Consistent CDRs of soil moisture and vegetation

Forestry, agriculture, and climate research need high quality information on the water that is stored in soils and plants to estimate accurately impacts of droughts on plant production and harvest yields. Observations from microwave satellites allow estimating the moisture content in soil and vegetation for the last 40 years. However, this information is not daily available and has a poor quality to draw clear conclusions about the short- to long-term changes in the storage and flows of water over land. In CONSOLIDATION, we will for the first time combine multiple satellite observations of soil and vegetation moisture with a new-generation model that simulates changes in water and vegetation over land. In this way, we will obtain worldwide estimates on soil and vegetation moisture and biomass since 1980 with high spatial (~10 km) and temporal (every 3 hours) detail. The resulting datasets will be of better quality than the current datasets that come either only from satellite observations or only from models, because we will combine the best information from both approaches. In addition, hourly to daily changes will be included and physical links between soil moisture and vegetation and all other land surface variables will be ensured. We will then use these datasets to forecasts crop yields. Furthermore, we will investigate how soil moisture and droughts affected plant biomass over large regions. The project will join the knowledge of two partners on combining observations with models (KU Leuven) and on estimating soil and vegetation moisture from microwave satellite observations (TU Wien).

Accurate estimates of the impact of droughts on plant production and harvest yields require high-quality information on the amount of water stored in soils and plants. Microwave satellite observations have allowed the moisture content of soil and vegetation to be estimated over the past 45 years. However, this information is not available daily and is subject to uncertainties that prevent clear conclusions being drawn about short- to long-term changes in water storage and flow over land. Conversely, large-scale land surface models provide consistent and contiguous simulations of numerous hydrological and ecosystem variables, but they inadequately describe the interactions between hydrology and vegetation functioning. In the CONSOLIDATION project, multiple satellite observations of soil moisture and vegetation productivity were combined with Noah-MP, a new-generation land surface model that simulates changes in water and vegetation over land. The resulting datasets are of better quality than current datasets derived solely from satellite observations or models because they combine the most reliable information from both sources. In addition, physical links between soil moisture and vegetation and all other land surface variables have been ensured. The project has focused on optimising methodologies for combining both data streams, as well as assessing the improved value of these model-data integration techniques for numerical weather prediction and machine learning-based crop yield forecasting. The project has brought together two partners with complementary expertise: KU Leuven in data assimilation and TU Wien in Earth observation of soil and vegetation moisture from microwave satellites.