Publications

Kang, J; Jin, R; Li, X; Ma, CF; Qin, J; Zhang, Y (2017). High spatio-temporal resolution mapping of soil moisture by integrating wireless sensor network observations and MODIS apparent thermal inertia in the Babao River Basin, China. REMOTE SENSING OF ENVIRONMENT, 191, 232-245.

Abstract
Soil moisture distributions with high spatio-temporal resolution are scarce but beneficial for understanding ecohydrological processes and closing the water cycle at the basin scale. Sensor networks are innovative in their ability to capture the spatio-temporal heterogeneity and dynamics of soil moisture; however, they cannot be used to directly derive spatially continuous soil moisture distributions. A Bayesian-based upscaling algorithm that utilizes MODIS-derived apparent thermal inertia is used to map daily soil moisture spatial patterns with a resolution of 1 km in the Babao River Basin, China. The 2-4 cm soil moisture observations from seven automatic meteorological stations located in different elevation zones from 3000 m to 3500 m are employed to validate the mapping algorithm. The correlation coefficient and unbiased root-mean-square error (RMSE) averaged 0.880 and 0.031 cm(3)/cm(3), respectively, which indicate satisfactory estimation accuracy. The 1 km resolution soil moisture products are re-sampled to a resolution of 25 km and then compared to the level 3 Soil Moisture and Ocean Salinity Mission (SMOS) soil moisture product. The results show that both products exhibit strong temporal consistency; however, due to complex topography, the SMOS soil moisture is generally lower than that of the upscaling results. Semivariograms and an empirical orthogonal function (EOF) analysis are used to analyze the space-time heterogeneities of soil moisture at the 1 km scale. In the summer, rainfall results in soil moisture with low spatial variability and a complex spatial structure. After the rainy season, the spatial heterogeneity of soil moisture is affected by other factors, such as soil texture, evapotranspiration and topography. From the perspective of temporal variation, the upscaled soil moisture shows a well-defined seasonal cycle, which represents the effects of decreased rainfall from August to October. Because more rain falls in the summer due to the mountain microclimate, the oscillation in soil moisture is more pronounced over 20% of the area compared to that in other regions. Based on a validation analysis of the mapping results, the upscaling method is proven feasible, and the upscaled soil moisture can be used to analyze eco-hydrological processes and validate remote sensing products. (C) 2017 Elsevier Inc. All rights reserved.

DOI:
10.1016/j.rse.2017.01.027

ISSN:
0034-4257