Showing 23 results for Kriging
Jahangard Mohammadi,
Volume 2, Issue 4 (1-1999)
Abstract
This study addresses the methodology of studying spatial variability of soil salinity. The information used is based on a semi-detailed soil survey, followed by a free survey, conducted in Ramhormoz, Khuzestan. The study of soil salinity variations was carried out using about 600 sampling points with an average distance of 500 m, at three depths of 0-50, 50-100, and 100-150 cm. To determine the spatial variability of soil salinity at different depths, the variogram which is a statistical function for the spatial variability analysis of the geographical variables was used. The results indicate that all variograms show almost the same range of 12 - 13 km which is closely related to the geographical distribution of the soil parent materials in the area. Ordinary block kriging was used to map salinity at different depths for a block dimension of 500 × 500 m. A comparison between the kriged estimates and the soil salinity map, produced during the soil survey, showed that the overall similarity between the test data and the classified kriging estimates was 40%, while the overall agreement between the test data and the soil survey salinity map was 36%. A detailed similarity calculation showed that the reliability of the classified kriging estimates representing the lowest salinity classes (S0, S1) is larger (75%) than the reliability of the soil survey salinity map representing these classes (50%). Consequently, the results indicate that geostatistical tools can be used to support the present-day procedures of soil salinity mapping.
Jahangard Mohammadi,
Volume 3, Issue 1 (4-1999)
Abstract
The analysis of the EC data set indicated that the spatial distribution of EC data of different depths are closely related to one another. It means that they are spatially cross correlated on one another and can be considered to be co-regionalized. It also implies that EC values at a particular depth contain useful information about the other depths which can be used to improve their estimation. In this research, we aimed to investigate the effects of using relevant ancillary information in the estimation procedure. To do this, cokriging was used. To evaluate this algorithm as a potential tool for mapping EC, its performance on the independent test data was evaluated and compared with the results obtained from studies using kriging. The results of the co-regionalization of EC at different depths indicated that cokriging the salinity data, although more rigorous from theoretical point of view, displayed no advantage over independent ordinary kriging at each depth. The results confirmed that cokriging improves little over ordinary kriging if the primary and auxiliary variables are almost equally sampled and all the variograms are identical. Also, ordinary kriging showed to be quite self-consistent since the predicted average salinity profile over the three depths was almost identical to the one predicted by cokriging. Considering the complexity of the cokriging and the LMC modeling, it is clear that there is no gain in using co-regionalization.
M. H. Mahdian, N. Ghiasi, S. M. Mousavy Nejad,
Volume 7, Issue 1 (4-2003)
Abstract
Point data of weather stations are not important in and by themselves. Therefore, it is necessary to change these point data into regional information. Undesirable distribution of weather stations and their data deficiency hinder the direct determination of the regional information, unless sufficient data in the study area could be provided. Providing extra data using the geostatistical methods is practical, scientific, simple and quick, but adopting a suitable method is the basic question. The objective of the present study is to find a suitable method to estimate monthly rainfall in the central region of Iran. In this regard, the methods of kriging (ordinary kriging, log-kriging, co-kriging), weighted moving average (WMA, with the power of 1 to 5), thin plate smoothing splines (TPSS, with the power of 2 and 3 and with covariable) were used. Cross validation technique was used to compare these methods.
Based on the variography analysis, the range of influence of monthly rainfall in the central region is about 450 km. The results show that TPSS, with the power of 2 and with elevation as a covariable, was the most accurate method to estimate monthly rainfall. In addition, it is preferable to use the selected interpolation method in the sub-basins with homogeneous climates instead of considering the whole region.
A. Siah-Marguee, M. H. Rashed-Mohassel, M. Nasiri-Mahallati, M. Banayan-Awal, H. Rahimiyan-Mashhadi,
Volume 10, Issue 3 (10-2006)
Abstract
This study was conducted in a sugar beet field at Collage of Agriculture Experimental Station, Ferdowsi University of Mashhad, Iran. In order to describe the pattern of spatial variations and density of Chenopodium album, Solanum nigrum, Amaranthus sp., Portulaca oleracea, Echinochla crus-galli, and Convulvulus arvense as the main prevalent annual and perennial weeds of sugar beet fields, geostatistic methods were used. Samples were taken by systematic method from the corners of (7m × 7m) grids, using (0.5m × 0.5m) quadrates in three stages (before application of herbicides, after herbicide treatment, and before harvesting sugar beets). The integrity of spatial variation of variables was determined by using variogram functions and distribution maps of species. The variograms indicated that variations of all variables did not happen by chance. The maximum and minimum ranges of variation were observed in Solanum nigrum (by 142.7m) and Portulaca oleracea (by 1.5m), respectively. Both maximum and minimum ranges of variations were related to pre herbicide application. The highest and the lowest spatial correlations were related to Amaranthus sp. (in the third sampling treatment) and Solanum nigrum (in the first stage of sampling), respectively. The spatial distribution maps confirmed the patchiness distribution of the weeds. The patch of weed was constructed from a dense point at the center, gradually tapering toward the edges. The patches were skewed across the rows and irrigation channels. The structure of patches altered during the growing season. Any information on the distribution of weeds in the fields can be useful to improve decision makings in relation to applying the herbicides, selecting the herbicide type or applying the amount of herbicide. Also it can be useful to better design of weed control programs.
S. Mohammad Zamani, Sh. Ayoubi, F. Khormali,
Volume 11, Issue 40 (7-2007)
Abstract
Evaluating agricultural land management practices requires a thorough knowledge of soil spatial variability and understanding their relationships. This study was conducted at a traditionally operated wheat field in Sorkhankalateh district, located about 25 km northeast of Gorgan, in Golestan province, Iran. Soil samples of the 0-30 cm depth were collected right after planting at the end of autumn 2004 , 100 × 180m plot in a nested grid pattern (n=101). A 1 m2 plot of wheat was harvested at each of 101 sites previously sampled at the end of spring. Statistical results showed that frequency distribution of all data was normal. The highest and lowest CV was related to grain yield (20.40%) and pH (0.59%) respectively. Variogram analysis showed that all parameters had spatial structure and the range values showed considerable variability among the measured parameters. The ranges of spatial dependence showed a variation from 23.99m for total N up to 93.92m for K. Among the parameters, total N and ESP had stronger spatial dependence while P had the lower spatial dependence. Interpolated maps of Kriging demonstrated that crop and soil properties did not have a random pattern but had a spatial distribution. The spatial distribution of total N was similar to organic matter and also there was similarity between spatial distribution of harvest index and available P. The results demonstrated that, the spatial distribution and spatial dependence level of soil properties can be different even within similarly managed farms. Variography and Kriging can be useful tools for designing soil sampling strategies, characterizing management zones and variable application rates of inputs in the precision agriculture.
A. Siah-Marguee, M.h. Rashed-Mohasel, M. Nasiri-Mahallati, M. Banayan-Aval, A. A. Mohammad-Abadi,
Volume 11, Issue 41 (10-2007)
Abstract
This study was performed in two barley fields, in Experimental Station, Agricultural College of Ferdowsi University of Mashhad in 2003. Sampling was done by systematic method in which samples were taken from the corners of 7m*7m grids using 0.5m 0.5m size quadrates in three stages (pre herbicide, post herbicide and pre harvesting stages). The results indicted that the density of annual weed seedlings in sugar beet- barley rotation was more than fallow- barley rotation, and the density of perennial weed seedlings in fallow-barley rotation was more than sugar beet- barley rotation. Map of species distribution and density confirmed patchiness distribution of the weeds. The shape and size of patches differed based on the field and weed species, but spatial distribution did not change considerably before and after the application of herbicide. Percentage of free weeds area was 11.5% and 1.5% in fallow-barley rotation and 0.6% and 0% in sugar beet- barley rotation in the first and second sampling stages, respectively. These results indicate beside emphasis on weed infestation. The result also indicates inefficacy of sugarbeet-barley rotation compared to follow-barley rotation. Apparently, the evaluation of management and paying special attention to weed dispersal within the field assist in the implementation of appropriate management strategy, which includes high efficacy, and profit for farmers as well as least damage to crops.
G Golmohamadi, S Maroufi, K Mohamadi,
Volume 12, Issue 46 (1-2009)
Abstract
In this research, using geographic information system (GIS) and different geostatistical methods including the kriging and co-kriging (ordinary, simple and universal) as well as the radial basis functions, the spatial distributions of runoff coefficient were evaluated in Hamedan province. To this end, the annual runoff were calculated in 18 existing hydrometery stations and another 11 auxiliary points, using digital elevation model (DEM) and 11 years available data of the stations. The performance criteria for evaluating the methods were mean absolute error (MAE), mean bias error (MBE), root mean square error (RMSE), and general standard deviation (GSD) along with the cross validation examination. A high regression between the runoff coefficient and watershed average slope was selected as auxiliary variable. The results showed that the runoff coefficient of the region changes between 3.5 and 85%. The findings also indicated that the universal co-krigings with spherical semi-variogram model had better performance with the values of MBE (-0.0014), MAE (0.036), RMSE (0.054) and GSD (20.152). The universal and simple kriging with spherical model were equal in runoff estimation of the region and were ranked as the second methods to this propose.
L. Khodakarami, A. Soffianian, N. Mirghafari, M. Afyuni, A. Golshahi,
Volume 15, Issue 58 (3-2012)
Abstract
Among the environmental pollutants, heavy metals according to their irresolvable and physiological effects on living organisms at low concentrations, are of special importance These elements due to low mobility are gradually accumulated in soil Being accumulated in soil, they eventually enter the food chains and threaten human health and other creatures Therefore, studying concentration distribution of heavy metals for soil pollution monitoring and maintaining environmental quality is essential In this study we investigated the effect of agricultural land use and geology on the concentration of heavy metals contamination of soil and spatial distribution map, using collected data, GIS and GeostatisticsUsing systematic stratified random sampling, 135 surface soil samples( 0-20 cm) from an area of 7262 sq km area and we measured total concentration of elements Nickel, Chromium and Cobalt and soil characteristics including pH, organic matter and texture. The mean value of elements concentrations turned out to be Cr: 88.9+22.7 Co: 17.6+3.5 Ni 63.1+17.7 mg per kg and the mean acidity is 7.8 which in the area is an indication …… property. Formetal concentrations interpolation procedures, Geostatistics was used. By the aid of spatial correlation analysis, appropriate interpolation method using functions mean absolute error and bias average error were selected. Interpolation map concentrations of heavy metals Chromium, Cobalt and Nickel with ordinary kriging method and the exponential model were developed Interpolation map analysis of heavy metals by the aid of geological and land use maps show that the distribution of the elements Chromium, Cobalt and Nickel are consistent with the geology classes However, they did not match the agriculture pattern Findings of this study in the area give us appropriate information about the concentration distribution of heavy metals Chromium, Cobalt and Nickel which can be used in monitoring and evaluation processes of heavy metals pollution in agricultural lands area. But on the other hand sampling in the areas far away from human effects, showed that the heavy metals concentration is naturally high.
Hadis Feizi, Mostafa Chorom, Arsalan Heidari,
Volume 17, Issue 64 (9-2013)
Abstract
In order to describe soils polluted with hydrocarbons, the amount and distribution pattern of soil heavy metals (Ni, Cd) in soils were studied. Soil samples were taken from one of the western oil field of Iran. The field was naturally exposed to crude oil spillage into soil and consequently was environmentally polluted during the development, production, transportation and storage of crude oil. Sampling was started near the oil wells with maximum relative contamination and continued to the remote places based on grid sampling pattern. Samples were characterized by physicochemical analysis. The results revealed different levels of total hydrocarbons (from 0.12 to 2.99 mg/kg of dry soil), Ni (from 32 to 136 mg/kg. of dry soil) and Cd (from 0 to 4mg/kg of dry soil). In addition, the role of soil agents such as pH and EC and sedimentary indexes was considerable in controlling the pollution trend in the studied area. Finally, by interpolation module and prediction of unknown values via Kriging techniques, the expansion plans were created. The extracted plans obviously illustrated the decrease in the levels of pollution indexes with the increase in distance from the given centers of pollution
E. Fathi Hafshejani, H. Beigi Harchegani,
Volume 17, Issue 65 (12-2013)
Abstract
Trends in groundwater pollution with nitrate and phosphate may be an indication of water resources management. The aims of this research were to determine changes in nitrate and phosphate concentration and changes in spatial variability patterns of nitrate and phosphate and distribution over a 5-year period. To do this, 100 agricultural wells were sampled in the years 2006, 2010 and 2011, and analyzed for nitrate and phosphate concentrations. From 2006 to 2011, the mean nitrate concentration increased from 18 to 27 mg/L and the mean phosphate concentration from 0.05 to 0.15 mg/L. Spatial patterns did not change, and spherical model described the patterns throughout this period. Maps showed that the nitrate and phosphate concentrations are higher in the south, and lower in the north of the aquifer. It seems that the presence of the municipality treatment plant, intensive cattle farming, shallower water-table and inward flow gradient may be the reasons for the higher concentration in the southern part of the aquifer. From the comparison of the maps, it was clear that the areas of less polluted classes had shrunk while the areas of more polluted classes had grown from 2006 to 2011.
H. Beigi Harchegani, S. S. Heshmati,
Volume 18, Issue 67 (6-2014)
Abstract
Shahrekord groundwater is the main source of water for drinking, and the agricultural and industrial activities of its
inhabitants. Water quality measures of scaling and corrosion can deteriorate steel-based systems used for storage or
supplying water for drinking and to industry and irrigation. The main aim of this study was to assess the spatial
variability and mapping of scaling and corrosion using Langelier index (LI) and Ryznar index (RI) and that of the
related parameters of pH, total dissolved solids (TDS), total hardness (TH), and total alkalinity (TA) in Shahrekord
groundwater. For this purpose, water samples from 97 wells were analyzed for pH, TDS, TH, and TA and LI and RI
indices were calculated. The Gaussian model best described the spatial variability of TDS while the Spherical model
was best for all other parameters. Based on LI and RI averages of, -0.13 and 7.9 respectively, Shahrekord groundwater
has a slight potential for corrosion. The values of all parameters, except RI, were lowest in the northwest and highest in
the southeast of the aquifer. In most parts and in the center of the aquifer, the values of LI ranged from -0.5 to zero
indicating negligible scaling potential. Spatial distribution of the RI index was almost inversely symmetrical to that of
LI index. LI showed strong positive correlations with its components (varying from 0.61 to 0.90) while RI had strong
negative correlations with its components (ranging from -0.66 to -0.98). LI and RI had the strongest correlations,
respectively, with pH (r=0.90) and total alkalinity (r=-0.90).
A. Taheri Tizro, H. Nozari, H. Alikhani,
Volume 20, Issue 76 (8-2016)
Abstract
To procure the status of groundwater level fluctuations in arid and semi-arid areas, it is necessary to obtain accurate forecast of fluctuations data. Time series as a linear model have been utilized to generate synthetic data and predict future groundwater level. Minitab17 software and monthly depth of groundwater level data of 20 years (1991-2011) for 25 piezometric wells of plain were used. Time series models of each well were selected and 5 years temporal forecasting was accomplished. The predicted depth of groundwater level data was converted to Groundwater level data using ARCGIS10 and GS+5.1.1 software. Ordinary kriging with a spherical variogram was selected for interpolation of groundwater level. Five years spatial forecasting was done and spatial forecasting and groundwater level drop forecasting maps were prepared. Forecasting results of groundwater level show that over the next 5 years, the area covered by two intervals of groundwater level, 1100-1140 m and 1140-1180 m, will increase and the area covered by three ranges of 1180 -1220 m, 1220-1260 m, and 1260-1300 m, will decline. Also, according to the 5-year groundwater level drop forecasting map of the plain, the highest level of groundwater level drop, more than 16 meters for Qasemabad bozorg areas, located in North East and central of the plain, and the lowest level of the groundwater level drop, about 0.5 m for Mohammad Abad Afkham Aldoleh Lands, located in outlet area of the plain, have been predicted.
M. Ayoubi, R. Sokouti, M. J. Malakouti,
Volume 20, Issue 76 (8-2016)
Abstract
This study is aimed to investigate the spatial variation of soil macronutrients such as phosphorus, potassium and organic matter using different methods of Geostatistics and Geostatistical method combined with Fuzzy logic to estimate the values of this element to provide a spatial distribution map for the proper distribution of fertilizer in the plain of Uremia. Spatial variations in soil nutrients are natural but knowing these changes for careful planning and management particularly in the agricultural lands is simply inevitable. This information is necessary to increase the profitability and sustainable agricultural management. Therefore, to estimate the changes in the elements of places not sampled, the Kriging, Fuzzy Kriging, Cokriging and Inverse Distance Weighting methods have been used in GS +. In this study, Matlab 9.1 software was used to fuzzification of the data and GIS was used for the final mapping. The parameters MAE, MBE and RMSE were used to compare these methods. The results showed that the combined method of Fuzzy Geostatistic with the mean absolute error values for the elements phosphorus, potassium and organic matter i.e. 0.17, 0.18 and 0.18, respectively, is recognized as the preferred method based on which zoning maps have been prepared for P, K and OC in GIS.
A. Khosravi-Dehkordi, M. Afyuni, A. Soffianian,
Volume 20, Issue 77 (11-2016)
Abstract
Pollutants are considered the disturbing factors of environment, and among them the heavy metals are more important considering their non-degradability and physiological effects on organisms in low concentrations. The goal of this research was to investigate the effect of industrial landuse on Cd and Pb concentrations in surface soils of the southwest Isfahan. According to satellite images and topographic maps (1:50000) of the study area, soil samples (depth: 0–20 cm) were collected using random sampling. A total of 38 surface soil samples were obtained from industrial areas (lowest distance = 1480 m) in the area of 73481 ha. Total concentrations of Cd and Pb in the digested solution were measured by Atomic Absorption Spectrophotometry (AAS). Using Arc GIS, the spatial distribution patterns and Cd and Pb variography of samples were analysed and finally the best models of spatial distribution of heavy metals were achieved. The primary results showed that the mean concentrations of Cd, and Pb of surface soil samples in industrial areas were 1.8 to 31.5 mg Kg-1 higher than the world’s mean values, respectively. Although the mean concentrations of Cd and Pb were respectively 8 to 700 mg Kg-1 lower than the standard of Iranian Department of Environment for industrial landuse.
Z. Savari, S. Hojati, R. Taghizadeh-Mehrjerdi,
Volume 20, Issue 77 (11-2016)
Abstract
Salinity and alkalinity decreases physical, chemical and biological quality of soils and as a result reduces crop yield. This study aims to evaluate spatial variability of soil salinity in Ahvaz using geostatistical approaches. Accordingly, 69 surface soil samples (0-10 cm) were collected and their electrical conductivities (EC) were measured in 1:1 soil: water extracts. The data were then analyzed using ordinary kriging (OK), log-normal kriging (LOK) and indicator kriging (IK) interpolation techniques to produce soil salinity maps. Finally, the quality control of soil maps was performed by calculation of root mean square error (RMSE) and coefficient of determination (R2). The results indicated that due to the lowest RMSE and the highest R2 values, the LOK interpolation method is the best approach in mapping soil salinity in Ahvaz. The results also illustrated that based on defined threshold values (4, 8, 16, and 32 dS m-1) the indicator kriging methods have been able to show risk of soil salinity in the area. Based on this, most of the area is covered by soils with salinity higher than 4 dS m-1. Evaluation of final soil maps showed that the highest concentrations of salts are related to the western and southwestern parts of Ahvaz city. In contrast, the lowest amounts of salinity were found in Eastern and Northern parts of the city.
M. Isazadeh, R. Arabzadeh, S. Darbandi,
Volume 20, Issue 77 (11-2016)
Abstract
Selection of optimum interpolation technique to estimate water quality parameters in unmeasured points plays an important role in managing the quality and quantity of water resources. The aim of this study is to evaluate the accuracy of interpolation methods using GIS and artificial neural network (ANNs) model. To this end, a series of qualitative parameters of samples from water taken from Dehgolan aquifer located in Kurdistan, Iran including CL, EC and PH were evaluated by any of the models. In this study, qualitative data from 56 observation wells with good dispersion in the whole plain was used. The data of 46 observation wells were used for calibration and the data of other 10 wells were used for verification of models. The results showed ANNs, IDW, and Kriging excellence and accuracy over other models in estimation of quality parameters CL, PH and EC. However the ANNs model is more accurate than other models. In case of lack of time and the need for acceptable accuracy and less risk in the estimation of qualitative parameters, the use of ANNs model is superior to other statistical models used.
M. Tayebi, M. Naderi, J. Mohammadi,
Volume 21, Issue 3 (11-2017)
Abstract
The aim of this work was to study distribution of some heavy metals in different soil particle-size fractions and to assess their spatial distribution. The study was carried out in Kafe Moor (Kerman, Iran) where the Gol-Gohar Iron Mine is located. One hundred twenty composite soil samples were randomly collected and transferred to the laboratory in bags. After air-drying, the samples were fractionated into six classes including 2- 0.5, 0.5-0.25, 0.25-0.125, 0.125- 0.075, 0.075-0.05 and <0.05 mm. Elemental concentrations (Fe, Mn, Cu, Zn, Pb and Ni) were determined using acid digestion method (HNO3, 4.0 N) and an atomic absorption spectrophotometer in each class. Ordinary Kriging technique was used for predicting spatial distribution of heavy metals. The results showed that content of metals in soil increased with decreasing particle size. The results also showed that the concentration of Fe, Mn, Cu, Zn, Pb and Ni in <0.05 mm size fraction were 2.13, 1.70, 4.79,2.43, 1.42, and 3.47 times higher than in 2-0.05 mm size fraction, respectively. In addition, mapping the concentrations of heavy metals with kiriging showed that metals pollution decreased with increasing distance from mines area.
F. Moosiri, N. Ganji Khorramdel, M. Moghaddasi,
Volume 22, Issue 1 (6-2018)
Abstract
To continue or develop the exploitation of underground water for different different uses and purposes, as well as building any water structure, set of quantitative features of aquifers can be detected. To achieve this goal, quantitative monitoring of groundwater level is only possible. Accordingly, this study compared the impact of both the concept of marginal entropy and ordinary kriging for groundwater level monitoring network design in the case Gotvand-Aghili Plain, Khuzestan province. It is important to note that a key aspect in groundwater level monitoring of the quantity measured was the variability or uncertainty in it. This created a considerable confidence to monitor and ultimately achieve favorable conditions in the future. In this study, the variability of the groundwater level was considered to evaluate the combined effects of marginal entropy and ordinary kriging. In order to determine the suitable areas for further monitoring or thinning as well as the compatibility of these two methods, the monitor network design was designed. The map classified according to the marginal entropy method, in a range between 0.07 to 5.26 of the marginal entropy change, areas with the higher rates of 2.13 in terms of density; this indicated the need for more observation wells. Ordinary Kriging method also changed the range of values; they also represented areas that needed monitoring more than 13.16. Comparison of the results obtained by the two methods showed that the marginal entropy of the kriging method with less uncertainty and by using it, there was less the need to be monitored and classified. Comparison of the two methods by the zoning map showed that fewer errors were taken to the marginal entropy method and it could be recommended for the groundwater level monitoring network design. The network was also based on the Cross validation estimation error evaluated. These tests and additional analysis were employed in this study to determine the suitable areas for the higher density of wells and the need for thinning areas. The results further confirmed the proper performance of the methods employed, as well as the superiority of the marginal entropy in the design of a small groundwater monitoring network.
E. Chavoshi, M. Afyuni, M. A. Hajabbasi,
Volume 22, Issue 2 (9-2018)
Abstract
This study covers a large agricultural and industrial area of Isfahan province, including three types of land use, i.e., agricultural, uncultivated, industrial and urban types. A total of 275 samples from surface soil (0-20 cm) were collected and water soluble fluoride concentrations of them were measured. The spatial structure of water soluble fluoride in the soils was determined by omnidirectional variogram in the GS+ software. The spatial distribution of water soluble fluoride in the soil was mapped by employing the point kriging method in the SURFER software. The results showed that the mean of the water soluble fluoride concentration in Isfahan soils (0.85 mg L-1) was higher than the mean world soils (0.53 mg L-1). The water soluble fluoride showed moderate spatial dependence, indicating that the spatial variability of water soluble fluoride was mainly controlled by intrinsic and extrinsic factors. The mean water soluble fluoride concentration was significantly higher in agricultural and urban areas, as compared with the uncultivated land. This could be due to application of phosphate fertilizer in agricultural areas and the atmospheric fallout of fluoride from the industrial sources such as steel factories. According to the generated kriging map, the higher concentration of fluoride was mainly recorded around the Zayande Rood River and in the central and western parts of the study area.
M. Ghandali, K. Shayesteh, M. Sadi Mesgari,
Volume 23, Issue 1 (6-2019)
Abstract
Determination of water quality is an essential issue in water resources management and its monitoring and zoning should be considered as an important principle in planning. In this study, in order to investigate the quality of groundwater resources (springs, wells and qanats) in Semnan watershed, first, the water quality index for drinking and agricultural purposes was obtained by means of measuring SO4, Cl, Na, Mg, PH, EC, SAR, TDS in 55 groundwater sources. For calculating the parameters weight in WQI, the fuzzy hierarchy analysis process was used with the Chang's development analysis. Due to the lack of sampling points for zoning of the entire area, regarding the existence of EC data for the majority of groundwater resources used in this catchment (354 sources), as well as the high correlation (Adjusted R2=0.99) between WQI with EC, the mentioned indexes of other resources were estimated based on the regression relationship with EC. To analyze the spatial distribution and monitor the zoning of the groundwater quality, the ArcGIS version 10.3 and Geostatistical method such as simple Kriging and ordinary Kriging were used; additionally certain methods including Inverse distance weighting and Radial Basis Function were utilized. The performance criteria for evaluating the used methods including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), %RMSE and R2 were used to select the appropriate method. Our results showed that the ordinary Kriging and Radial Basis Function were the best methods to estimate the groundwater quality.
