Can We Use Satellite-Based Soil-Moisture Products at High Resolution to Investigate Land-Use Differences and Land–Atmosphere Interactions? A Case Study in the Savanna

Remote Sensing

Nitu Ojha¹  Aurore Brut¹  Olivier Merlin¹  Thierry Pellarin²  Carlos Román-Cascón³  Carlos Yagüe⁴  RamónC. Soriguer⁵  MaríaP. González-Dugo⁶  Ana Andreu⁶  Fabienne Lohou⁷  Marie Lothon⁷  Oscar Hartogensis⁸  Ricardo Díaz-Delgado⁹  David Aragonés⁹ 

[1] CESBIO, Université de Toulouse, CNES/CNRS/INRAE/IRD/UPS, 31400 Toulouse, France;CNRS, IRD, Univ. Grenoble Alpes, Grenoble INP, IGE, F-38000 Grenoble, France;Centre National d’Études Spatiales (CNES), 31400 Toulouse, France;Departamento de Física de la Tierra y Astrofísica. Universidad Complutense de Madrid, 28040 Madrid, Spain;Estación Biológica de Doñana, CSIC, 41092 Sevilla, Spain;IFAPA. Avd. Menéndez Pidal s/n, 14071 Córdoba, Spain;Laboratorie d’Aerologie, CNRS, Université de Toulouse, 31400 Toulouse, France;Meteorology and Air Quality Section, Wageningen University, 6700AA Wageningen, The Netherlands;Remote Sensing & GIS Lab. Estación Biológica de Doñana-CSIC, 41092 Sevilla, Spain;

关键词: soil moisture;    satellite data;    land use;    heterogeneity;    savanna;    DISPATCH;   

DOI  :  10.3390/rs12111701

来源: DOAJ

【 摘 要 】

The use of soil moisture (SM) measurements from satellites has grown in recent years, fostering the development of new products at high resolution. This opens the possibility of using them for certain applications that were normally carried out using in situ data. We investigated this hypothesis through two main analyses using two high-resolution satellite-based soil moisture (SBSM) products that combined microwave with thermal and optical data: (1) The Disaggregation based on Physical And Theoretical scale Change (DISPATCH) and, (2) The Soil Moisture Ocean Salinity-Barcelona Expert Center (SMOS-BEC Level 4). We used these products to analyse the SM differences among pixels with contrasting vegetation. This was done through the comparison of the SM measurements from satellites and the measurements simulated with a simple antecedent precipitation index (API) model, which did not account for the surface characteristics. Subsequently, the deviation of the SM from satellite with respect to the API model (bias) was analysed and compared for contrasting land use categories. We hypothesised that the differences in the biases of the varied categories could provide information regarding the water retention capacity associated with each type of vegetation. From the satellite measurements, we determined how the SM depended on the tree cover, i.e., the denser the tree cover, the higher the SM. However, in winter periods with light rain events, the tree canopy could dampen the moistening of the soil through interception and conducted higher SM in the open areas. This evolution of the SM differences that depended on the characteristics of each season was observed both from satellite and from in situ measurements taken beneath a tree and in grass on the savanna landscape. The agreement between both types of measurements highlighted the potential of the SBSM products to investigate the SM of each type of vegetation. We found that the results were clearer for DISPATCH, whose data was not smoothed spatially as it was in SMOS-BEC. We also tested whether the relationships between SM and evapotranspiration could be investigated using satellite data. The answer to this question was also positive but required removing the unrealistic high-frequency SM oscillations from the satellite data using a low pass filter. This improved the performance scores of the products and the agreement with the results from the in situ data. These results demonstrated the possibility of using SM data from satellites to substitute ground measurements for the study of land–atmosphere interactions, which encourages efforts to improve the quality and resolution of these measurements.

【 授权许可】

Unknown