Are Geostatistical Interpolation Methods Better than Deterministic Interpolation Methods in Mapping Salinity of Groundwater?

Research areas:

Volume 4,Issue 5, Sept. 2017

Year:

2016

Type of Publication:

Article

Keywords:

Geostatistical, GIS, Groundwater, Interpolation, Salinity

Authors:

M. H. J. P. Gunarathna; K. G. S. Nirmanee; M. K. N. Kumari

Journal:

IJRIES

Volume:

Number:

Pages:

59-64

Month:

May

ISSN:

2394-1375

BibTex:

@article{117,
author = "M.H.J.P. Gunarathna and K.G.S. Nirmanee and M.K.N. Kumari",
abstract = "Groundwater is one of the most valuable natural resources, which supports human health, economic development and ecological diversity. It has the ability to mobilize and transport soluble salts when it moves through the regolith and the geological formations. Anthropogenic activities can influence this process affecting its quality. Salinity is a global problem, which is more severe in water scarce areas. In the management of natural resources, geostatistical methods in geographic information systems are applied, because it shows the spatial relationship between neighbouring values of the hydrochemical parameters, in order to predict the values from non-sampled areas. In this study the geostatistical analyst tool was used to compare the accuracy of different interpolation methods by using spatial pattern of electrical conductivity (EC) and total dissolved solids (TDS) of groundwater in Malwathu Oya cascade-I in Anuradhapura District, using Arc GIS 10.2. The cascade was divided into 1 km2 grids and forty wells were purposely selected at two wells per grid, to assess the groundwater EC and TDS during the dry period. Inverse Distance Weighted (IDW), Radial Basis Function (RBF), which are deterministic interpolation methods and Kriging and Empirical Bayesian Kriging, which are geostatistical interpolation methods were used with different parameters in this study.  Results revealed that, Empirical Bayesian Kriging Method with thin plate semi variogram model reported the least Root Mean Square Error (RMSE) values for the spatial variation of both groundwater EC and TDS.  Therefore geostatistical interpolation method performs better than deterministic interpolation methods for mapping groundwater salinity.",
issn = "23941375",
journal = "IJRIES",
keywords = "Geostatistical, GIS, Groundwater, Interpolation, Salinity",
month = "May",
number = "3",
pages = "59-64",
title = "{A}re {G}eostatistical {I}nterpolation {M}ethods {B}etter than {D}eterministic {I}nterpolation {M}ethods in {M}apping {S}alinity of {G}roundwater?",
volume = "3",
year = "2016",
}

Abstract:

Groundwater is one of the most valuable natural resources, which supports human health, economic development and ecological diversity. It has the ability to mobilize and transport soluble salts when it moves through the regolith and the geological formations. Anthropogenic activities can influence this process affecting its quality. Salinity is a global problem, which is more severe in water scarce areas. In the management of natural resources, geostatistical methods in geographic information systems are applied, because it shows the spatial relationship between neighbouring values of the hydrochemical parameters, in order to predict the values from non-sampled areas. In this study the geostatistical analyst tool was used to compare the accuracy of different interpolation methods by using spatial pattern of electrical conductivity (EC) and total dissolved solids (TDS) of groundwater in Malwathu Oya cascade-I in Anuradhapura District, using Arc GIS 10.2. The cascade was divided into 1 km2 grids and forty wells were purposely selected at two wells per grid, to assess the groundwater EC and TDS during the dry period. Inverse Distance Weighted (IDW), Radial Basis Function (RBF), which are deterministic interpolation methods and Kriging and Empirical Bayesian Kriging, which are geostatistical interpolation methods were used with different parameters in this study. Results revealed that, Empirical Bayesian Kriging Method with thin plate semi variogram model reported the least Root Mean Square Error (RMSE) values for the spatial variation of both groundwater EC and TDS. Therefore geostatistical interpolation method performs better than deterministic interpolation methods for mapping groundwater salinity.

Full text:

IJRIES_117_Final.pdf Online version

Back