A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products
The 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; howe...
| 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/9247250/ |
| _version_ | 1818643707308539904 |
|---|---|
| author | Yan Jin Yong Ge Yaojie Liu Yuehong Chen Haitao Zhang Gerard B. M. Heuvelink |
| author_facet | Yan Jin Yong Ge Yaojie Liu Yuehong Chen Haitao Zhang Gerard B. M. Heuvelink |
| author_sort | Yan Jin |
| collection | DOAJ |
| description | The 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 kinds of 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 eight days over a six-year period (2010-2015). Compared with five other downscaling methods, the downscaled predictions from the SVATARK method performs the best with in situ observations, resulting in a 24.4% reduction in root-mean-square error with 0.08 m<sup>3</sup>·m<sup>-3</sup> and a 8.2% increase in correlation coefficient with 0.72, 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. |
| first_indexed | 2024-12-17T00:03:14Z |
| format | Article |
| id | doaj.art-ae0848b1dea74bbab9b7bab2d3be1303 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-12-17T00:03:14Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-ae0848b1dea74bbab9b7bab2d3be13032022-12-21T22:11:02ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352021-01-01141025103710.1109/JSTARS.2020.30353869247250A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture ProductsYan Jin0https://orcid.org/0000-0002-6978-163XYong Ge1https://orcid.org/0000-0002-5175-5812Yaojie Liu2https://orcid.org/0000-0002-3112-9590Yuehong Chen3Haitao Zhang4Gerard B. M. Heuvelink5https://orcid.org/0000-0003-0959-9358School of Geographic and Biologic Information, Nanjing University of Posts and Telecommunications, Nanjing, ChinaState Key Laboratory of Resources and Environmental Information Systems, Institute of Geographic Sciences and Natural Resources Research, Beijing, ChinaInternational Institute for Earth System Science, Nanjing University, Nanjing, ChinaSchool of Earth Sciences and Engineering, Hohai University, Nanjing, ChinaSchool of Geographic and Biologic Information, Nanjing University of Posts and Telecommunications, Nanjing, ChinaSoil Geography and Landscape Group, Wageningen University, Wageningen, AA, The NetherlandsThe 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 kinds of 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 eight days over a six-year period (2010-2015). Compared with five other downscaling methods, the downscaled predictions from the SVATARK method performs the best with in situ observations, resulting in a 24.4% reduction in root-mean-square error with 0.08 m<sup>3</sup>·m<sup>-3</sup> and a 8.2% increase in correlation coefficient with 0.72, 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.https://ieeexplore.ieee.org/document/9247250/Area-to-area krigingdownscalingsoil moisturesupport vector regression |
| spellingShingle | Yan Jin Yong Ge Yaojie Liu Yuehong Chen Haitao Zhang Gerard B. M. Heuvelink A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Area-to-area kriging downscaling soil moisture support vector regression |
| title | A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products |
| title_full | A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products |
| title_fullStr | A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products |
| title_full_unstemmed | A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products |
| title_short | A Machine Learning-Based Geostatistical Downscaling Method for Coarse-Resolution Soil Moisture Products |
| title_sort | machine learning based geostatistical downscaling method for coarse resolution soil moisture products |
| topic | Area-to-area kriging downscaling soil moisture support vector regression |
| url | https://ieeexplore.ieee.org/document/9247250/ |
| work_keys_str_mv | AT yanjin amachinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yongge amachinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yaojieliu amachinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yuehongchen amachinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT haitaozhang amachinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT gerardbmheuvelink amachinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yanjin machinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yongge machinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yaojieliu machinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT yuehongchen machinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT haitaozhang machinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts AT gerardbmheuvelink machinelearningbasedgeostatisticaldownscalingmethodforcoarseresolutionsoilmoistureproducts |