The Potential Evaluation of Multisource Remote Sensing Data for Extracting Soil Moisture Based on the Method of BP Neural Network

As an important component of the earth ecosystem, soil moisture monitoring is of great significance in the fields of crop growth monitoring, crop yield estimation, variable irrigation, and other related applications. In order to mitigate or eliminate the impacts of sparse vegetation covers in farmland areas, this study combines multi-source remote sensing data from Sentinel-1 radar and Sentinel-2 optical satellites to quantitatively retrieve soil moisture content. Firstly, a traditional Oh model was applied to estimate soil moisture content after removing vegetation influence by a water cloud model. Secondly, support vector regression (SVR) and generalized regression neural network (GRNN) models were used to establish the relationships between various remote sensing features and real soil moisture. Finally, a regression convolutional neural network (CNNR) model is constructed to extract deep-level features of remote sensing data to increase soil moisture retrieval accuracy. In addition, polarimetric decomposition features for real Sentinel-1 PolSAR data are also included in the construction of inversion models. Based on the established soil moisture retrieval models, this study analyzes the influence of each input feature on the inversion accuracy in detail. The experimental results show that the optimal combination of R2 and root mean square error (RMSE) for SVR is 0.7619 and 0.0257 cm3/cm3, respectively. The optimal combination of R2 and RMSE for GRNN is 0.7098 and 0.0264 cm3/cm3, respectively. Especially, the CNNR model with optimal feature combination can generate inversion results with the highest accuracy, whose R2 and RMSE reach up to 0.8947 and 0.0208 cm3/cm3, respectively. Compared to other methods, the proposed algorithm improves the accuracy of soil moisture retrieval from synthetic aperture radar (SAR) and optical data. Furthermore, after adding polarization decomposition features, the R2 of CNNR is raised by 0.1524 and the RMSE of CNNR decreased by 0.0019 cm3/cm3 on average, which means that the addition of polarimetric decomposition features effectively improves the accuracy of soil moisture retrieval results.

Download Full-text

A NEW BP NEURAL NETWORK FUSION ALGORITHM FOR MULTI-SOURCE REMOTE SENSING DATA ON GROUNDWATER

Applied Ecology and Environmental Research ◽

10.15666/aeer/1704_90839095 ◽

2019 ◽

Vol 17 (4) ◽

Cited By ~ 1

Author(s):

F ZHANG

Keyword(s):

Neural Network ◽

Remote Sensing ◽

Bp Neural Network ◽

Remote Sensing Data ◽

Fusion Algorithm ◽

Sensing Data

Download Full-text

Sensitive Feature Evaluation for Soil Moisture Retrieval Based on Multi-Source Remote Sensing Data with Few In-Situ Measurements: A Case Study of the Continental U.S.

Water ◽

10.3390/w13152003 ◽

2021 ◽

Vol 13 (15) ◽

pp. 2003

Author(s):

Ling Zhang ◽

Zixuan Zhang ◽

Zhaohui Xue ◽

Hao Li

Keyword(s):

Neural Network ◽

Remote Sensing ◽

Feature Selection ◽

Soil Moisture ◽

Remote Sensing Data ◽

In Situ Measurements ◽

Feature Subset ◽

Selection Methods ◽

Sensing Data

Soil moisture (SM) plays an important role for understanding Earth’s land and near-surface atmosphere interactions. Existing studies rarely considered using multi-source data and their sensitiveness to SM retrieval with few in-situ measurements. To solve this issue, we designed a SM retrieval method (Multi-MDA-RF) using random forest (RF) based on 29 features derived from passive microwave remote sensing data, optical remote sensing data, land surface models (LSMs), and other auxiliary data. To evaluate the importance of different features to SM retrieval, we first compared 10 filter or embedded type feature selection methods with sequential forward selection (SFS). Then, RF was employed to establish a nonlinear relationship between the in-situ SM measurements from sparse network stations and the optimal feature subset. The experiments were conducted in the continental U.S. (CONUS) using in-situ measurements during August 2015, with only 5225 training samples covering the selected feature subset. The experimental results show that mean decrease accuracy (MDA) is better than other feature selection methods, and Multi-MDA-RF outperforms the back-propagation neural network (BPNN) and generalized regression neural network (GRNN), with the R and unbiased root-mean-square error (ubRMSE) values being 0.93 and 0.032 cm3/cm3, respectively. In comparison with other SM products, Multi-MDA-RF is more accurate and can well capture the SM spatial dynamics.

Download Full-text