Recharge Assessment in Greek Karst Systems: Methodological Considerations and Implications

Abstract

The recharge of karst aquifers is significant for the effective management of groundwater resources, and its estimation should be tailored to accommodate their specific hydrogeological characteristics. This study provides a two-step methodological approach for the determination of karst aquifer recharge. Initially, mean annual recharge rates were estimated in the karst system of Ziria (Southern Greece) utilizing the APLIS and modified APLIS methods in order to decipher which was the most suitable version for recharge assessments. The results indicated similar mean recharge rate values at 42.7% and 41.4%, respectively, but significant differences in the spatial distribution. The modified methodology emerged as a more accurate and realistic approach, mainly due to the incorporation of permeability assessments. The final phase of the methodological approach involved the application of modified APLIS in two additional karst hydrosystems, Planitero and **romero, while a quantitative cross-comparison of the recharge rates was obtained for a deeper understanding of the factors controlling the groundwater recharge process. In Ziria, recharge rates exhibit a relatively uniform distribution throughout the area, with a median value of 46.7%. Conversely, in Planitero, High recharge rates (60–80%) occupy 56.8% of the surface, while in **romero, Moderate recharge rates (40–60%) dominate, representing 53.4% of the land coverage. These variations underscore the spatial heterogeneity of recharge within the karst systems, highlighting the importance of considering local geological and hydrological conditions in its assessments. The methodological approach of this study is flexible and can be adapted to different karst sites for the determination of recharge regimes, contributing to the alleviation of the groundwater depletion issue.

1. Introduction

Groundwater resources play an immense role in today’s society [1,2], since the universal demand for water supply is crucially increasing. The achievement of sustainable groundwater management is a pressing and often addressed issue due to its critical role in meeting basic human requirements [3]. The escalating need for freshwater, driven by global population growth and rapid urbanization, can cause groundwater reservoirs to face over-extraction and depletion, compromising their sustainability. Simultaneously, the issue of groundwater contamination has become extensive, with pollutants from agricultural runoff, industrial discharges, and improper waste disposal infiltrating aquifers.

Besides the ramifications that can arise from human activities, climatic causes may burden aquifers further, with dire impacts on their resources. In an era dominated by the pervasive effects of climate change, Earth’s hydrological cycle faces unprecedented challenges with shifts in temperature regimes, alterations in precipitation patterns, and intensified extreme weather events. These climate-induced changes can cause direct and indirect repercussions on global reserves [4], disrupt traditional recharge mechanisms, and intensify water scarcity. The increasing occurrence of droughts may lead to changes in the water balance, while rising global temperatures contribute to heightened evaporation rates, further amplifying water scarcity in vulnerable regions. The challenges facing groundwater resources have reached a critical juncture, potentially influencing their global distribution in the future. This underscores the imperative to address numerous water-related issues in a sustainable manner. Groundwater depletion constitutes a universal concern due to the unbalanced abstractions in comparison to the natural recharge. Obviously, the study of groundwater recharge is quite challenging, especially in karst aquifers.

At the forefront of hydrological environments, karst aquifers provide high-quality groundwater that is utilized either for drinking water supply or irrigation purposes. Climate-induced shifts can cause severe repercussions, leading to both a degradation in the availability and a compromise in the quality of water resources [5,6,7,8]. At the same time, with distinct characteristics and processes, karst aquifers are able to provide substantial amounts of groundwater, often resulting in their overexploitation [9,10], in order to meet the water demands of the constantly increasing population.

In Greece, there are numerous carbonate massifs distinguished by a diverse array of geological, geomorphological, and hydrogeological features, leading to a substantial variation in the degree of karstification across the country. Over the last decade, increasing research efforts have been directed towards the characterization, protection, and sustainable management of native karst aquifers, highlighting their significant value as water sources in Greece [11]. Up to 80% of the domestic karst systems are defined by good groundwater quality and quantity conditions, while qualitative issues are recorded mainly in coastal aquifers on islands, represented by a percentage of 5% [12].

The viable management of groundwater use requires the aversion of resource depletion. Recharge is one of the most significant parameters to be assessed in order to sustain the overall effective function of an exploited aquifer. Its approximation is essential to evaluating the maximum volume of water obtainable and to avoid overutilization [13]. Stable isotopes [14,15,16,17], lump and spatial models (e.g., Karstmod, VarKarst, Modflow-CFP), as well as index-based methods (e.g., APLIS, KARSTLOP), are widely used in order to determine the recharge process in karst environments [11,18,19,20,21,22,23,24]. Due to the heterogeneity of porosity and permeability that transcends karst aquifers [25,26] and significantly affects many of their qualities, the quantification of recharge constitutes a challenging task. The APLIS method was developed by Andreo et al. [27] and later modified by Marin [28] for the estimation of the recharge rate, expressed as a percentage of precipitation. Its parametric methodology is based on a geographic information system (GIS) and was developed specifically for carbonate aquifers under Mediterranean climatic conditions. APLIS has been implemented as well as compared to several other methods by numerous researchers [29,30,31,32], while many modifications have been made since to accommodate varying circumstances [33,34,35].

In this research, we applied a two-step-based methodological approach. Initially, the efficiency of APLIS and its modified version were compared in the karst system of Ziria in Southern Greece. The quantitative comparison of the two methods showed that modified APLIS is more suitable for Greek karst systems. The next step was the application of modified APLIS at two additional karst sites, Planitero and ** on the surface. To apply the method, the prior variables were developed as information layers within a Geographic Information System (GIS). Scoring values, ranging from 1 (minimum) to 10 (maximum), were then assigned according to the influence on recharge. All the map layers that were developed for the application of APLIS and modified APLIS in Ziria are depicted in Figure 3 and Figure 4.

The altitude map (A) was derived from a Digital Elevation Model (DEM) of the country, which in turn originated the slope map (P) in percent rise as required by the method. The elevation in the study area ranges from 600 to 2368 m.a.s.l., and the altitude was classified in 6 sequences of 300 m intervals. The slopes span from 0%, mainly in the polje of Stymphalia, to over 100% on the steep inclinations of Mount Ziria. The scoring of the first two variables was attributed so that the higher the altitude, the greater the recharge due to the increase in precipitation, and the greater the slope, the lower the recharge into the aquifer [27] (

Table 1. Scoring of the Altitude (A) and Slope (P) variables for the karst system of Ziria.

Scoring	1	2	3	4	5	6	7	8	9	10
Altitude (m)	-	-	(600–900]	(900–1200]	(1200–1500]	(1500–1800]	(1800–2100]	(2100–2400]	-	-
Slope (%)
APLIS	>100	(76–100]	(46–76]	(31–46]	(21–31]	-	(16–21]	(8–16]	(3–8]	<3
Modified APLIS	>100	(65–100]	(45–65]	(30–45]	(20–30]	(15–20]	(10–15]	(5–10]	(3–5]	<3

). In both scenarios, the altitude map and its corresponding values remain consistent. However, the classification and scoring of the slopes differ in APLIS and its modified version, with the latter presenting an additional class.

Considering the lithological aspect (L), there are seven major units dominating the site that all received scoring based on imputes from previous studies [47,48] and field observations. All the carbonates received high scores due to their karstification degree and tectonic stress. The limestones of Tripolis are considered to be more karstified than the Cretaceous limestones of Pindos because the latter often intervene with thin layers of radiolarite [48], therefore they were attributed the values 10 and 9, respectively. The Quaternary formations exhibit fluctuations due to changes in grain size and composition, along with the presence of argillaceous materials. The alluvial deposits in the polje, along with the outcrop at the north-eastern part of the area, received lower values than the rest. Finally, the conglomerates were attributed a significantly higher score than what is proposed by APLIS since they are considered to be karstic [49].

The areas of preferential infiltration (I) were located with remote sensing techniques, using aerial photographs and stereo pairs in Erdas Imagine 2010 software that enabled the three-dimensional representation of the system. Predominantly, these landforms were observed on the limestones of the Mountain, where the dominance of bare rocks and the absence of vegetation facilitated the clear delineation of such topographic depressions. On the Cretaceous limestones of Pindos, fields of dolines can be visually distinguished through satellite images and photographs, with no further assistance from remote sensing software. These landforms were collectively grouped rather than individually digitized, occupying, in such a way, a significant section of the limestones. Regarding this variable, APLIS provides only two rankings: a score of 10, which was attributed to the preferential infiltration landforms, and a score of 1, which was assigned to the rest of the system. Karst landforms that are filled with Quaternary sediments were assigned the lowest value (1), due to the method’s limited scoring options. In modified APLIS, an additional intermediate value of 5 was introduced. It was received by the polje and other karstic features that are covered by various materials, as well as attributed to all the carbonate outcrops, in order to be distinguished from the non-carbonate formations.

Due to the lack of specific soil data, the evaluation of the soil types (S) was based on the combination of information provided by different Soil maps [50,51,52]. A substantial section of the Mountain is characterized by exposed rock, devoid of soil and vegetation. Leptosols, assigned the highest value of 10, were designated as the soil type for limestone areas exhibiting minimal to no vegetation. Chromic luvisols were attributed to formations with denser vegetation, while calcareous regosols and fluvisols were predominantly assigned to Quaternary sediments. There are minimum scoring differences between the two methods.

For the estimation of the correction factor in modified APLIS, permeable formations such as carbonates, conglomerates, and most of the Quaternary deposits were identified with aquifer potential and were assigned the correction coefficient (Fh) 1. Impermeable formations like the metamorphic series and argillaceous sediments were assigned the value 0.1.

The reliability of the used data for the application of the methods is provided in the Supplementary Materials.

The scorings that were attributed to the rest of the variables for the application of modified APLIS are displayed in

Table 2. Scoring values of the variables L, I, and S for the application of modified APLIS in Ziria.

Lithology (L)	Scoring	Infiltration Landforms (I)	Scoring
Argillaceous sediment	1	Rest of the area	1
Metamorphic bedrock	2	Carbonates/karstic landforms filled with sediment	5
Breccia	3	Doline/Swallow hole	10
Cone of Debris	4	Soil Type (S)	Scoring
Conglomerate	7	Chromic luvisols	3
Cretaceous Limestone (Pindos)	9	Calcareous regosols and fluvisols	6
Limestone/Dolomite (Tripoli)	10	Leptosols	10

. For the estimation of the annual recharge rate and the creation of maps with its spatial distribution in both scenarios, the expressions of the original (1) and modified method (2) were utilized in ArcGIS 10.8 software:

R = (A + P + 3 × L + 2 × I + S)/0.9

(1)

R = [(A + P + 3 × L + 2 × I + S)/0.9] × Fh.

(2)

2.2.2. Application of Modified APLIS to Planitero and ** the final outcome. Modified APLIS distinguishes the infiltration landforms in three categories (1, 5, and 10), in contrast to the initial method, which has only two categories (1 and 10). Consequently, the category with a score of 5 is more representative for many formations of the studied hydrosystem compared to the initial score of 1. The greater extent of the lowest recharge rates is due to the correction factor (Fh) that was introduced in the modified methodology and empowers users to provide a more precise permeability assessment for the geological formations. Inevitably, in the final map, extreme values of recharge rates (e.g., <20% and 60–80%) cover a greater area. Consequently, the mean annual value of the recharge rate estimated by the modified APLIS is lower than the initial method. The permeability factor has notably addressed the limitation of the original method, which could not estimate recharge rates below 8%. In non-permeable formations, APLIS estimated recharge rates between 16% and 33%, as it relied solely on the minimum attributed values of the lithological parameter. High elevations, flattened slopes, and thin soil coverage, however, can increase the recharge rate of an impermeable formation that receives high scores in all three categories.