Identification of groundwater recharge potential zones using AHP …
133
Saaty’s Analytic Hierarchy Process (AHP) was used to delineate artificial recharge
zones as well as to identify favorable artificial recharge sites in the West Medinipur
district of West Bengal, India [11]. A study over a sub-watershed of River Kanhan, in
Nagpur District, Maharashtra, India, was carried out in GIS using the land use, soil
types, topography, and rainfall-runoff model data, where AHP with expert’s judgment was used for ranking the sites [12]. The next advancement in the method came
when [13] coupled GIS with MCDM using fuzzy rules to yield more precise results
in selecting the site for managed aquifer recharge.
Later on, [14] used modern satellite imagery to revise the GIS and AHP coupling
techniques. In current years, software to process high-resolution satellite data has
been improved drastically. [15] used high-resolution layers to find out groundwater (GW) recharge potential zones using AHP. A combination of state-of-the-art
groundwater potential mapping (GPM) tool C5.0 along with random forest (RF),
and multivariate adaptive regression splines (MARS) algorithms were used recently
for generating GPMs in the eastern part of Mashhad Plain, Iran [16]. Other known
machine learning-based methods adopted in recent studies include boosted regression tree (BRT), classification and regression tree (CART), support vector machine
(SVM), and genetic algorithm optimized random forest (RFGA) [17, 18]. A sensitivity analysis of the effect of sample size on the accuracy of different individual
and hybrid models (i.e., adaptive neuro-fuzzy inference system (ANFIS), ANFISimperial competitive algorithm (ANFIS-ICA), alternating decision tree (ADT), and
random forest (RF)) was carried out by [19].
Although very effective and accurate machine learning and artificial intelligencebased GIS techniques are adopted in current studies, the accuracy and efficiency
of most of the approaches heavily depend on the available data size and accuracy.
However, most of the areas of acute groundwater depletion problem in the Indian
subcontinent suffer from the scarcity of accurate and updated spatial data of most
of the critical parameters. Hence, we wanted to evaluate the applicability of a relatively simple and computationally inexpensive approach to delineate the groundwater
potential zones with minimal available data. For the present study, AHP is used as
a scientific tool to estimate weightage, coupled with a simple Mamdani-based hierarchical fuzzy controller to delineate groundwater recharge potential zones over a
data-scarce area.
Purulia district of West Bengal is one of the most water-scarce areas of India.
Although this district receives 1400 mm of average rainfall during the monsoon
season, most of the precipitation disappears through Kumari and Kangsabati river
systems as runoff. Low retention capacity of the soil and the presence of hard crystalline rocks beneath the ground create constraints to groundwater development in
the area. In the present study, an endeavor has been made to delineate groundwater
recharge potential zones in some severely water-scarce blocks of Purulia district of
West Bengal using GIS and fuzzy logic techniques to overcome the limitations due
to unavailability of ground-based observational data.
133
Saaty’s Analytic Hierarchy Process (AHP) was used to delineate artificial recharge
zones as well as to identify favorable artificial recharge sites in the West Medinipur
district of West Bengal, India [11]. A study over a sub-watershed of River Kanhan, in
Nagpur District, Maharashtra, India, was carried out in GIS using the land use, soil
types, topography, and rainfall-runoff model data, where AHP with expert’s judgment was used for ranking the sites [12]. The next advancement in the method came
when [13] coupled GIS with MCDM using fuzzy rules to yield more precise results
in selecting the site for managed aquifer recharge.
Later on, [14] used modern satellite imagery to revise the GIS and AHP coupling
techniques. In current years, software to process high-resolution satellite data has
been improved drastically. [15] used high-resolution layers to find out groundwater (GW) recharge potential zones using AHP. A combination of state-of-the-art
groundwater potential mapping (GPM) tool C5.0 along with random forest (RF),
and multivariate adaptive regression splines (MARS) algorithms were used recently
for generating GPMs in the eastern part of Mashhad Plain, Iran [16]. Other known
machine learning-based methods adopted in recent studies include boosted regression tree (BRT), classification and regression tree (CART), support vector machine
(SVM), and genetic algorithm optimized random forest (RFGA) [17, 18]. A sensitivity analysis of the effect of sample size on the accuracy of different individual
and hybrid models (i.e., adaptive neuro-fuzzy inference system (ANFIS), ANFISimperial competitive algorithm (ANFIS-ICA), alternating decision tree (ADT), and
random forest (RF)) was carried out by [19].
Although very effective and accurate machine learning and artificial intelligencebased GIS techniques are adopted in current studies, the accuracy and efficiency
of most of the approaches heavily depend on the available data size and accuracy.
However, most of the areas of acute groundwater depletion problem in the Indian
subcontinent suffer from the scarcity of accurate and updated spatial data of most
of the critical parameters. Hence, we wanted to evaluate the applicability of a relatively simple and computationally inexpensive approach to delineate the groundwater
potential zones with minimal available data. For the present study, AHP is used as
a scientific tool to estimate weightage, coupled with a simple Mamdani-based hierarchical fuzzy controller to delineate groundwater recharge potential zones over a
data-scarce area.
Purulia district of West Bengal is one of the most water-scarce areas of India.
Although this district receives 1400 mm of average rainfall during the monsoon
season, most of the precipitation disappears through Kumari and Kangsabati river
systems as runoff. Low retention capacity of the soil and the presence of hard crystalline rocks beneath the ground create constraints to groundwater development in
the area. In the present study, an endeavor has been made to delineate groundwater
recharge potential zones in some severely water-scarce blocks of Purulia district of
West Bengal using GIS and fuzzy logic techniques to overcome the limitations due
to unavailability of ground-based observational data.