Comparison of parametric and non-parametric methods for trend identification in groundwater levels in Sirjan plain aquifer, Iran

Rasoul Mirabbasi,Deepak Jhajharia,Farshad Ahmadi

doi:10.2166/nh.2020.041

Rasoul Mirabbasi, Deepak Jhajharia + Show 1 more

Open Access

https://doi.org/10.2166/nh.2020.041

Copy DOI

Abstract

Abstract In the present study, five parametric and non-parametric methods: linear regression (LR), conventional Mann–Kendall (MK), modified Mann–Kendall (MMK), Spearman's Rho (SR) and Innovative Trend Method (ITM) were used to identify trends in the groundwater levels of 60 piezometers distributed uniformly across Sirjan plain, Iran, from 2005 to 2018. The LR method was found to be affected by the presence of outliers and autocorrelation. The conventional non-parametric tests (MK and SR) were not able to offset the effects of the autocorrelations between the groundwater level data. The ITM method was also found to be a not so comprehensive and precise statistical tool for trend analysis because it does not provide a quantitative index for identifying trend significance. Therefore, the MMK test was found to be the most appropriate trend analysis method among the five trend identification methods used in this study by eliminating the effect of all significant autocorrelation coefficients. The results of the MMK test showed that the groundwater levels in Sirjan plain had witnessed significant decreasing trends during the study period. In only 24 months (out of a total 10,080 studied months), no significant decreasing trends in groundwater levels were observed.

Highlights

IntroductionThe main objective of trend analysis is to identify the existence or non-existence of significant increasing or
Five tests used in identifying trends, namely, linear regression (LR), Spearmen’s Rho (SR), classic Mann–Kendall (MK), modified Mann–Kendall (MMK), and Innovative Trend Method (ITM), were applied with the where ri is the i delayed autocorrelation coefficient and V(S) is estimated using Equation [11]
In the case of the three non-parametric tests used in this study, non-significant increasing trend in groundwater level is observed in the spring season in the Sirjan plain

Summary

Introduction

The main objective of trend analysis is to identify the existence or non-existence of significant increasing or. While applying the non-parametric test for trend analysis, it is assumed that there is no significant autocorrelation in the given time series. Hamed & Rao ( ) investigated the effect of existence of autocorrelation in time series on trend detection by using non-parametric tests. They proposed a methodology for eliminating the effect of autocorrelation from a data series and applied the proposed method for identifying trends in precipitation and stream flow time series. On using the MK test after eliminating the autocorrelation effect from time series instead of the classic MK test, an increase in the accuracy of trend detection was observed (Hamed & Rao )

Methods

Results

Conclusion