RESUMO
Groundwater resources are one of the essential aspects of achieving self-sufficiency in a country's agricultural production, poverty alleviation, and socioeconomic development, particularly in agricultural heritage management and practices. In the Barind Tract in Bangladesh, groundwater levels have steadily declined due to growing irrigation demand. Surface water sources become scarce during the dry season, and groundwater levels fall to levels that make minimum cultivation challenging. In these circumstances, determining the current status of groundwater is key to any action in the future. This study investigated the existing geospatial pattern and critical zone of groundwater level in Chapainawabganj District, a significant area of the Barind Tract of Bangladesh, and predicted future groundwater levels considering multiple factors. Kriging, a sophisticated geostatistical method, was performed to examine the geographical pattern and groundwater variations, and time series analysis was employed to determine data trends and make future projections. The current study used groundwater level data from 23 monitoring stations over 10 years (2009-2018). Exponential, Gaussian, and Spherical models were cross-matched here for the best predictor model in four fitness measures to determine groundwater concentrations (RMSE, ME, RMSS, ASE), and Box-Jenkins ARIMA (3,1,0) was found best-fit for predictions, and variance estimation. Likewise, cross-validation has been assessed for the accuracy of anticipated results across spatial scales. Although more research is needed to identify the underlying mechanisms, critical zones, and their pattern of modification, possible recharge zones and their locations have been identified. Future groundwater levels, critical zones, and recharge locations have been indicated for the research area and potential recommendations.
