The "Soil MOisture retrieval from multi-temporal SAR data" (SMOSAR) software implements a Soil Moisture (SM) retrieval algorithm from Sentinel-1 SAR data based on a Short Term Change Detection (STCD) approach. The rationale of the algorithm is to exploit the frequent revisit of Sentinel-1 in order to decouple the soil moisture contribution from that of soil roughness and vegetation. The software has been developed and validated in the ESA SEOM LAND project: "Exploitation of Sentinel-1 for Surface Soil Moisture Retrieval at High Resolution (Exploit-S-1), to produce SM maps at 1 km resolution. Successively, the software has been improved in the H2020 EU project: "Sentinels Synergy for Agriculture" (SENSAGRI), to integrate spectral information from Sentinel-2 images, to support the masking of abrupt changes of surface parameters (roughness or vegetation) at field scale, which are not related to SSM changes, thus improving the SMOSAR performance at high resolution. If NDVI is not available due to, e.g., cloud cover, then the ratio of VH/VV is considered as a proxy of NDVI. The output consists of SM maps characterized by a pixel size of 50 m and by a hybrid spatial resolution, ranging between 0.1 km to 1.0 km. More precisely, the initial SM retrieval is performed at 40 m pixel size (i.e., ~100 m resolution), then the maps are averaged at field scale, over those areas where the parcel borders (e.g., from the LPIS EU system) are available, and at 1 km resolution over the remaining areas. Together with the map of the SM spatial mean at hybrid resolution, the SENSAGRI output includes a companion map of SM spatial standard deviation. The SMOSAR prototype code consists of IDL executable files and other additional bash scripts developed and tested under the Linux operating system (kernel 4.9.0-9-amd64), running IDL in 64-bit mode, and a platform using 2xCPU 14 core Intel Xeon E5-2660v4 2.0Ghz, 35M cache, 128GB RAM DIMM 2400MT/s RDIMM. The routines in IDL have been developed by using the IDL version 8.6.0 [W1]. The IDL executable files (*.sav files) have been obtained by using the IDL SAVE procedure, which saves all IDL routines by using the XDR (eXternal Data Representation) format.
SMOSAR (Surface Soil moisture retrieval from Sentinel data) software
G Satalino;A Balenzano;F Lovergine;F Mattia
2020
Abstract
The "Soil MOisture retrieval from multi-temporal SAR data" (SMOSAR) software implements a Soil Moisture (SM) retrieval algorithm from Sentinel-1 SAR data based on a Short Term Change Detection (STCD) approach. The rationale of the algorithm is to exploit the frequent revisit of Sentinel-1 in order to decouple the soil moisture contribution from that of soil roughness and vegetation. The software has been developed and validated in the ESA SEOM LAND project: "Exploitation of Sentinel-1 for Surface Soil Moisture Retrieval at High Resolution (Exploit-S-1), to produce SM maps at 1 km resolution. Successively, the software has been improved in the H2020 EU project: "Sentinels Synergy for Agriculture" (SENSAGRI), to integrate spectral information from Sentinel-2 images, to support the masking of abrupt changes of surface parameters (roughness or vegetation) at field scale, which are not related to SSM changes, thus improving the SMOSAR performance at high resolution. If NDVI is not available due to, e.g., cloud cover, then the ratio of VH/VV is considered as a proxy of NDVI. The output consists of SM maps characterized by a pixel size of 50 m and by a hybrid spatial resolution, ranging between 0.1 km to 1.0 km. More precisely, the initial SM retrieval is performed at 40 m pixel size (i.e., ~100 m resolution), then the maps are averaged at field scale, over those areas where the parcel borders (e.g., from the LPIS EU system) are available, and at 1 km resolution over the remaining areas. Together with the map of the SM spatial mean at hybrid resolution, the SENSAGRI output includes a companion map of SM spatial standard deviation. The SMOSAR prototype code consists of IDL executable files and other additional bash scripts developed and tested under the Linux operating system (kernel 4.9.0-9-amd64), running IDL in 64-bit mode, and a platform using 2xCPU 14 core Intel Xeon E5-2660v4 2.0Ghz, 35M cache, 128GB RAM DIMM 2400MT/s RDIMM. The routines in IDL have been developed by using the IDL version 8.6.0 [W1]. The IDL executable files (*.sav files) have been obtained by using the IDL SAVE procedure, which saves all IDL routines by using the XDR (eXternal Data Representation) format.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
