Reply on RC3

Abstract

High-resolution (HR) surface soil moisture (SM) observations are important for applications in hydrology and agriculture, among other purposes. For instance, the S2MP (Sentinel-1/Sentinel-2 derived Soil Moisture Product) algorithm was designed to retrieve surface SM at agricultural plot scale using simultaneously Sentinel-1 (S1) backscatter coefficients and Sentinel-2 (S2) NDVI (Normalized Difference Vegetation Index) as inputs to a neural network trained with Water Cloud Model simulations. However, for many applications, including future climate impact assessment at regional level, a resolution of 1 km is already a significant improvement with respect to most of the publicly available SM data sets, which have resolutions of about 25 km. Therefore, in this study, the S2MP algorithm was adapted to work at a resolution of 1 km and extended from croplands (cereals and grasslands) to herbaceous vegetation types. A target resolution of 1 km also allows to explore the use of NDVI derived from Sentinel-3 (S3) instead of S2. The algorithm improvements are evaluated both over Europe and other regions of the globe, for which S1 coverage is poorer. Two sets of SM maps at 1-km resolution were produced with S2MP over six regions of about 104 km2 in the southwest and southeast of France, Spain, Tunisia, North America, as well as Australia from 2017 to 2019. The first set of maps was derived from the combination of S1 and S2 data (S1+S2 maps), while the second one was derived from the combination of S1 and S3 (S1+S3 maps). S1+S2 and S1+S3 SM maps were compared to each other and to those of the 1-km resolution Copernicus Global Land Service (CGLS) SM and Soil Water Index (SWI) data sets as well as to the SMAP+S1 product. The S2MP S1+S2 and S1+S3 SM maps are in very good agreement in terms of correlation (R ≥ 0.9), bias (≤ 0.04 m3 m−3) and standard deviation of the difference (STDD ≤ 0.03 m3 m−3) over the 6 domains investigated in this study. The S2MP maps are well correlated to those from the CGLS SM product (R ∼ 0.7–0.8), but the correlations with respect to the other HR maps (CGLS SWI and SMAP+S1) drop significantly over many areas of the 6 domains investigated in this study. In addition, higher correlations between the HR maps were found over croplands and when the 1-km pixels have a very homogeneous land cover. The bias in between the different maps was found to be significant over some areas of the six domains, reaching values of ± 0.1 m3 m−3. The S1+S2 maps show a lower STDD with respect to CGLS maps (≤ 0.06 m3 m−3) than with respect to the SMAP+S1 maps (≤ 0.1 m3 m−3) for all the 6 domains. Finally, all the HR data sets were also compared to in situ measurements from 5 networks across 5 countries along with coarse resolution (CR) SM products from SMAP, SMOS and the ESA Climate Change Initiative (CCI). While all the CR and HR products show different bias and STDD, the HR products show lower correlations than the CR ones with respect to in situ measurements.