Reconstructing Monthly ECV Global Soil Moisture with an Improved Spatial Resolution
- 150 Downloads
Remote sensing techniques have provided global covered soil moisture at high temporal resolution, however, the coarse spatial resolution and the data gaps have greatly reduced their potential values in large numbers of practical and regional applications. This study proposed a two-steps reconstruction approach for reconstructing satellite-based soil moisture products (ECV) at an improved spatial resolution. The reconstruction model implemented the Random Forests (RF) regression algorithm to simulate the relationships between soil moisture and environmental variables, and takes advantages of the high spatial resolution of optical remote sensing products: the data gaps of ECV soil moisture products were firstly filled by the estimation model trained using available pixels of the ECV products and corresponding environmental variables; then a spatial downscaling was carried out to the gap-filled ECV products to obtain the reconstructed soil moisture with fine spatial resolution (0.05°). As a result, the reconstructed soil moisture well fill the data gaps of the original ECV products and nicely reproduced the original soil moisture values (R2 > 0.98). The spatial resolution and variation details of the soil moisture products were also improved significantly. Validation results indicated that the reconstructed soil moisture showed comparable good performance (average R2 = 0.66) as the original ECV products (average R2 = 0.65) and nicely reflect the temporal behavior of ground-based measurements. As a result, the reconstructed soil moisture well filled the data gaps and greatly improved the spatial resolution of ECV products.
KeywordsSoil moisture Reconstruction ECV Remote sensing Downscaling
This study were jointly supported by the GDAS’ Special Project of Science and Technology Development (2017GDASCX-0101, 2018GDASCX-0904), the Guangdong Innovative and Entrepreneurial Research Team Program (2016ZT06D336), the National Natural Science Foundation of China (41601175, 41401430), and the National Earth System Science Data Sharing Infrastructure (http://www.geodata.cn/).
- Dorigo W et al (2017) ESA CCI Soil Moisture for improved Earth system understanding: State-of-the art and future directions. Remote Sens Environ. https://doi.org/10.1016/j.rse.2017.07.001
- Fern N-DM, Cernadas E, Barro S, Amorim D (2014) Do we need hundreds of classifiers to solve real world classification problems? J Mach Learn Res 15:3133–3181Google Scholar
- Im J, Park S, Rhee J, Baik J, Choi M (2016) Downscaling of AMSR-E soil moisture with MODIS products using machine learning approaches. Environ Earth Sci:75–1120Google Scholar
- Koike T, Nakamura Y, Kaihotsu I, Davaa G, Matsuura N, Tamagawa K, Fujii H (2004) Development of an advanced microwave scanning radiometer (AMSR-E) Algorithm for soil moisture and vegetation water content. Doboku Gakkai Ronbunshuu B 48:217–222Google Scholar