Comment on essd-2021-362

doi:10.5194/essd-2021-362-cc1

Abstract

Abstract. Land surface soil moisture (SM) plays a critical role in hydrological processes and terrestrial ecosystems in desertification areas. Passive microwave remote-sensing products such as the Soil Moisture Active Passive (SMAP) satellite have been shown to monitor surface soil water well. However, the coarse spatial resolution and lack of full coverage of these products greatly limit their application in areas undergoing desertification. In order to overcome these limitations, a combination of multiple machine learning methods, including multiple linear regression (MLR), support vector regression (SVR), artificial neural networks (ANNs), random forest (RF) and extreme gradient boosting (XGB), have been applied to downscale the 36âkm SMAP SM products and produce higher-spatial-resolution SM data based on related surface variables, such as vegetation index and surface temperature. Desertification areas in northern China, which are sensitive to SM, were selected as the study area, and the downscaled SM with a resolution of 1âkm on a daily scale from 2015 to 2020 was produced. The results showed a good performance compared with in situ observed SM data, with an average unbiased root mean square error value of 0.057âm3âmâ3. In addition, their time series were consistent with precipitation and performed better than common gridded SM products. The data can be used to assess soil drought and provide a reference for reversing desertification in the study area. This dataset is freely available at <a href="https://doi.org/10.6084/m9.figshare.16430478.v6">https://doi.org/10.6084/m9.figshare.16430478.v6</a> (Rao et al., 2022).

Highlights

Surface soil moisture (SM) plays a very important role in water-energy cycle processes (Sandholt et al, 2002; De Santis et al, 2021) and is an important source of water for plants and soil microbes (Wang et al, 2007; Gu et al, 2008; Mallick et al., 2009)
SM data were mainly obtained through ground measurements or the assimilation of products based on land surface models such as the Global Land Data Assimilation System (GLDAS)
We studied five methods: Multiple linear regression (MLR), support vector regression (SVR), artificial neural networks (ANN), random forest (RF) and extreme gradient boosting (XGB)

Summary

Introduction

Surface soil moisture (SM) plays a very important role in water-energy cycle processes (Sandholt et al, 2002; De Santis et al, 2021) and is an important source of water for plants and soil microbes (Wang et al, 2007; Gu et al, 2008; Mallick et al., 2009). SM data were mainly obtained through ground measurements or the assimilation of products based on land surface models such as the Global Land Data Assimilation System (GLDAS). Data soil depths can be obtained, field measurements and in situ observations are limited due to the high cost and labor intensity involved in their collection and are generally not representative of soil water status over larger areas (Rahimzadeh-Bajgiran et al, 2013; Zhao et al, 2018; Bai et al, 2019). Data assimilation products largely depend on the accuracy of the land surface model and the original data (Zawadzki and Kędzior, 2016). They generally have low accuracy in areas where ground measurements are scarce, which is a problem that can be overcome with remote sensing

Objectives

Results

Discussion

Conclusion