A MULTI-DECADAL HIGH-RESOLUTION SOIL WATER CONTENT PRODUCT BASED ON THE SMAP CLIMATOLOGY

T. Frederikse¹, A. Geers¹, A. de Nijs¹, A. Mulder¹

¹Planet Labs

Over the last decade, the SMAP mission has provided global high-quality soil water content (SWC) estimates. However, for many applications, especially in the domain of climate science and anomaly detection, long records are fundamental. Previous missions, such as AMSR-E, provide an opportunity to extend the record back in time and to create a 25-year SWC record. Furthermore, the current AMSR-2 and recently-launched AMSR-3 microwave radiometers provide additional SWC estimates and ensure continuity for the coming decade.

Here, we present a novel approach to combine SWC estimates from multiple missions to create a long unified high-resolution record. First, we utilize the overlapping microwave footprints to extract kilometer-scale brightness temperature estimates. We use the Land Parameter Retrieval Model (LPRM) to compute SWC from these brightness temperatures. Finally, we harmonize and combine the multiple estimates from each radiometer and band using a robust statistical framework. This framework uses the decade-long SMAP record as its baseline. This procedure results in a near-daily soil water content estimate that is consistent in space and time.

With this method, we can combine AMSR2 C-band and X-band signals to almost fully replicate SMAP-derived SWC, except for regions with very dense vegetation. The method is able to combine the outputs from upcoming multi-frequency microwave radiometer missions, such as AMSR3 and CIMR, into a single robust SWC estimate.

The figure below shows the unified SWC time series over the Netherlands versus SWC derived from SMAP-only observations. The green estimate uses all available radiometer missions, except SMAP, and the yellow line is only based on AMSR2 data.