To promote the radiometric quality of satellite remote sensing products and assure the comparability of product quality amongst multi-series satellites, we should build a continuous transfer chain from remote sensor to the radiometric reference standard, estimate those parameters characterizing sensor’s performance, and conduct quality controls on remote sensing data and products during sensor lifetime. However, when the sensor is on-orbit, its radiometric quality is usually hard to be traced to SI because of breakage of the reference transfer chain. The traditional field vicarious radiometric calibration method which uses ground target measurement value as radiometric reference, can be influenced by various uncertainties due to scaling effect, atmospheric condition, environment change, etc., so it is hard to reach high calibration accuracy. In pursuit of on-orbit optical sensor high-accuracy calibration and product quality consistence, we carried out the following exploratory work:
1) Research on the ground-based radiometric reference transfer chain in sensor calibration under the framework of RadCalNet (the Radiometric Calibration Network, initiated by CEOS/WGCV). The RadCalNet initiative was sponsored cooperatively by NASA, ESA, CNES and AOE, with object to verify EO sensor’s radiometric calibration accuracy and ensure radiometric measurement consistency from different EO sensors. Through the activities of establishment of calibration sites, increasing observation frequencies, decreasing of total calibration errors, RadCalNet ensures the traceability of radiometric measurement benchmark for space remotely sensors. And moreover, it can provide the opportunities to carry out on-obit calibration for those space agencies who do not have the conditions for organizing the in-situ calibration activities. Under the framework of RadCalNet, AOE designed and manufactured automated surface spectrum measurement system. Now, this system has been deployed on different permanent targets in Baotou site. It can automatically measure the surface reflected radiance and atmospheric parameters continuously. Then, following the requirement of data products defined by RadCalNet data center, Level 1 BOA (bottom of atmosphere) reflectance standard products and level 2 TOA (top of atmosphere) sensor-independent reflectance products are generated automatically. Besides, AOE and NPL carried out cooperation research on uncertainty analysis of the whole chain of radiometric calibration benchmark product generation for Baotou site. In this study, the contributions to the total uncertainty for different factors, including obtaining of surface reflectance, homogeneity and BRDF of surface target, measurement of atmospheric parameters, atmospheric radiative transfer calculation, solar irradiance spectra model, and etc., have been elaborately analyzed to calculate the uncertainty of radiometric calibration in Baotou site.
2) Research on space radiometric benchmark transfer calibration. Introducing radiometric benchmark sources, which can be traced to SI, in some specific benchmark satellite, via synchronously observation of surface scenes between benchmark satellite and other satellite, is a feasible way to transfer the space radiometric measurement benchmark from spaceborne standard radiometric calibration system to operational optical remote sensing satellite with high accuracy. Under this framework, we can provide high accurate and consistent space radiometric measurement benchmark for those optical satellite, to significantly promote calibration accuracy and ensure the consistency and stability of data quality amongst multi-series satellites. Based on this idea, we design a space radiometric benchmark transfer calibration plan with accurate spatial-temporal-spectral-angular matching and coupling transform as core techniques. In this plan, we comprehensively use the ground surface scene model, atmospheric radiative transfer model and optical sensor model, build the satellite-satellite radiometric measurement mapping relationship, carry out the spatial-temporal variety and anisotropy characteristic analysis for surface stable targets, establish the constraint mechanism in spatial-temporal-spectral-angular matching, construct the satellite-satellite radiometric measurement transform model though simulation and observation data, to finally realize space radiometric benchmark transfer calibration.
The above two researches have been validated and the superiority of each method have been analyzed with Chinese and European satellite, such as Sentinel-2a/2b、GF-1/2、ZY301/02, with ground-based and aerial-based campaign. This work points out the direction of the next step.

Long-term aerosol data records provide information on changes in the aerosol properties due to both natural (meteorological, climatological, dust storm, fires) and anthropogenic effects (such as industrialization, urbanization and policy measures aimed at improvement of air quality). An often-used indicator for the aerosol burden is the aerosol optical depth (AOD), i.e. the column-integrated extinction coefficient. AOD can be monitored from satellites using radiometers, but instruments used for this purpose have a limited lifetime and often are not designed for aerosol monitoring. As part of the Aerosol-cci project, the Along Track Scanning Radiometer (ATSR-2) flying on the European Space Agency (ESA) ERS-2 satellite from 1995 to 2003 and the advanced ATSR (AATSR) flying on ESA’s ENVISAT (2002-2012) were used together to create a 17-years (1995-2012) global AOD data record over land and ocean (Popp et al., 2016). This data set is planned to be extended with AOD retrieved from the Sea and Land Surface Temperature Radiometer (SLSTR), an instrument similar to ATSR but with a backward instead of forward view, the first of which flies on Sentinel-3A launched in early 2016. However, this leaves a gap of about 4 years between the end of the AATSR and the start of the SLSTR data records. To fill this gap, we investigated the use of AOD data available from the application of the ALAD algorithm developed by RADI for AOD retrieval over land using data from the Advanced Very High Resolution Radiometer (AVHRR) (Xue et al., 2017). ALAD combines eight different AVHRR instruments to produce an AOD time series starting in 1983 (Xue et al., 2017). Hence, the ALAD AOD data set could also be used to extent the information from the ATSR data record backward from 1995 to 1983, provided that a satisfactory match can be obtained between the overlapping ATRS and AVHRR data sets.

In this study we used ATSR AOD data produced with FMI’s Aerosol Dual View (ADV) algorithm (Kolmonen et al., 2016; Sogacheva et al., 2017) and ALAD data sets retrieved for the whole period from 1983 to 2014 over the North China Plain (NCP). In addition, MODIS-Terra AOD C6.1 data are used for comparison and ground-based sun photometer AOD data from AERONET (Holben et al., 1998) are used as reference. The validation versus AERONET shows the good performance of the AVHRR AOD up to about 0.5 and for ATSR up to about 1.3, which leads to large differences during high AOD episodes such as often observed over the NCP in the summer. However, during the winter, when AOD is often moderate, AVHRR provides better coverage. Part of the difference between AVHRR and ATSR AOD may be explained by the difference in wavelength between the ATSR- and AVHRR-retrieved AOD (550 nm and 630 nm, respectively).

References

Kolmonen, P., Sogacheva, L., Virtanen, T.H., de Leeuw, G. , and Kulmala, M.: The ADV/ASV AATSR aerosol retrieval algorithm: current status and presentation of a full-mission AOD data set, International Journal of Digital Earth, 9:6, 545-561, doi: 10.1080/17538947.2015.1111450, 2016.

Popp, T., de Leeuw, G., Bingen, C., Brühl, C., Capelle, V., Chedin, A., Clarisse, L., Dubovik, O., Grainger, R., Griesfeller, J., Heckel, A., Kinne, S., Klüser, L., Kosmale, M., Kolmonen, P., Lelli, L., Litvinov, P., Mei, L., North, P., Pinnock, S., Povey, A., Robert, C., Schulz, M., Sogacheva, L., Stebel, K., Stein Zweers, D., Thomas, G., Tilstra, L.G., Vandenbussche, S., Veefkind, P., Vountas, M., and Xue, Y.: Development, production and evaluation of aerosol Climate Data Records from European satellite observations (Aerosol_cci), Remote Sens. 2016, 8, 421; doi:10.3390/rs8050421, 2016.

Sogacheva, L., Kolmonen, P., Virtanen, T. H., Rodriguez, E., Saponaro, G., and de Leeuw, G.: Post-processing to remove residual clouds from aerosol optical depth retrieved using the Advanced Along Track Scanning Radiometer, Atmos. Meas. Tech., 10, 491-505, doi:10.5194/amt-10-491-2017, 2017.

Xue, Y., He, X., de Leeuw, G., Mei, L., Che, Y., Rippin, W., Guang, J., Hu, Y. : Long-time series aerosol optical depth retrieval from AVHRR data over land in North China and Central Europe. Remote Sensing of Environment, 198: 471-489, 2017.

Different from the work last year, the paper develops a all-weather and all-day passive sub-millimeter precipitation retrievals algorithm for Microwave Humidity and Temperature Sounder (MWHTS) onboard the Chinese Feng Yun 3C (FY-3C) satellite. The retrieval algorithm employs a number of neural network estimators trained and evaluated using the validated global reference physical model NCEP/WRF/ARTS, and works for seawater. NCEP data per 6 hours are downloaded to run the Weather Research and Forecast model WRF, and derive the typical precipitation data from the whole world. The Atmospheric Radiative Transfer Simulator ARTS is feasible for performing simulations of atmospheric radiative transfer. Rain detection algorithm has been used to generate level 2 products. Retrievals are reliable for surface precipitation rate higher than 0.1 mm/h at 15km resolution, which is in good agreement with those retrieved using the Precipitation retrieval algorithm version 1(ATMP-1) for Advanced Technology Microwave Sounder (ATMS) aboard Suomi NPP satellite.

Meanwhile, calibration and validation between similar instruments onboard different satellites are also important to ensure the validity of observations and accuracy of precipitation retrievals. In the ongoing work, we are going to carry out the calibration and validation among FY-3C MWHTS, FY-3B MWHS and ATMS，and some preliminary results can be shown in the conference materials.

A ground-based MAX-DOAS and a Bruker IFS 125HR have been deployed in Xianghe Station, Northern China, of the Institute of Atmospheric Physics, Chinese Academy of Sciences, and another Bruker IFS 125M has been installed in Xinglong Station. The MAX-DOAS has been running for more than ten years, providing a large number of high quality data of NO₂, SO₂, etc., for deriving their trends, and for validating the satellite products of OMI, GOME-2, and SCIMACHY. In Xianghe station, CIMEL sunphotometer, gas analyzers, automatic meteorological station, and a 100-meter tower can provide aerosol optical properties, air quality status, and meteorological conditions in the planet boundary layers. The two Bruker FTIR instruments in Xianghe and Xinglong stations aim at providing the greenhouse gas such as CO₂, CH₄, N₂O, and for validating GOSAT, OCO-2, and TanSat products in future. The FTIR in Xinglong station has been operating for more than one year, and some data has been used for validation of GOSAT products. Some new progresses in validating satellite products over northern China using ground-based FTIR and MAX-DOAS instruments will be reported here.

Atmospheric aerosol optical remote sensing makes use of ultraviolet, visible and infrared sensors to collect information of the particles in atmosphere by detecting radiation scattered from targets. Black carbon (BC) is the most important light-absorbing aerosol in the current atmosphere because of its strong positive climate forcing from direct radiative and snow albedo effects. Both effects are significantly affected by BC optical properties. Observations have shown that BC particles have complex structures due to stochastic aggregating. Thus, a reliable remote sensing of BC aerosols and estimate of BC climatic effects requires accurate computations of optical properties for BC particles with complex structures. Currently most simulation methods employ single-sized spherical particle as primary spherule to construct the whole aggregates, this may introduce some extra uncertainty. In this study we use DDA method to compute optical properties of aggregates constructed by nonspherical and spherical particle with same structural parameters. Preliminary results are presented and show great differences exist between them, the results could be used for evaluating uncertainty introduced by particle modelling and calibrating results of remote sensing.

The Baotou site is one of four instrumented sites being established as part of the Radiometric Calibration Network (RadCalNet). RadCalNet is an initiative of the CEOS WGCV. It has been designed to provide satellite operators with SI (Système International d'Unités)-traceable top-of-atmosphere (TOA) spectrally-resolved reflectances from a coordinated network of instrumented land-based test sites. An automated radiometric calibration system was established on the Baotou Cal&Val test site in China to provide an operational high-accuracy and high-stability vicarious calibration and validation site for high resolution remote sensing instruments.
The composition of the system is first described in this paper, which provides bottom-of-atmosphere (BOA) reflectance in 30 minutes intervals in 10 nm steps over the spectral range from 400 nm to 1000 nm. The sources of uncertainty from the laboratory calibration of the spectrometer, through to field observations of radiance and the calculation of ground reflectance, through to the propagation to TOA reflectance were analyzed in this paper. An initial uncertainty analysis of the automated radiometric calibration was made by considering three aspects: uncertainty associated with the spectrometer (ground viewing instrument) measurements; uncertainty associated with the BOA reflectance calculation; and the uncertainty associated with averaging this to the requirements of the Radiometric Calibration Network, RadCalNet. Preliminary analysis shows that the combined uncertainty associated with RadCalNet BOA reflectance would be approximatively 3.0 % within the spectral range 500 nm ~ 900 nm. The uncertainty associated with the TOA reflectance propagated from BOA reflectance measured by single observation is approximatively estimated as 3.5% ~ 5.0%, ignoring the influence of the surrounding pixels.

Here we applied our tropospheric nitrogen dioxide (NO2) retrieval algorithm, which was implemented for the Chinese Environmental trace gases Monitoring Instrument (EMI), to the TROPOspheric Monitoring Instrument (TROPOMI). Generally, the Differential Optical Absorption Spectroscopy (DOAS) technique was used to retrieve slant column densities (SCDs) of NO2, and air mass factor (AMF) are calculated for light path correction. The solar and viewing geometries, surface albedo and pressure, cloud properties, and modelled gas profile was used as input parameters for online AMF calculations. The tropospheric NO2 was estimated from the total column by using the modified reference sector method. Results show good agreements with independent ground-based MAX-DOAS measurements.