Advancing peatland water level monitoring by combining Sentinel-1, Sentinel-2, and peat-specific SMAP Level-4 data

Peatlands are global hot spots of soil organic carbon, regionally important regulators of the water cycle, and provide several more critical ecosystem services. However, monitoring peatland hydrology remains challenging due to the complex surface properties and hydrodynamics in these areas. This study presents the development of a peatland water level product by integrating Sentinel-1 synthetic aperture radar, Sentinel-2 optical imagery, and the Soil Moisture Active Passive (SMAP) Level-4 (L4) product to advance the monitoring of peatland hydrology at high spatial resolution.

Our approach downscales the 9 km SMAP L4 product, which includes a specialized model parameterized for peatland processes, to 100 m using Sentinel-1 and Sentinel-2, addressing the spatial variability of peatland hydrology. SMAP L4 aids in resolving ambiguities in backscatter-to-water level relationships from Sentinel-1, distinguishing between subsurface and surface water level fluctuations. Additionally, the Normalized Difference Water Index (NDWI) and the optical trapezoid model (OPTRAM), derived from Sentinel-2, contribute to resolving ambiguities of the Sentinel-1 backscatter dynamics and to enhance the accuracy of water level estimates. NDWI assists in the identification of open water surfaces while OPTRAM mainly adds information on the interannual water level anomalies. Our product is provided with retrieval uncertainty estimates for each pixel.

We present the validation of our product across boreal, temperate, and tropical peatlands using time series of in situ water level data and surface water maps from high-resolution optical imagery. Our preliminary results highlight considerable variability in the quality of the new product over different peatlands and biomes. We discuss how quality differences relate to site characteristics and the retrieval uncertainty estimates.

Our approach targets a scalable and transferable method for monitoring peatland hydrology, addressing critical needs in management and conservation. Understanding hydrological state variables is essential due to their primary role in regulating ecosystem services. While SMAP L4-SM may not be directly useful for stakeholders at the management scale, the downscaled product holds significant potential for management applications. This method could become an operational tool for researchers and practitioners across diverse peatland research and application fields. This work is part of the ESA WorldPeatland project.