Fusion of Remotely Sensed Data with Monitoring Well Measurements for Groundwater Level Management

Abstract

In the realm of hydrological engineering, integrating extensive geospatial raster data from remote sensing (Big Data) with sparse field measurements offers a promising approach to improve prediction accuracy in groundwater studies. In this study, we integrated multisource data by applying the LMC to model the spatial relationships of variables and then utilized block support regularization with collocated block cokriging (CBCK) to enhance our predictions. A critical engineering challenge addressed in this study is support homogenization, where we adjusted punctual variances to block variances and ensure consistency in spatial predictions. Our case study focused on mapping groundwater table depth to improve water management and planning in a mixed land use area in Southeast Brazil that is occupied by sugarcane crops, silviculture (Eucalyptus), regenerating fields, and natural vegetation. We utilized the 90 m resolution TanDEM-X digital surface model and STEEP (Seasonal Tropical Ecosystem Energy Partitioning) data with a 500 m resolution to support the spatial interpolation of groundwater table depth measurements collected from 56 locations during the hydrological year 2015–16. Ordinary block kriging (OBK) and CBCK methods were employed. The CBCK method provided more reliable and accurate spatial predictions of groundwater depth levels (RMSE = 0.49 m), outperforming the OBK method (RMSE = 2.89 m). An OBK-based map concentrated deeper measurements near their wells and gave shallow depths for most of the points during estimation. The CBCK-based map shows more deeper predicted points due to its relationship with the covariates. Using covariates improved the groundwater table depth mapping by detecting the interconnection of varied land uses, supporting the water management for agronomic planning connected with ecosystem sustainability.