Development of spatially varying groundwater-drawdown functions for land subsidence estimation

Study region: Choshui River alluvial fan, Taiwan. Study focus: Land subsidence caused by groundwater overexploitation is a critical global problem. The spatial distribution of land subsidence is crucial for effective environmental management and land planning in subsidence prone areas. Because of the nonlinear relationship between subsidence and drawdown due to groundwater exploitation in heterogeneous aquifers, a spatial regression (SR) model is developed to effectively estimate nonlinear and spatially varying land subsidence. Considering various data inputs in the Choshui River alluvial fan, the SR model offers a robust method for accurately estimating the spatial patterns of subsidence using only drawdown as input data. New hydrological insights for the region: Without requiring extensive calibration or an elaborate numerical groundwater flow and subsidence model, the model provides annual subsidence patterns using a spatially varying relationship between drawdown and resulting land subsidence. Results show that the largest water-level cone of depression occurs in the distal fan area. Nonetheless, the calculated subsidence bowl closely approximates the observed one located much farther inland. The root-mean-square-errors (RMSEs) of annual subsidence is less or equal to 0.76 cm for the SR. Results indicate that the SR model reasonably estimates the spatial distribution of the skeletal storage coefficient in the aquifer system. The large coefficient that represents high potential of inelastic compaction occurs in the southern inland area, whereas the small coefficient that represents elastic compaction occurs in the northern area and proximal fan. Furthermore, this method can be used efficiently for subsidence management/ regulation and might be widely used for subsidence estimation solely based on drawdown.

Automated summary

The study focused on land subsidence caused by groundwater overexploitation in the Choshui River alluvial fan in Taiwan. A spatial regression model was developed to estimate the spatial distribution of subsidence, using drawdown as input data and showing promising results with low RMSE. The model provided new hydrological insights for the region and could be used effectively for subsidence management based solely on drawdown data.