Climate change and anthropological impacts on a karst aquifer: a multi- statistical assessment

Abstract

Many karst aquifers are the sole source of municipal (drinking) water in Iran. The purpose of this study is to detect trends in rainfall and temperature series as climate variables and the groundwater level time series as a hydrologic variable to assess the anthropogenic and climate change impact on the strategic Sarbalesh aquifer in Fars province, Iran, employing the Mann-Kendall (MK) and modified MK tests, the combination of discrete wavelet transform (DWT), MK, and Sequential MK analyses, and the Pettitt-Mann-Whitney (PMW) and Cumulative sum methods. The MK and MMK tests analyses showed a significant declining trend in the groundwater level and a negative/positive trend in the rainfall/temperature series, particularly in the annual series at the 5% level of significance. The time series were decomposed into the approximation (A) and details (Ds) sub-signals by the DWT. The dominant periodicities contributing to the production of trend in the original series were found by comparing the approximate equality of the MK Z-values, the correlation coefficient, and the likeness of the Sequential MK curves of the original data and their respective A + Ds components. The monthly rainfall, temperature, and groundwater series showed 8-month, 8-month, and 64-month prominent periodicities, respectively. The annual rainfall, temperature, and groundwater series exhibited the prominent periodicities of 4-year, 2-year, and 8-year, respectively. By PMW and Cumulative sum approaches, we detected abrupt shifts (change points) in the studied time series. The same change year point in the rainfall series and the Southern Oscillation Index (SOI) series and the negative correlation between rainfall and El Niño-Southern Oscillation (ENSO) cycles indicated that the climate change and hence the La Niño phenomenon increased SOI after the change year led to a decrease in precipitation from November to April in the study area. Our multi-statistical approach proved that the drastic decline in groundwater level is due to the continuous over-exploitation of water storage from the aquifer in the last 32 years and reduced rainfall and increased temperature (particularly in the annual series) have triggered the declining trend. Our analysis provides an elaborate view of future hydro-climatic conditions, it can be used as a foundation for designing and planning water resources within the study area and elsewhere.