A low-cost approach for soil moisture prediction using multi-sensor data and machine learning algorithm.

A high-resolution soil moisture prediction method has recently gained its importance in various fields such as forestry, agricultural and land management. However, accurate, robust and non- cost prohibitive spatially monitoring of soil moisture is challenging. In this research, a new approach involving the use of advance machine learning (ML) models, and multi-sensor data fusion including Sentinel-1(S1) C-band dual polarimetric synthetic aperture radar (SAR), Sentinel-2 (S2) multispectral data, and ALOS Global Digital Surface Model (ALOS DSM) to predict precisely soil moisture at 10 m spatial resolution across research areas in Australia. The total of 52 predictor variables generated from S1, S2 and ALOS DSM data fusion, including vegetation indices, soil indices, water index, SAR transformation indices, ALOS DSM derived indices like digital model elevation (DEM), slope, and topographic wetness index (TWI). The field soil data from Western Australia was employed. The performance capability of extreme gradient boosting regression (XGBR) together with the genetic algorithm (GA) optimizer for features selection and optimization for soil moisture prediction in bare lands was examined and compared with various scenarios and ML models. The proposed model (the XGBR-GA model) with 21 optimal features obtained from GA was yielded the highest performance (R2 = 0. 891; RMSE = 0.875%) compared to random forest regression (RFR), support vector machine (SVM), and CatBoost gradient boosting regression (CBR). Conclusively, the new approach using the XGBR-GA with features from combination of reliable free-of-charge remotely sensed data from Sentinel and ALOS imagery can effectively estimate the spatial variability of soil moisture. The described framework can further support precision agriculture and drought resilience programs via water use efficiency and smart irrigation management for crop production.

A novel intelligence approach based active and ensemble learning for agricultural soil organic carbon prediction using multispectral and SAR data fusion.

Monitoring agricultural soil organic carbon (SOC) has played an essential role in sustainable agricultural management. Precise and robust prediction of SOC greatly contributes to carbon neutrality in the agricultural industry. To create more knowledge regarding the ability of remote sensing to monitor carbon soil, this research devises a state-of-the-art low cost machine learning model for quantifying agricultural soil carbon using active and ensemble-based decision tree learning combined with multi-sensor data fusion at a national and world scale. This work explores the use of Sentinel-1 (S1) C-band dual polarimetric synthetic aperture radar (SAR), Sentinel-2 (S2) multispectral data, and an innovative machine learning (ML) approach using an integration of active learning for land-use mapping and advanced Extreme Gradient Boosting (XGBoost) for robustness of the SOC estimates. The collected soil samples from a field survey in Western Australia were used for the model validation. The indicators including the coefficient of determination (R2) and root - mean - square - error (RMSE) were applied to evaluate the model's performance. A numerous features computed from optical and SAR data fusion were employed to build and test the proposed model performance. The effectiveness of the proposed machine learning model was assessed by comparing with the two well-known algorithms such as Random Forests (RF) and Support Vector Machine (SVM) to predict agricultural SOC. Results suggest that a combination of S1 and S2 sensors could effectively estimate SOC in farming areas by using ML techniques. Satisfactory accuracy of the proposed XGBoost with optimal features was achieved the highest performance (R2 = 0.870; RMSE = 1.818 tonC/ha) which outperformed RF and SVM. Thus, multi-sensor data fusion combined with the XGBoost lead to the best prediction results for agricultural SOC at 10 m spatial resolution. In short, this new approach could significantly contribute to various agricultural SOC retrieval studies globally.