{"id":{"repo_id":"sdstate","oai_identifier":"oai:openprairie.sdstate.edu:etd2-2875"},"canonical_url":"https://search.ndltd.org/etd/sdstate/oai:openprairie.sdstate.edu:etd2-2875","repository":{"repo_id":"sdstate","name":"South Dakota State University","base_url":"https://openprairie.sdstate.edu/do/oai/"},"display":{"title":"Multi-Source Remote Sensing–Based Soil Moisture Prediction Using Machine Learning","abstract":"Soil moisture (SM) plays a central role in climatic and environmental processes, influencing shear strength of soil, agricultural productivity, land–atmosphere interactions, and hydrologic functioning. However, accurately estimating SM across diverse climatic regions remains challenging due to spatial heterogeneity, limited in situ measurements, and inconsistencies in sensor resolution. Machine learning (ML) and remote sensing offer promising avenues for improving SM prediction, yet many existing approaches struggle with generalization across climatic gradients and often fail to capture temporal variability. This study integrates multi-source satellite and climate datasets, including SMAP L4_SM, MODIS land surface temperature, Daymet meteorological variables, and in situ observations from the U.S. Climate Reference Network (USCRN), to evaluate the potential of ML models for daily SM prediction across nine U.S. climatic zones. All datasets were harmonized to a common 9 km spatial and daily temporal resolution using Google Earth Engine and Python preprocessing workflows. Three ML algorithms, such as, Extreme Gradient Boosting (XGBoost), Support Vector Regression (SVR), and Multilayer Perceptron (MLP), were trained and evaluated under zone-specific experimental cases. Across the nine climatic zones, XGBoost consistently provided the strongest predictive performance, achieving an R² of 0.43 in Zone 6A, particularly in arid and mixed-dry environments. MLP demonstrated comparable performance in colder regions, while SVR showed greater stability in humid zones. These findings highlight the sensitivity of model performance to climatic regimes and emphasize the importance of region-specific modeling strategies. The results also suggest that future hybrid frameworks incorporating additional variables, such as vegetation indices, topographic features, and soil properties, could further enhance SM prediction accuracy and transferability. Overall, this study provides a scalable, data-driven approach for regional soil moisture estimation and offers practical insights for applications in precision agriculture, drought monitoring, hydrological modeling, and climate resilience planning. The work also underscores the need for continued research in temporal modeling, multi-sensor data fusion, and spatial validation techniques to strengthen the robustness of SM prediction across heterogeneous landscapes.","abstract_html":"Soil moisture (SM) plays a central role in climatic and environmental processes, influencing shear strength of soil, agricultural productivity, land–atmosphere interactions, and hydrologic functioning. However, accurately estimating SM across diverse climatic regions remains challenging due to spatial heterogeneity, limited in situ measurements, and inconsistencies in sensor resolution. Machine learning (ML) and remote sensing offer promising avenues for improving SM prediction, yet many existing approaches struggle with generalization across climatic gradients and often fail to capture temporal variability. This study integrates multi-source satellite and climate datasets, including SMAP L4_SM, MODIS land surface temperature, Daymet meteorological variables, and in situ observations from the U.S. Climate Reference Network (USCRN), to evaluate the potential of ML models for daily SM prediction across nine U.S. climatic zones. All datasets were harmonized to a common 9 km spatial and daily temporal resolution using Google Earth Engine and Python preprocessing workflows. Three ML algorithms, such as, Extreme Gradient Boosting (XGBoost), Support Vector Regression (SVR), and Multilayer Perceptron (MLP), were trained and evaluated under zone-specific experimental cases. Across the nine climatic zones, XGBoost consistently provided the strongest predictive performance, achieving an R² of 0.43 in Zone 6A, particularly in arid and mixed-dry environments. MLP demonstrated comparable performance in colder regions, while SVR showed greater stability in humid zones. These findings highlight the sensitivity of model performance to climatic regimes and emphasize the importance of region-specific modeling strategies. The results also suggest that future hybrid frameworks incorporating additional variables, such as vegetation indices, topographic features, and soil properties, could further enhance SM prediction accuracy and transferability. Overall, this study provides a scalable, data-driven approach for regional soil moisture estimation and offers practical insights for applications in precision agriculture, drought monitoring, hydrological modeling, and climate resilience planning. The work also underscores the need for continued research in temporal modeling, multi-sensor data fusion, and spatial validation techniques to strengthen the robustness of SM prediction across heterogeneous landscapes.","abstract_has_math":false,"creators":["Neupane, Niraj"],"institution":null,"degree_name":"Master of Science (MS)","degree_level":"Thesis - Open Access","degree_discipline":"Civil and Environmental Engineering","degree_department":null,"school":null,"contributors":["Aritra Banerjee"],"advisors":[],"committee_chairs":[],"committee_members":[],"year":null,"date_issued":"","date_published":null,"updated_at":"2026-11-19T09:05:56Z","subjects":["Civil and Environmental Engineering","Remote Sensing","Soil Science"],"languages":[],"rights":[],"rights_urls":["http://rightsstatements.org/vocab/InC/1.0/"],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://openprairie.sdstate.edu/etd2/1862","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://openprairie.sdstate.edu/etd2/1862","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://openprairie.sdstate.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Aopenprairie.sdstate.edu%3Aetd2-2875","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Aritra Banerjee"]},{"key":"dc:creator","label":"Author","values":["Neupane, Niraj"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.available","label":"Dc Date Available","values":["2026-01-23T08:00:00Z"]},{"key":"dc:date.created","label":"Dc Date Created","values":["2025-01-01T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Civil and Environmental Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Thesis - Open Access"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master of Science (MS)"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Civil and Environmental Engineering","Remote Sensing","Soil Science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["http://rightsstatements.org/vocab/InC/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://openprairie.sdstate.edu/etd2/1862"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Soil moisture (SM) plays a central role in climatic and environmental processes, influencing shear strength of soil, agricultural productivity, land–atmosphere interactions, and hydrologic functioning. However, accurately estimating SM across diverse climatic regions remains challenging due to spatial heterogeneity, limited in situ measurements, and inconsistencies in sensor resolution. Machine learning (ML) and remote sensing offer promising avenues for improving SM prediction, yet many existing approaches struggle with generalization across climatic gradients and often fail to capture temporal variability. This study integrates multi-source satellite and climate datasets, including SMAP L4_SM, MODIS land surface temperature, Daymet meteorological variables, and in situ observations from the U.S. Climate Reference Network (USCRN), to evaluate the potential of ML models for daily SM prediction across nine U.S. climatic zones. All datasets were harmonized to a common 9 km spatial and daily temporal resolution using Google Earth Engine and Python preprocessing workflows. Three ML algorithms, such as, Extreme Gradient Boosting (XGBoost), Support Vector Regression (SVR), and Multilayer Perceptron (MLP), were trained and evaluated under zone-specific experimental cases. Across the nine climatic zones, XGBoost consistently provided the strongest predictive performance, achieving an R² of 0.43 in Zone 6A, particularly in arid and mixed-dry environments. MLP demonstrated comparable performance in colder regions, while SVR showed greater stability in humid zones. These findings highlight the sensitivity of model performance to climatic regimes and emphasize the importance of region-specific modeling strategies. The results also suggest that future hybrid frameworks incorporating additional variables, such as vegetation indices, topographic features, and soil properties, could further enhance SM prediction accuracy and transferability. Overall, this study provides a scalable, data-driven approach for regional soil moisture estimation and offers practical insights for applications in precision agriculture, drought monitoring, hydrological modeling, and climate resilience planning. The work also underscores the need for continued research in temporal modeling, multi-sensor data fusion, and spatial validation techniques to strengthen the robustness of SM prediction across heterogeneous landscapes."]},{"key":"dc:title","label":"Title","values":["Multi-Source Remote Sensing–Based Soil Moisture Prediction Using Machine Learning"]}]}],"canonical_facts":{"dc:contributor":["Aritra Banerjee"],"dc:creator":["Neupane, Niraj"],"dc:date.available":["2026-01-23T08:00:00Z"],"dc:date.created":["2025-01-01T08:00:00Z"],"dc:description.abstract":["Soil moisture (SM) plays a central role in climatic and environmental processes, influencing shear strength of soil, agricultural productivity, land–atmosphere interactions, and hydrologic functioning. However, accurately estimating SM across diverse climatic regions remains challenging due to spatial heterogeneity, limited in situ measurements, and inconsistencies in sensor resolution. Machine learning (ML) and remote sensing offer promising avenues for improving SM prediction, yet many existing approaches struggle with generalization across climatic gradients and often fail to capture temporal variability. This study integrates multi-source satellite and climate datasets, including SMAP L4_SM, MODIS land surface temperature, Daymet meteorological variables, and in situ observations from the U.S. Climate Reference Network (USCRN), to evaluate the potential of ML models for daily SM prediction across nine U.S. climatic zones. All datasets were harmonized to a common 9 km spatial and daily temporal resolution using Google Earth Engine and Python preprocessing workflows. Three ML algorithms, such as, Extreme Gradient Boosting (XGBoost), Support Vector Regression (SVR), and Multilayer Perceptron (MLP), were trained and evaluated under zone-specific experimental cases. Across the nine climatic zones, XGBoost consistently provided the strongest predictive performance, achieving an R² of 0.43 in Zone 6A, particularly in arid and mixed-dry environments. MLP demonstrated comparable performance in colder regions, while SVR showed greater stability in humid zones. These findings highlight the sensitivity of model performance to climatic regimes and emphasize the importance of region-specific modeling strategies. The results also suggest that future hybrid frameworks incorporating additional variables, such as vegetation indices, topographic features, and soil properties, could further enhance SM prediction accuracy and transferability. Overall, this study provides a scalable, data-driven approach for regional soil moisture estimation and offers practical insights for applications in precision agriculture, drought monitoring, hydrological modeling, and climate resilience planning. The work also underscores the need for continued research in temporal modeling, multi-sensor data fusion, and spatial validation techniques to strengthen the robustness of SM prediction across heterogeneous landscapes."],"dc:identifier":["https://openprairie.sdstate.edu/etd2/1862"],"dc:rights":["http://rightsstatements.org/vocab/InC/1.0/"],"dc:subject":["Civil and Environmental Engineering","Remote Sensing","Soil Science"],"dc:title":["Multi-Source Remote Sensing–Based Soil Moisture Prediction Using Machine Learning"],"thesis:degree_discipline":["Civil and Environmental Engineering"],"thesis:degree_level":["Thesis - Open Access"],"thesis:degree_name":["Master of Science (MS)"]},"updated_at":"2026-11-19T09:05:56Z"}