute several meters to even several tens of meters to the pulse shape.

The spatial variability of spectral reflectance and albedo within a glacier was evaluated using Landsat TM and GuoFeng multi –spectral images. Areas covered by debris and lakes or ponds linked with a glacier were clearly delineated, particularly using the Guofeng very high spatial resolution images. Differences in spectral reflectance and albedo were rather large and the consequences in terms of spatial variability, within a glacier, of radiative forcing are being evaluated.

The work on time series of glacier albedo revealed large and rapid variations in the radiative forcing on glaciers. These variations are not reproduced by the land surface schemes currently implemented in advanced atmospheric models such as WRF. We have combined WRF snow depth with MODIS albedo to parameterize the dependence of albedo on snow depth and snow age. This parameterization has been evaluated against in – situ measurements at the Plateau permanent observatories.

The parameterization constructed in this way has been implemented in WRF and the sensitivity of the WRF land surface energy balance has been evaluated, showing rather large impacts on both radiative and convective fluxes.

One great challenge of cryosphere and hydrosphere science in high elevation regions is the scarcity and sparseness of data on the multiple variables and processes relevant to the understanding of the water cycle.

Quantitative information on water losses is important to understand the global terrestrial water cycle and land – atmosphere interactions. The global evapotrasnspiration in 2008-2013 with a spatial resolution of 1 km was determined using ETMonitor as the sum of the evapotranspiration components, i.e. plant transpiration, soil evaporation, open water evaporation, rainfall interception, snow and ice sublimation. All these variables were retrieved using the ETMonitor model driven by multiple satellite data products. The ASCAT (Advanced Scatterometer) soil moisture data product was applied as a key input to scale ET between 0 and ET_max. To allow the estimation of ET at high spatial resolution, the 0.1° resolution ASCAT data product was downscaled to 1km spatial resolution globally using bilinear resampling method. We have developed a different, bio-phyisical, downscaling procedure applicable for regional studies and based on high resolution surface temperature and vegetation index data products. The estimated water losses agreed well with the in situ tower based observations at a number of FLUXNET sites, with high correlation, low bias, and low root mean square error. The retrieved ET captures the expected global patterns and the details of spatial and temporal patterns were consistent with the current available global evapotranspiration products such as data from GLEAM (Global Land Evaporation Amsterdam Model) and the GLDAS (Global Land Data Assimilation System) Noah product. The ETMonitor data product is superior due to the high spatial and temporal resolutions. We have also experimented with the ESA-CCI (European Space Agency - Climate Change Initiative) soil moisture data product to replace the ASCAT one. A first evaluation was carried out by retrieving ET in China during 2001-2015 and the results used to contribute to the National Remote Sensing Monitoring for Sustanable Devepoment Report in China (2016).

To improve the accuracy of ETMonitor in cold and high elevation regions, the algorithm to estimate snow and ice sublimation was improved, by adapting the Penman – Monteith combination equation. This method was evaluated against eddy – covariance measurements of latent heat flux at high elevation sites in the Heihe river Basin N – W China: on average the RMSE was 8.75 W m^-2. We have also evaluated the bulk aerodynamic (BA) method against the same measurements and the BA performance was slightly worse. The main driver of the P – M equation is net radiation, which is very variable at high elevation due to the variability of albedo, which enhances the scope of using satellite observations to estimate and monitor sublimation. A case – study on the upper reach of the Heihe River Basin has been carried out using MODIS data products on surface albedo and temperature.

Work contiuned on improving other algorithms and data products. An algorithm to retrieve the total precipitable water was developed, based on the ratio of brightness temperature changes ΔTb_18.7/ ΔTb_23.8 , using atmospheric profiles obtained from the globally distributed radiosonde observations and applying a microwave radiative transfer model. This algorithm can retrieve total precipitable water under both clear and cloudy sky condition over land, and can be easily transferred to MWRI on board the FY-3 satellites. To retrieve the surface freeze and thaw condition, an innovative freeze/thaw index based on microwave observations at 18.7 and 36.5 GHz was defined and assumed to be linearly correlated with the radiometric land surface temperature retrieved with thermal infrared observations. It was found that this linear relationship is quite reliable for most areas, and can provide high-resolution information on near surface soil freeze/thaw state. The validation of the high-resolution freeze/thaw state against soil temperature measured at active layer monitoring sites along the Qinghai-Tibet Highway illustrated a moderate accuracy over a decade scale.The daily snow cover fraction at 500m resolution was retrieved in the Tibet Plateau in 2013, and missing values due to the cloud cover were filled to obtain a cloud-free dataset.

Mun river basin is largest basin in Mekong river region, where both flood and drought frequently occur. Water management by hydrological model is difficult due to lack of accurate and spatial/temporal continuous products to force and localize the model. The new Precipitation and soil microwave products gradually plays important role for hydrological modelling. The study introduced the precipitation product at 0.1 degree spatial resolution (GPM mission) for SM simulation by Variable Infiltration Capacity (VIC) model. Then we assessed accuracy of 10cm soil moisture simulation by comparison with SMAP soil moisture product at a spatial resolution of 9km. The results show that:(1) GPM precipitation can be compared with gauge-based monthly precipitation in Mun river;(2) GPM products can be used to simulate spatial pattern of soil moisture in dry season; the simulated soil moisture has a high accuracy in the upperstream of basin but poor in the downstream due to spatial pattern of irrigation application in Mun river basin. It suggests that based on GPM and SMAP products, we can make model calculation and cross validation by remote sensing technology. Therefore microwave remote sensing products are expected to be introduced for water management in remote or data-lacking regions.

Soil hydrologic parameterization, as well known control the energy and water fluxes of hydrologic basin surface playing a crucial role in hydrological model simulation for operative application in the field of water engineering. Despite their importance their definition for large areas is always a source of uncertainty due to difficulty of representative ground measurements , their spatial variability also strongly affected by land use change and agricultural practices.

In the framework of Dragon 4 Project “Forcing, calibration, validation and data assimilation in basin scale hydrological models using satellite data products”, the paper presents a procedure for soil hydrologic parameterization based on the assimilation of satellite LST data into a thermodynamic distributed water balance model (FEST-EWB). The model algorithm solves the system of energy and mass balances in terms of a representative equilibrium temperature (RET), that is the land surface temperature that closes the energy balance equation and so governs the fluxes of energy and mass over the basin domain. This equilibrium surface temperature, which is a critical model state variable, is comparable to LST as retrieved from operational remote sensing data from MOST, ESA and NASA agencies. This approach will be compared with traditional ones based the pixel wise use of pedo transfer function, calibrated for available soil maps.

The case study is the upper part of the Heihe River basin where a consistent historical data set will allow to test this approach.

A simple two-source evapotranspiration (ET) model was applied to the Yingke and Daman irrigation districts of the Zhangye Oasis, which is located in the middle reaches of the Heihe River, China. The ET model was composed of two parts, including an evaporation (E) sub-model and a transpiration (T) sub-model. A separated parameter estimation scheme was conducted using Bayesian inference. First, an empirical multiplier was estimated for an E sub-model using observations that were collected after crop harvests. The empirical multiplier was then assigned to the most-likely value in the simple two-source ET model. Second, a global sensitivity analysis was performed to identify the key parameters that were responsible for most of the variability in the λET results within the T sub-model. To avoid equifinality or over-parameterization, Bayesian inference was applied to estimate the key parameters that induced the most variability in the first set. A second set of Bayesian inference was then performed by fixing the most-likely values of these parameters, and the other parameters were defined one-by-one as Bayesian parameters. These parameters were estimated for seven sites. The coefficient of determination for the modeled λET and the observed values exceeded 0.9. Next, a cluster analysis was conducted using the canopy height, leaf area index and soil moisture content to classify the fields with the highest similarities and then to distribute the same parameter values to similar fields. Finally, λET was estimated using the most-likely values of the parameters at the regional scale. The root-mean-square error of the remotely sensed estimates was less than 20 Wm^-2, the mean absolute percent error did not exceed 4%, and the correlation coefficient was greater than 0.97. The validation was conducted for both the modeled λET at the point scale and for the remotely sensed λET at the satellite pixel scale. The results demonstrate that using cluster analysis, the most-likely values of the parameters can be effectively applied to estimate remotely sensed λET.

The effect of climate change on snow is complicated at regional scale. The high spatio-temporal resolution snow related variables simulated by weather research and forecast model including snowfall, snow water equivalent and physical snow depth and the high spatial resolution fractional snow cover data extracted from MODIS/Terra are adopted to evaluate the effect of climate change on snow over the Heihe River Basin (HRB) in last 15 years using Empirical Orthogonal Function (EOF) analysis and Mann-Kendall / Theil-Sen trend analysis. The results indicate 1) Due to the air temperature increasing, the fractional snow cover, snow water equivalent, physical snow depth over the whole HRB region decrease in last 15 years, especially at the height over 4500 m, however, the snowfall increases at mid-altitude ranges over the upstream of HRB. 2) Over the upstream of HRB, the total snow flux increased, however, the number of snowfall days decreased in last 15 years, so the occurrence of extreme snow events over the upstream of the HRB might increase. 3) The air temperate over the downstream increased at the most of the HRB, however, the snowfall over this region decreased in last 15 years, so the weak ecological system over the downstream may be exacerbated in future.

In contrast to the glacier mass losses observed at other locations around the world, some glaciers in the High Mountains of Asia appear to have gained mass in recent decades, which was called as ‘Karakoram anomaly’ or ‘Karakoram-Pamir anomaly’. Recent study to laser altimetry data found the centre of the anomaly might locates at the West Kunlun instead of the Karakoram. We performed differential interferometry to 14 pairs of bistatic TerraSAR-X and TanDEM-X observed the West Kunlun and its surroundings obtained at ~2013 by referring to SRTM observed in 2000. After removing seasonal effect and penetration depth differences, it found during 2000 and 2013, glacier mass balance rate at the West Kunlun was 0.128 ± 0.055 m w.e.a^-1 and most of its surrounding area also experienced a mass gain that varied from 0.043 to 0.363 m w.e.a^-1, with a decreasing gradient from northeast to southwest. At southwest of this study region, glacier presented significant mass lost at -0.286 ± 0.067 m w.e.a^-1. For the West Kunlun region, northern slope gained mass quicker than southern slope, eastern and western part gained mass quicker than central part. Comparing to previous studies applied ICESat satellite laser altimetry data, similar results of glacier height changing was obtained at their footprints. Our results suggested a decreasing gradient of glacier mass balance from the Upper Tarim edge to the Karakoram, which is similar to previous ICESat derived gradient rather than topographic difference results derived with SPOT/HRS and SRTM. Glacier surging was common at the West Kunlun. Surging and quiescent glaciers identified by glacier height changing pattern was almost the same to previous study derived with feature tracking. They cover almost one third of the total glacierized region. For the West Kunlun, glacier height changes in different elevation bins for non-surging glaciers present significant and homogeneous height increasing above 5450 m, while below 5400 glaciers shows significant thinning, which indicate the warming and moisturizing trend in the centre of the anomaly area.

Glacier is one of the most important climate change indicators in both regional and global scale. High Mountain Asia has the largest extend of glacier outside the polar region. As the Asian Water Tower, glacier in this region also has distinct ecological significance due to its vast water source. Its movement has close correlation with the risk of glacier lake outburst in highland. As a consequence, glacier dynamics in High Mountain Asia including mass balance, surface velocity and outline detection have been research hotspot. In recent years, synthetic aperture radar (SAR) observation has been regarded as an effective tool for glacier surface motion monitoring with wide range and high resolution. Therefore, it is essential to employ SAR observation to evaluate the velocity of glaciers in Tibet, China, which may be the basis for glacier dynamics and even climate change conditions in High Mountain Asia region.

SAR echo records both amplitude and phase information. From one aspect, the phase-based traditional differential SAR interferometry (D-InSAR) can achieve a millimeter accuracy in deformation detection theoretically while its application to glacier dynamics is generally limited by decorrelation. From another aspect, the intensity-based pixel offset tracking (POT), taking advantages of the intensity from backscatter signals, can implement large displacement in both range and azimuth direction whereas redundancy and error estimation in matching decorrelated patches arises in the calculation procedure. Hence an integrated method combining these two complements is applied to sub-region in High Mountain Asia in exploring glacier dynamics which can improve utilization of SAR observation.

In this study, acquisitions from Sentinel-1A and Sentinel-1B constitutes the data stack with a 6-day temporal baseline. Qualitative and quantitative evaluations of the glacier velocity are made for understanding the surface motion and glacier dynamics. Compared with other space-borne SAR acquisitions, the C-band Sentinel-1 data have smaller temporal decorrelation effects with shorter revisit time, which might be attributed to the good interferometry analysis.

Rockglaciers are the best visual expression of mountain permafrost and are widespread in the Tien Shan. These ice-debris landforms can, in contrast to permafrost itself, be mapped and monitored directly using remotely sensed data. Previous studies showed that changes in rockglacier flow can be related to climate conditions. However, no consistent rock glacier inventory of the whole Tien Shan exists and information about rockglacier flow is rare. Most previous studies concentrated in a few valleys in the Ile Range of Northern Tien Shan (Kazakhstan).

We have systematically mapped active rock glaciers of Northern Tien Shan located in Kazakhstan, Kyrgyzstan and Xinjiang, China’s north-western-most province, based on differential SAR interferograms and the best available optical imagery from Google Earth or other sources. Different SAR interferograms from various sources, including ERS-1/2, ALOS-1/2 and Sentinel-1, were used to identify and manually map surface deformations at elevations where rockglaciers can occur. The optical imagery were subsequently applied to distinguish rockglaciers from other deformations, e.g. due to subsidence, landsliding or solifluction. The rockglaciers were finally classified according to their state of activity (surface velocity), origin of the debris and topographic parameters (e.g. aspect, slope).

We identified so far more than 700 objects with an extent of about 250 km² within an area of 4000 km². Most of the rockglaciers are moraine-derived and have a northern exposition. The altitude distribution varies significantly depending on the location. Work is ongoing to extend the study region, refine the inventory, classify the rock glaciers according to their activity, and investigate the changes in velocity and surface elevation of selected rockglaciers over time.

Global warming caused significant changes in mountain glaciers. Observations showed on average clear glacier mass loss. However, recent studies revealed also regions with balanced mass budgets especially in parts of High Mountain Asia (HMA). These heterogeneous changes significantly influence the hydrology, i.e. regionally they alter the river run-off and cause the rise of endorheic lakes on the Tibetan Plateau and globally they affect the sea-level. Glaciers are a major contributor of sea-level rise and affect population that rely on fresh water from glaciers. Rock glaciers have so far only rarely been investigated in HMA but may also of hydrological importance.

The purpose of this sub-project within the more general cryosphere project is to generate an up to date glacier and a rock glacier inventory for selected benchmark regions located in different climatic settings in HMA. Glacier mapping will be based both on optical and radar imagery distributed by the Chinese and European Space programs and combine information about surface flow (as derived in the subproject 2), surface reflectance and backscatter. The generated outlines will be compared to existing ones of previous periods to detect changes in glacier area. As area changes can only indirectly related to climate and hydrology, we will also investigate glacier mass changes using digital elevation models from different time periods (DEM differencing) and altimetry data. Both data sources are complimentary in regard to their spatiotemporal coverage. We will apply ICESat and Cryosat-2 altimetry data to investigate the trends over the whole of HMA and apply DEM differencing in the benchmark regions using existing DEMs (e.g. SRTM, ASTER or TanDEM-X DEMs) or DEMs derived from stereo data. Field measurements and high resolution data will be employed to validate and calibrate the remote-sensing derived results.

The outcome of this project will be improved methodologies for glacier mapping and glacier change assessments and a better knowledge about rock glacier occurrence, the spatial and temporal variability of glacier area and mass changes in HMA, its influence on hydrology and its control by local and climatic forcing. This will be realised thanks to the large archive of satellite data available via Dragon, data available from other sources and thanks to the coordinated effort of the several institutions partnering in the project. The link to hydrology, local and climate forcing will be investigated within this subproject via data assimilation into mass balance models, interaction with the two other sub-projects under the umbrella of cryosphere, and via interaction with the hydrology consortium.

Glacier fluctuations are regarded one of the most significant indicators of climate change. The expansion and contraction of glaciers can be observed by outlining glacier boundaries or measuring snow lines with optical Earth observation satellites. New satellites, such as Sentinel-2A/B, provide high spatial resolution images and short revisit times that can be used to make ample measurements to accurately determine glacial variability. Likewise, ever increasing volumes of satellite data make automated boundary and snow line detection a desirable solution for researchers. Two regions of interest for boundary and snow line detection are the Himalaya and Tibetan Plateau. They are home to the world's highest mountains and some of the world's largest non-polar glaciers. These regions also provided valuable water resources to over a billion people in nearby countries, and therefore are not just ecologically, but also economically important. Clouds, however, present a challenge to obtaining useful image data. Mountainous regions are often surrounded or covered by clouds. Clouds can be a menacing phenomenon in remote sensing because they greatly attenuate and reflect short wavelengths used by optical Earth observation satellites. Currently, many techniques exist to automatically detect clouds and classify them. However, they are not perfect. Many techniques have encountered difficulties in cases where snowy and icy landscapes share similar properties with clouds. The Himalaya topographic relief also adds to the challenge. Steep slopes and topographic shadows have a profound effect on surface reflectances and can lead to misclassifications. To address these issues, this study presents an assessment of multiple cloud detection techniques. For initial analysis, 19 Sentinel-2A images were acquired at various times between 12/07/15 and 31/12/16. The images are centered on the Bara Shigri Glacier in Himachal Pradesh, India, a large 10 km long glacier which drains into the Chenab River (an Indus River tributary). The images vary from cloudy to clear (cloud-free), but also have variations in snow cover and cloud shadows. The set was reduced to 6 images that were selected for further classification analysis, as partly shown in Figure 1. Figure 1 shows, starting from the top-left, a natural color image for reference, a manually created cloud mask for validation, and well-known spectral analysis methods: Fmask and maximum likelihood classification. Goal of the study is to first evaluate the performance of existing methods in the automatic identification of pixels contaminated by clouds, and secondly, if necessary, design an improved method, for example, by incorporating the high temporal revisit time of the Sentinel-2 imagery. In doing so, this research seeks to better understand cloud cover over mountainous regions and distinguish them from snow cover and glaciers. It should be noted that cloud detection is a valuable pre-processing step that, when successful, will increase data availability to glacier researchers. Thus, the final goal is to incorporate the cloud-free pixel identification in automated workflows for snow cover studies.

Catastrophic event - fast moving glacier destroying pastures and killing livestock near the glacier tongue in Kongur Mountains, Xinjiang Uygur Autonomous Region, was reported on May 14^th, 2015. In this letter, Sentinel-1 SAR and ALOS/PALSAR data feature-tracking was employed to obtain the glacier surface velocities. Time series of the estimated glacier surface velocities suggested that the left tributary had been a fast moving flow for eight years at least. Analysis of the obtained glacier surface velocities Variation made the surge occurrence confirmed, and from Mar 24th, 2015 to May 11th, 2015, the left tributary has pushed the trunk gradually detected by the analysis of glacier flow vector. Surface velocities of three stages, prior to the surging, during surging and post surging were mapped and analyzed. This research suggested that monitoring glacier surface velocities could be regarded as an effective way for glacier catastrophic warning.

The mass balance history (1980–2010) of a monsoon-dominated glacier in the southeast Tibetan Plateau is reconstructed using an energy balance model and later interpreted with regard to macroscale atmospheric variables. The results show that this glacier is characterized by significant interannual mass fluctuations over the past three decades, with a remarkably high mass loss during the recent period of 2003–2010. Analysis of the relationships between glacier mass balance and climatic variables shows that interannual temperature variability in the monsoonal season (June–September) is a primary driver of its mass balance fluctuations, but monsoonal precipitation tends to play an accentuated role for driving the observed glacier mass changes due to their covariation (concurrence of warm/dry and cold/wet climates) in the monsoon-influenced southeast Tibetan Plateau. Analysis of the atmospheric circulation pattern reveals that the predominance of anticyclonic/cyclonic circulations prevailing in the southeastern/northern Tibetan Plateau during 2003–2010 contributes to increased air temperature and decreased precipitation in the southeast Tibetan Plateau. Regionally contrasting atmospheric circulations explain the distinct mass changes between in the monsoon-influenced southeast Tibetan Plateau and in the north Tibetan Plateau/Tien Shan Mountains during 2003–2010. The macroscale climate change seems to be linked with the Europe-Asia teleconnection.

The Tibetan Plateau (TP), the highest and largest plateau in the world, with complex and competing cryospheric-hydrologic-geodynamic processes, is particularly sensitive to anthropogenic warming. The quantitative water mass budget in the TP is poorly known. Here we examine annual changes in lake area, level, and volume during 1970s −2015. We find that a complex pattern of lake volume change during 1970s−2015: a slight decrease of –2.78 Gt yr^-1 during 1970s−1995, followed by a rapid increase of 12.53 Gt yr^-1 during 1996−2010, and then a recent deceleration (1.46 Gt yr^-1) during 2011−2015. We then estimated the recent water mass budget for the Inner TP, 2003−2009, including changes in terrestrial water storage (TWS), lake volume, glacier mass, snow water equivalent (SWE), soil moisture, and permafrost. The dominant components of water mass budget, namely changes in lake volume (7.72 ± 0.63 Gt yr^-1) and groundwater storage (5.01 ± 1.59 Gt yr^-1), increased at similar rates. We find that increased net precipitation contributes the majority of water supply (74%) for the lake volume increase, followed by glacier mass loss (13%), and ground ice melt due to permafrost degradation (12%). Other term such as SWE (1%) make a relatively small contribution. These results suggest that the hydrologic cycle in the TP has intensified remarkably during recent decades.

This paper presents a new water and energy budget-based glacier mass balance model. Enthalpy, rather than temperature, is used in the energy balance equations to simplify the computation of the energy transfers through the water phase change and the movement of liquid water in the snow. A new parameterization for albedo estimation and state-of-the-art parameterization schemes for rainfall/snowfall type identification and surface turbulent heat flux calculations are implemented in the model. This model was driven with meteorological data and evaluated using mass balance and turbulent flux data collected during a field experiment implemented in the ablation zone of the Parlung No. 4 Glacier on the Southeast Tibetan Plateau during 2009 and 2015–2016. The evaluation shows that the model can reproduce the observed glacier ablation depth, surface albedo, surface temperature, sensible heat flux, and latent heat flux with high accuracy. Comparing with a traditional energy budget-based glacier mass balance model, this enthalpy-based model shows a superior capacity in simulation accuracy. Therefore, this model can reasonably simulate the energy budget and mass balance of glacier melting in this region and be used as a component of land surface models and hydrological models.

Most estimates of melt and surface sublimation rates for the glaciers of the Tibetan Plateau have been obtained at the point scale, and it is not entirely established how ablation components vary across the entire extent of a glacier in this environment. Energy balance models can provide accurate simulations of energy fluxes and ablation components, thus fostering understanding of the main processes at the glacier-atmosphere interface, but need a high number of meteorological and surface input variables that can be available at the point scale of Automatic Weather Stations but are difficult to extrapolate across the distributed domain of an entire glacier.

Here, we simulate the distributed energy and mass balance of Parlung No. 4 Glacier, for the ablation season 2016, using a distributed energy balance model. We use input data at one AWS and two novel methods, which combine in-situ and reanalysis data, to generate fields of near-surface air temperature and wind speed that are needed to force the model. We calculate the spatial distribution of energy fluxes and ablation rates over the glacier surface at a high spatial resolution (50 m), which allows the inclusion of small-scale processes, such as katabatic and valley winds, refreezing and topographic shading, and knowledge of the local variability of topographic parameters, such as sky-view factors and local slopes.

We also compare our model results with the simulations of a point-scale energy-balance model based on enthalpy calculations which includes advanced schemes for calculation of albedo and turbulent fluxes; and with large-scale, regional energy balance simulations driven by satellite input data. The model is validated with in-situ and satellite observations of ablation, surface temperature and turbulent fluxes.

Our main objectives are to: i) advance our understanding of glacier ablation in the Tibetan Plateau by providing one of the first spatially-distributed quantifications of energy fluxes and ablation rates on a glacier in this region, including understanding the role that surface sublimation plays; ii) explore the use of two relatively new methods to generate fields of air temperature and wind speed, which provide alternatives to solve some of the interactions between large-scale meteorological forcing and the atmospheric boundary layer over a mountain glacier during the summer period; and iii) understand the level of model complexity that is needed for accurate simulations of energy fluxes and ablation components at different spatial scales.

The project is focusing on monitoring water resources in the Red River basin using microwave remote sensing. The Red River, also known as the Yuan River in Chinese, is a trans-boundary river basin with its total area of 169,000km2 shared by Vietnam (51%), China (48%) and Laos (1%). The Red River basin has a tropical or subtropical climate, dominated by the southwest monsoon from May to September and the northeast monsoon from October to April. The monsoon results in massive flow volume fluctuations. Flooding is a significant problem during the rainy season, particularly in July and August. At the same time, large area of karst landform causes the river flow loss and water shortage. One of the greatest challenges for flood prediction and integrated water management in the Red River basin is a lack of information on reservoir management as a consequence it is not easy to estimate the water resources. Since it is a trans-boundary river, there are difficulties to manage the area as a whole, and the information might not be in time for flood and drought early warning.

Hydrological data on major rivers, lakes and wetlands can often be difficult to obtain due to a region's inaccessibility, sparse distribution of gauge stations or the slow dissemination of data. Remote sensing technologies can be used to overcome such shortcomings.

The main objective of this project is to develop the algorithms and synergies between different Microwave Remote Sensing sensors to be able to monitoring water resources in the Red River Basin. For that purpose, the water elevation information, precipitation, soil moisture and evapotranspiration by remote sensing will be integrated in a hydrological model to improve its accuracy.

In this presentation we will be showing the first results of the variables monitored during this first year.