Downscaling SMAP Soil Moisture Products With Convolutional Neural Network
Soil moisture (SM) downscaling has been extensively investigated in recent years to improve coarse resolution of SM products. However, available methods for downscaling are generally based on pixel-to-pixel strategy, which ignores the information among pixels. Hence, a new downscaling method based o...
Main Authors: | , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2021-01-01
|
Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9390292/ |
id |
doaj-0f5cd82769964e618819347174a021e1 |
---|---|
record_format |
Article |
spelling |
doaj-0f5cd82769964e618819347174a021e12021-06-03T23:03:54ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352021-01-01144051406210.1109/JSTARS.2021.30697749390292Downscaling SMAP Soil Moisture Products With Convolutional Neural NetworkWei Xu0https://orcid.org/0000-0002-8341-2021Zhaoxu Zhang1https://orcid.org/0000-0003-1236-1463Zehao Long2Qiming Qin3School of Earth and Space Sciences, Peking University, Beijing, ChinaSchool of Earth and Space Sciences, Peking University, Beijing, ChinaSchool of Earth and Space Sciences, Peking University, Beijing, ChinaSchool of Earth and Space Sciences, Peking University, Beijing, ChinaSoil moisture (SM) downscaling has been extensively investigated in recent years to improve coarse resolution of SM products. However, available methods for downscaling are generally based on pixel-to-pixel strategy, which ignores the information among pixels. Hence, a new downscaling method based on a convolutional neural network (CNN) is proposed to solve the problem. Furthermore, a weight layer is designed for the input, and residual SM is treated as the output of the CNN to improve the accuracy. This method is applied to downscale Soil Moisture Active Passive (SMAP) SM products (i.e., 36-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P}$</tex-math></inline-formula> and 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula>) from January 1, 2018 to December 30, 2018. Compared with 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula>, the 9-km downscaling result is satisfactory with obtained correlation coefficient (<inline-formula><tex-math notation="LaTeX">$R$</tex-math></inline-formula>), root mean square error (RMSE), and unbiased RMSE (ubRMSE) values of 95.81%, 2.77%, and 2.67%, respectively. Moreover, SMAP SM products (36 and 9 km) and downscaling SM (3 and 1 km) are validated by the <italic>in situ</italic> data, which are collected by the 109 stations of the Oklahoma Mesonet SM monitoring network. Mean <inline-formula><tex-math notation="LaTeX">$R$</tex-math></inline-formula>, RMSE, and ubRMSE values are 67.92%, 7.94%, and 4.87% for 36-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P}$</tex-math></inline-formula>; 67.78%, 8.35%, and 4.95% for 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula>; 67.28%, 8.34%, and 4.97% for 3-km downscaling SM; 65.90%, 8.40%, and 5.18% for 1-km downscaling SM, respectively. The 3-km downscaling SM generated by this method can improve the coarse resolution of 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula> while preserving its accuracy. However, error will remarkably increase in the 1-km downscaling SM. Therefore, the proposed method provides a new strategy for SM downscaling and obtains satisfactory results in practice. Additional studies can be conducted in the future.https://ieeexplore.ieee.org/document/9390292/CNNdownscalingSMAPsoil moisture (SM) |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Wei Xu Zhaoxu Zhang Zehao Long Qiming Qin |
spellingShingle |
Wei Xu Zhaoxu Zhang Zehao Long Qiming Qin Downscaling SMAP Soil Moisture Products With Convolutional Neural Network IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing CNN downscaling SMAP soil moisture (SM) |
author_facet |
Wei Xu Zhaoxu Zhang Zehao Long Qiming Qin |
author_sort |
Wei Xu |
title |
Downscaling SMAP Soil Moisture Products With Convolutional Neural Network |
title_short |
Downscaling SMAP Soil Moisture Products With Convolutional Neural Network |
title_full |
Downscaling SMAP Soil Moisture Products With Convolutional Neural Network |
title_fullStr |
Downscaling SMAP Soil Moisture Products With Convolutional Neural Network |
title_full_unstemmed |
Downscaling SMAP Soil Moisture Products With Convolutional Neural Network |
title_sort |
downscaling smap soil moisture products with convolutional neural network |
publisher |
IEEE |
series |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
issn |
2151-1535 |
publishDate |
2021-01-01 |
description |
Soil moisture (SM) downscaling has been extensively investigated in recent years to improve coarse resolution of SM products. However, available methods for downscaling are generally based on pixel-to-pixel strategy, which ignores the information among pixels. Hence, a new downscaling method based on a convolutional neural network (CNN) is proposed to solve the problem. Furthermore, a weight layer is designed for the input, and residual SM is treated as the output of the CNN to improve the accuracy. This method is applied to downscale Soil Moisture Active Passive (SMAP) SM products (i.e., 36-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P}$</tex-math></inline-formula> and 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula>) from January 1, 2018 to December 30, 2018. Compared with 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula>, the 9-km downscaling result is satisfactory with obtained correlation coefficient (<inline-formula><tex-math notation="LaTeX">$R$</tex-math></inline-formula>), root mean square error (RMSE), and unbiased RMSE (ubRMSE) values of 95.81%, 2.77%, and 2.67%, respectively. Moreover, SMAP SM products (36 and 9 km) and downscaling SM (3 and 1 km) are validated by the <italic>in situ</italic> data, which are collected by the 109 stations of the Oklahoma Mesonet SM monitoring network. Mean <inline-formula><tex-math notation="LaTeX">$R$</tex-math></inline-formula>, RMSE, and ubRMSE values are 67.92%, 7.94%, and 4.87% for 36-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P}$</tex-math></inline-formula>; 67.78%, 8.35%, and 4.95% for 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula>; 67.28%, 8.34%, and 4.97% for 3-km downscaling SM; 65.90%, 8.40%, and 5.18% for 1-km downscaling SM, respectively. The 3-km downscaling SM generated by this method can improve the coarse resolution of 9-km <inline-formula><tex-math notation="LaTeX">$\mathbf {L3{\_}SM{\_}P{\_}E}$</tex-math></inline-formula> while preserving its accuracy. However, error will remarkably increase in the 1-km downscaling SM. Therefore, the proposed method provides a new strategy for SM downscaling and obtains satisfactory results in practice. Additional studies can be conducted in the future. |
topic |
CNN downscaling SMAP soil moisture (SM) |
url |
https://ieeexplore.ieee.org/document/9390292/ |
work_keys_str_mv |
AT weixu downscalingsmapsoilmoistureproductswithconvolutionalneuralnetwork AT zhaoxuzhang downscalingsmapsoilmoistureproductswithconvolutionalneuralnetwork AT zehaolong downscalingsmapsoilmoistureproductswithconvolutionalneuralnetwork AT qimingqin downscalingsmapsoilmoistureproductswithconvolutionalneuralnetwork |
_version_ |
1721398639630221312 |