Conference Agenda

The monitoring and forecasting of global water cycle under climate changes indeed require enhancement of satellite remote sensing products in both of spatial resolution and accuracy. To strengthen the ability of microwave remote sensing in global water cycle studies and seek for new opportunities of satellite missions, we put forward research contents as follows in the first year of project implementation:

(1) Refinement of the SMOS Multiangular Brightness Temperature with the Adoption of SMAP Observations

The Soil Moisture Ocean Salinity (SMOS) was the first mission to provides L-band multiangular brightness temperature (TB) at a global scale. However, radio frequency interference (RFI) and aliasing issues jeopardize part of its scientific applications in certain areas of the world. The Soil Moisture Active Passive (SMAP) mission provides the L-band brightness temperature at a fixed incidence angle of 40° with the RFI effects being well detected and filtered. In this study, we proposed a method called three-step regression to refine the SMOS multiangular TB with the adoption of SMAP observations through two options of anchor regression and translation transform, resulting in a multiangular TB dataset which is highly consistent with the SMAP TB. Results show the three-step regression can represent the multiangular L-band TB even for areas strongly affected by RFI, and improve the spatial and temporal coverage of SMOS. The evaluation results with 12 soil moisture networks show that the R2 between the refined TB dataset and in situ soil moisture has a significant improvement at strong RFI contaminate regions as compared with that of Centre Aval de Traitement des Données (CATDS) L3 daily TB product. The cumulative density function (CDF) of the refined TB at 40° are consistent with that of SMAP at a global or regional scale, which would promote the development of a consistent SMOS-SMAP TB and soil moisture products. (Submitted to IEEE TGRS)

(2) Retrievals of soil moisture and vegetation optical depth using a multi-channel collaborative algorithm

We explore multi-angular and multi-frequency approaches for the retrieval of soil moisture and vegetation tau, considering the payload configurations of current and future satellite missions (such as the Copernicus Imaging Microwave Radiometer, the Water Cycle Observation Mission, and the Terrestrial Water Resources Satellite) using a new set of ground observations. Two ground-based microwave radiometry datasets collected in Inner Mongolia during the Soil Moisture Experiment in the Luan River from July to August 2017 (cropland) and August to September 2018 (grassland) are used for this study. The corn field, which covers an entire growth period, indicated that the degree of information increases linearly as the number of channels (in terms of the incidence angle and frequency) increases, and that the multi-frequency observations contain slightly more independent information than do the multi-angular observations under the same number of channels. A multi-channel collaborative algorithm (MCCA) is developed based on the two-component version of the omega-tau model, which utilizes information from collaborative channels expressed as an analytical form of brightness temperature at the core channel to rule out the parameters to be retrieved. Results of soil moisture retrieval show that the multi-angular approach used by the MCCA generally has a better performance, unbiased root mean square difference (ubRMSD) varying from 0.028 cm3/cm3 to 0.037 cm3/cm3, than the multi-frequency approach (ubRMSD from 0.028 cm3/cm3 to 0.089 cm3/cm3) for the corn field. This is attributed to the dependence of vegetation tau on the frequency being more significant than that on the incidence angle. It is affirmed that increasing the number of observation channels could make the soil moisture retrieval more robust, but might also limit the retrieval performance, as the probability that the model estimations will not match the observations is increased. This study provides new insights into the design of potential satellite missions to improve soil moisture retrieval. A satellite with simultaneous multi-angular and multi-frequency observation capabilities is highly recommended. (Published in Remote Sensing of Environment)

(3) A long term global daily soil moisture dataset derived from AMSR-E and AMSR2 (2002-2019)

Long term surface soil moisture (SSM) data with stable and consistent quality are critical for global environment and climate change monitoring. L band radiometers onboard the recently lunched Soil Moisture Active Passive (SMAP) Mission can provide the state-of-the-art accuracy SSM, while Advanced Microwave Scanning Radiometer for EOS (AMSR-E) and AMSR2 series provide long term observational records of multi-frequency radiometers (C, X, and K bands). This study transfers the merits of SMAP to AMSR-E/2, and develops a global daily SSM dataset (named as NNsm) with stable and consistent quality at a 36 km resolution (2002-2019). The NNsm can reproduce the SMAP SSM accurately, with a global Root Mean Square Error (RMSE) of 0.029 m3/m3. NNsm also compares well with in situ SSM observations, and outperforms AMSR-E/2 standard SSM products from JAXA and LPRM. This global observation-driven dataset spans nearly two decades at present, and is extendable though the ongoing AMSR2 and upcoming AMSR3 missions for long-term studies of climate extremes, trends, and decadal variability. (Published in Scientific Data)

Soil moisture plays an important role in land surface processes by controlling the partitioning of global radiation into latent and sensible heat fluxes and the partitioning of precipitation into surface runoff and infiltration. Acquiring accurate soil moisture information over large areas remains a challenge. Assimilation of remotely sensed soil moisture into land surface models has been proven an effective way to generate more accurate soil moisture data products, but there are still several limitations. For example, it is observed that evapotranspiration estimates are hardly improved by soil moisture assimilation. Land surface models have in general an over-simplified representation of groundwater dynamics. In this work, we investigate the assimilation of soil moisture information from the SMAP satellite into the coupled land surface-subsurface model CLM-ParFlow, components of the Terrestrial Systems Modeling Platform (TSMP). CLM-ParFlow solves the 3D Richards´ equation for water flow in the subsurface, as well as overland flow by routing. We investigated whether this mechanistic representation of subsurface flow processes in combination with data assimilation results in a better characterization of soil moisture and evapotranspiraton than with the stand-alone land surface model CLM. A study was carried out over parts of western Germany, for the years 2017 and 2018. The SMAP soil moisture data is assimilated with the Ensemble Kalman Filter (EnKF), in some simulation scenarios including hydraulic parameter estimation. The simulated soil moisture and evapotranspiration (ET) time series are evaluated with in-situ measurements from Cosmic Ray Neutron Sensors (CRNS) and Eddy Covariance (EC) stations. Simulations illustrate that there is no systematic bias between soil moisture from SMAP and CLM-ParFlow. The soil moisture characterization improves with data assimilation and CLM-ParFLow captures better spatial patterns than CLM stand-alone.

The objective of this project is to assess the feasibility of remotely sensed products of key cryospheric and hydrological elements (snow, evapotranspiration, soil moisture and precipitation) in representative regions across the Third Pole region and the Heihe River Basin of China and selected test sites in other regions, e.g. northern Finland. The in-situ measurements used to validate remotely sensed products have been collected from several ground-based observation networks including the Finnish Meteorological Institute (FMI), the TERrestrial ENvironmental Observatories (TERENO), the Agrosphere institute (IBG-3) and The Qilian Mountain Observatories (QMO). Essential remote sensing products e.g. the GlobSnow data sets covering northern hemisphere and the soil moisture data set from SMOS, were evaluated by referencing ground-based observations in representative regions. The upscaling methods were developed to improve the representativeness of ground-based observations to remote sensing pixels. The validated products were also inter-compared with other gridded products, and the spatiotemporal trends were diagnosed by statistical indexes, e.g., RMSE and correlation coefficient. The performance of each product will be further evaluated in different landscapes, topographic conditions in the representative regions selected in China and Europe. The research results have been submitted to or published in international journals such as Remote Sensing and the Cryosphere. In addition, young scientists on this project made considerable efforts to observe snow, evapotranspiration, soil moisture and precipitation. They also assist with the validation of remotely sensed products on preprocessing data, developing validation algorithms and writing validation reports.

In the context of global climate change, drought, flood and wetland vegetation change caused by hydrometeorological change have caused great variations to the food, water, soil resources and ecological environment on which human beings depend for survival. The hydrometeorological characteristics in middle and lower Yangtze River Basin have significant spatiotemporal heterogeneity. A series of in-depth study of the hydrometeorological changes in middle and lower Yangtze River Basin and its impact on wetland vegetation is of great significance to natural, economic and ecological environment constructions, and satellite remote sensing data, meteorological observation data, hydrological observation data, and statistical yearbook data were collected for this research.

Soil moisture and reference crop evapotranspiration (ET0) are of great importance in assessing the potential impacts of climate changes on energy and water cycles, and they are key indicators of drought assessment. The history and future drought conditions were studied. The applicability of ESA CCI soil moisture data in Yangtze River Basin was verified, and a concept of lag time was proposed to quantify the hysteresis between soil moisture and meteorological elements, such as precipitation, temperature and evapotranspiration, under different climatic conditions and timescales. A novel Comprehensive Agricultural Drought Index (CADI) was then constructed to reflect the feedback of time lag effects in drought assessment. Results showed that the climate generally regulated the lag times, and the lag time in arid region is shorter than that in humid region. The CADI was able to effectively monitor the annual and seasonal variations and spatial pattern of agricultural drought, particularly better identify summer droughts, from which the crop phenology related agriculture drought monitoring can benefit. The spatiotemporal change of ET0 and the drought response over Poyang Lake watershed from 2011 to 2100 were investigated based on the meteorological data and the output of the general circulation model (GCM) from the CMIP5. We found that ET0 will increase in the future under the representative concentration pathway (RCP) 4.5 and RCP 8.5 scenarios, and the spatial distribution of ET0 is generally high in the east and low in the west. The drought index (DI) of the watershed showed an increasing trend, the seasonal distribution of DI is fall >summer >spring >winter, and the Ganjiang River Basin of Poyang lake will suffer high risks of future drought.

In 2020, Poyang Lake suffered the most serious flood hazard since the 21st century, which presents the characteristics of sharp shift from drought to flood. A multi-criteria model combining the analytic hierarchy process and Entropy weight method (AHP-Entropy) was proposed to assess the long and short flood risk. Validation of the flood risk assessment results shows that the flood risk assessment model has a great consistency with Sentinel-1 synthetic aperture radar data, which indicated that the presented flood risk model is reliable. Over all, the northeastern parts of the Poyang Lake basin are prone to floods and the risk of floods gradually decreased from the Poyang Lake area towards the surrounding areas. The sharp rise in water level and long-term, high-intensity precipitation are important causes of the flood. The water level in 2020 from July to October were at least 17% higher than the same period from 2010 to 2019 on average, and the average precipitation from June to September were all higher than the same period in previous years. The cropland areas were the most heavily inundated compared with wetland, grassland, impervious surface, forest and bare land.

Hydrology is a critical environmental condition for the evolution of wetland ecosystems. The hydrological influences on wetland cover distribution and transition in a large complex lake-floodplain system, Poyang Lake were then investigated. The statistical results of annual inundation conditions for different wetland cover types indicated that vegetation communities were preferential to hydrological environments with shorter annual inundation than water and mudflats, and different vegetation communities were distributed in areas with considerable variations in annual inundation, which suggested a substantial hydrological influence on the distribution of wetland cover in Poyang Lake. The spatial analysis indicated that hydrological changes were probably the dominant factor for the wetland cover evolution in the floodplain areas of the northern and central parts of Poyang Lake, but not the unique determined factor for wetland cover transitions in the shallow floodplains near the sink of the inflows in the eastern and southwestern parts of Poyang Lake.