A GIS tool for hydrogeological water balance evaluation on a regional scale in semi-arid environments

Abstract

A GIS tool to evaluate hydrogeological water balance based on a mass-balance model applied to surface and subsurface systems is discussed. The tool is designed for managers responsible for groundwater resource planning during conditions of water shortage. In developing the tool, the natural groundwater recharge was evaluated through the application of a soil water balance equation, and defined as the difference between the inflows (rainfall, irrigation) and the outflows (plant evapotranspiration, surface run-off). A distributed approach was used in the soil water balance equation to account for the spatial variability of climate and landscape features. Conversely the groundwater balance was calculated on a watershed or aquifer scale, using a lumped water balance equation, in which withdrawals for different uses were estimated together with inflows from other water bodies and coastal outflows. The model was implemented on a GIS platform with an automatic routine that manages all the data sets required and allows for the forecasting of groundwater storage volumes. Furthermore, the model was able to evaluate agricultural water demands under different climatic and management scenarios. A tool which provides a summary of the results and performs a statistical analysis for any portion of the study area was also implemented. The model was applied to a coastal region of Southern Italy. The averaged groundwater balance calculated by the model was in agreement with the piezometric head and chlorine concentration trends measured in selected monitoring wells.

Introduction

The European Union, in its water framework directive (WFD) states new principles and guidelines in water resources policies. This document points to “the need to establish procedures for the regulation of abstraction of freshwater and for the monitoring of freshwater quality and quantity” (European Union, 2000). The WFD established a framework for the protection of groundwater, which promotes “sustainable water use based on a long-term protection of available water resources”. Furthermore, member States are required to “ensure a balance between abstraction and recharge of groundwater”, with the aim of achieving good groundwater quality and quantity within at most 15 years. Moreover, according to the Committee on Hydrologic Science (1999), research on climate-hydrology linkages should be broadened to address issues of groundwater recharge, evaporation, and basin-scale water balance.

In this paper we present a GIS model based on a mass-balance approach, applied to surface and subsurface systems, in order to provide reasonable estimations of groundwater recharge in poorly gauged environments.

In semi-arid regions groundwater balance is strictly connected with surface water balance, which in turn is directly governed by climate, and in particular, precipitation and evaporative energy. However, climate alone is not sufficient to explain the variability showed on smaller time scales because of the complex dynamic interactions between climate and landscape characteristics. Eagleson (1978) investigated the influence of climate–soil–vegetation interactions on expected annual water balance, introducing the concept of dynamic water balance, through physically based process descriptions and landscape variables. Milly (1994), in order to explore the basis of regional variability in water balance, demonstrated that variability of soil storage capacity and seasonal climate patterns can explain departures from long-term expected behaviour.

From these studies, it emerges that relatively simple approaches can be used to evaluate water balance correctly, especially on a large scale, provided they take into account the fundamental variables of precipitation, potential evaporation and soil storage capacity. Several examples in the literature show that regional and continental scale simulations can be performed based on simple lumped storage representations of water balance (Reed, 1999; Milly and Dunne, 1994). However, when a more detailed description is required (i.e. on a smaller spatial and temporal scale), additional variables and processes must be considered to account for the causes of variability of a smaller magnitude (Sivapalan and Woods, 1995; Boughton, 1986; Mohseni and Stefan, 1998). For natural groundwater recharge estimations, some authors have proposed approaches based on the interface between surface hydrological water balance models and groundwater flow models, in which the deep infiltration terms, output of the 2-dimensional surface models, are used as source terms for the groundwater 3-dimensional transient flow models (Ruud et al., 1999; Kim et al., 2000). This approach, though very useful, is rather complicated because of a demanding hydrogeological parameterisation and feedback calibration.

The present study was motivated by the need to develop a reliable means for estimating ground water balance on a watershed scale, in the context of the controlling climatic and landscape characteristics. Particularly in semi-arid environments, often lacking in hydrometric records, the reliability of groundwater recharge estimation depends on the accuracy of landscape feature description inside the water balance model, without any possible calibration. The proposed model yields natural groundwater recharge on a monthly basis, through the distributed application of the soil water balance equation, evaluated as the difference between the inflows (rainfall, irrigation) and the outflows (evapotranspiration, surface runoff). In order to calculate the groundwater balance on a watershed or hydrogeological unit scale, withdrawals for different uses are estimated together with the inflows from other water bodies and the coastal outflows where aquifer borders correspond to the coast line. For this purpose, GIS offers an effective support for hydrological modelling. Such a tool is capable of archiving, analysing and handling the large amounts of data required to describe hydrological processes on different scales (Fuhrmann, 2000; Asadi and Hale, 2001). The proposed model was implemented on a PC using the object-oriented AVENUE scripting language available with ArcView^® (Environmental Systems Research Institute Inc., 1996).

The model provides estimations of recharge rates and groundwater withdrawals due to irrigation demands, avoiding the hydrogeological parameterisation required by groundwater flow models.

An application of the model to a region of about 19,000 km² in Southern Italy is presented here, as an example of monthly and annual water balance calculation and analysis.