A
machine learning-based geostatistical downscaling method for
coarse-resolution soil moisture products
Yan Jin1, 2, Yong Ge3,*, Yaojie
Liu4, Yuehong Chen5, Haitao
Zhang1, 2 and Gerard B. M.
Heuvelink6
1 School of Geographic and Biologic Information,
Nanjing University of Posts and Telecommunications, Jiangsu Province,
Nanjing 210023, China; jinyan@njupt.edu.cn
2 Smart Health Big Data Analysis and Location Services
Engineering Lab of Jiangsu Province, Nanjing 210023, China
3 State Key Laboratory of Resources and Environmental
Information Systems, Institute of Geographic Sciences & Natural
Resources Research, Chinese Academy of Sciences, Beijing 100101, China
4 International Institute for Earth System Science,
Nanjing University, Jiangsu Province, Nanjing 210023, China
5 School of Earth Sciences and Engineering, Hohai
University, Nanjing 210098, China
6 Soil Geography and Landscape Group, Wageningen
University, P.O. Box 47, 6700 AAWageningen, The Netherlands
*Corresponding author: Prof. Yong Ge, State Key Laboratory of Resources
and Environmental Information Systems, Institute of Geographic Sciences
& Natural Resources Research, Chinese Academy of Sciences, Beijing
100101, China. Email: gey@lreis.ac.cn.
Abstract: The land Surface
Soil Moisture (SSM) products derived from microwave remote sensing have
a coarse spatial resolution, therefore downscaling is required to obtain
accurate SSM at high spatial resolution. An effective way to handle the
stratified heterogeneity is to model for various stratifications,
however the number of samples is often limited under each
stratification, influencing the downscaling accuracy. In this study, a
machine learning-based geostatistical model, which combines various
ancillary information at fine spatial scale, is developed for spatial
downscaling. The proposed support vector area-to-area regression kriging
(SVATARK) model incorporates support vector regression and area-to-area
kriging by considering the nonlinear relationships among variables for
various stratifications. SVATARK also considers the change of support
problem in the downscaling interpolation process as well as for solving
the small sample size in trend prediction. The SVATARK method is
evaluated in the Naqu region on the Tibetan Plateau, China to downscale
the European Space Agency’s (ESA) 25-km-resolution SSM product. The
1-km-resolution SSM predictions have been produced every 8 days over a
six-year period (2010-2015). Compared with other two methods, the
downscaled predictions from the SVATARK method performs the best with
in-situ observations, resulting in a 23.6 percent reduction in root mean
square error and a 10.7 percent increase in correlation coefficient, on
average. Additionally, anomalously low SSM values, an indicator of
drought, had a record low anomaly in mid-July for 2015, as noted by
previous studies, indicating that SVATARK could be utilized for drought
monitoring.
Key Words: Downscaling, Support vector regression, Area-to-area
kriging, Soil moisture