Spatiotemporal distribution and prediction of groundwater level linked to ENSO teleconnection indices in the northwestern region of Bangladesh

摘要

This study aims to investigate spatiotemporal trends and magnitudes of groundwater level for the period of 1981-2017 in the northwestern region of Bangladesh and to predict the groundwater level changes from 2018 to 2025 through autoregressive integrated moving average (ARIMA) modeling. The study explores the relationship between groundwater level and El-Nino Southern Oscillation (ENSO) teleconnection indices during 1981-2017 in the study area. Results of MMK (Modified Mann-Kendall) Z statistics showed a significant downward trend of groundwater level (p<0.05), and the years 1994, 2002, 2004, 2006 and 2007 were the trend turning points of groundwater level depletion identified by Pettitt's test. The overall decreasing rates of groundwater level vary from 0.1 to 1.3 (m/decade). Wavelet spectrum was employed to show the temporal distribution patterns of groundwater level, suggesting a prolonged drought period in Rangpur subregion of northwestern Bangladesh. The Pearson correlation and cross-correlation results revealed that there was a weak relation between ENSO indices and groundwater level due to less anthropogenic influence in the study area. The outcomes of ARIMA model were consistent with MMK test trend analysis which predicted a similar downward tendency of groundwater level from 2018 to 2025, indicating an alarming situation, especially for Rajshahi subregion (0.8 m/decade). Overall, all the test results indicate a strong concurrence relationship among them. The results of the study recommend the need for rainwater harvesting during monsoon season, multi-disciplinary groundwater researches and monitoring programs as measures for future sustainable groundwater management on northwestern Bangladesh.