Assessment and spatial distribution of drought Contact with reduced quality of groundwater based on GRI indicators in GIS

Drought of Groundwater happens when aquifers as an important source of supplying water, are effected by long-term drought. Groundwater resource index (GRI) have been used for long-term change of piezometric and observational data depth of aquifers. In this research, the results of calculations about drought happens with regard to spatial distribution of it within 609 plain of Iran, in GIS software have been compared with calculations results about drinkable water quality index of aquifers resources. This comparison has been done by using FAHP method and in two consecutive ten-year periods and long-term period from 1994 until 2015. In order to calculate of GRI index, in first step in the spreadsheet, 76% of primary missing data, by using R-squared method have been reconstructed, scrutinized, and outlier data have been eliminated. In order to create the interpolation map, we needed to evaluate and process raw data. In this regard, first, by using existence assessment tool of improbability trend in data, this purpose has been achieved in 3 dimensional diagrams. The existence of second-rate improbability trend from north to south and east to west in the first mean data of GRI index in long term period was approved. By recognizing the main factor approach of creating interpolating layers through geostatistical tool we created continuous level of GRI index number for all years and for three intervals of ten, ten and twenty years. Selection of best method based on the comparative evaluation among created samples and average comprehensive range of 1994-2015 period. For this purpose, in two class of deterministic methods of Inverse Distance Weighting (IDW) with three powers of 1, 2, 3. In addition by using GPI with three rates of 1, 2, 3 and RBF method and geostatistical class methods including; ordinary kriging, universal kriging and simple kriging with 4 mathematical functions: Stable, J-Bessel, Gaussian and spherical continuous maps have been created. With another exclusive method of geostatistical method called; Empirical Bayesian kriging by using 100 repetitions of Semivariogram continuous amount was created. The reason for selecting simple Kriging with stable function as final approach was that, with comparison with all samples, mean error decreased to 0.002 unit, Root-Mean-Square (RMS) decreased to 0.669 unit, Mean-Square (MS) decreased to 0.003 unit, Root-Mean-Square Standardized (RMSS) decreased to 0.958 unit, and last ASE decreased and reached close to 0.699, equal to amount of RMS. Similar correlation of changes in two categories of time for GRI and drinkable water quality, revealed a close and direct relation between decrease of aquifer water level and increase of class number that means the decrease of water quality. The result of GRI index calculation was shown that only 3.3% drought close to normal in first decade, in second decade reach to 70.8%. And in similar manner, Wet year class number close to normal have been decreased from 87.1% to 27.7%. these changes in term of type, were also seen in Wet year mild class, sever class and very severe class. Finding revealed this fact that central regions of Iran, some parts of Khorasan Razavi province, northern Khorasan province, extensive area of Fars province and major part of south of Iran have experienced severe drought class. Based on the evaluation of drought maps and diagram of increase the volume of withdrawal from the Groundwater operation resources, between 1994 until 2015, it can be stated that, the most important factor for changes is the wasteful withdrawal from operation wells with agricultural function. Specifically, in second decade of evaluation period in this research, drought index near to normal has continued with sever slope. Therefore, replacement a major part of cells in middle drought class and severe drought class have been observed, especially in last year. Only in 2001-2007 period, there is no special drought and wet near to normal has been happened repeatedly. In order to evaluate the relation between decrease of Groundwater and decrease of water resource quality by using current calculation it was necessary to continuous level map of first and second period of GRI and Groundwater quality index by using date functions be standardize. Current statistics has been placed between two amounts of 0 and 1 by using linear date function, with regard to these principles that minimum amount be referred to maximum happens of drought and maximum amount be referred to maximum happens of wet year. The output results of FAHP were between 0 and 1, and higher values represent the areas with higher qualitative potential with aim of using for drinkable water sources or more desirable water resources. In FAHP method the quantitative final results value depends completely on the kind of operator. By selecting Gama operator, for 20 years period, maximum calculated amount was equal to 0.148 with 0.136 average and 0.011standard deviation. In contrast, for ten years period from 1994 to 2004, in sum, 13657 Raster cells layer, for maximum, average and standard deviation; were respectively 0.149, 0.136 and 0.011. And in ten years period between 2004-2015 years, decrease of water quality was determined with average of 0.124 and maximum amount of 0.147 and standard deviation of 0.015. scatter points vector, revealed this fact that there is 43% correlation between two columns of GRI standard statistics in two periods, 1994-2004 and 2004-2015. In addition, by drawing the scattered points between two column of standard value data of drinkable water quality index in FAHP method and in two ten periods, 1994-2004 and 2004-2015; 39% correlation was approved.