Recent Advances in Karstic Hydrogeology, 2nd Edition

Karst hydrogeology is the branch of hydrogeology that studies how groundwater flows and behaves in karst systems, characterized by the presence of soluble rocks (mainly limestone and dolomite) that favour chemical dissolution by meteoric water. This process creates distinctive karst landscapes with sinkholes, swallow holes, caves, conduits and large springs as preferential final outlets of the water flowpath through the underground system [1,2].

The key characteristic of a karst system is the high heterogeneity and anisotropy that allows water moving both slowly through rock matrix (pores/fractures) and rapidly through the conduit network. This leads to observe high groundwater flow velocities, similar to underground rivers, with high turbulence and non-laminar flows. The response to rainfall events is generally very fast, making karst systems preferentially vulnerable to pollutant transport due to low filtration capacity and conduit-dominated flow.

Karst areas spread over the 7–12% of the global continental area where their related aquifers often represent the most important and high quality freshwater resource for many different purposes (drinking, industrial, agriculture, etc.) [3].

In the last decades these have been a continuous challenge and source of inspiration for many studies carried on by researchers, field technicians and water managers.

Some of the most recent and current research trends in karst hydrogeology focused or are focusing on the following topics:

• flow modelling,
• monitoring and tracer techniques,
• machine learning and AI methods,
• sustainability and karst water management,
• contaminant transport and vulnerability,
• climate change effects on karst spring response,
• exploration and speleogenesis.

In the modelling, the development of hybrid hydrogeological models (dual porosity/ dual permeability) has opened a door to a deeper understanding of the subsurface flow and water transfer from the infiltration areas to the outlets [4]. The use of increasingly advanced software tools like MODFLOW-CFP and GIS-integrated tools to map conduit networks helped in representing groundwater flowing dynamics in karst. The major challenge is related to the difficulty in representing both diffuse and concentrated flow paths and calibrating models. All these models are evolving toward open-source transparency: openKARST is a good example of that, with a simulator that handles transitions from free-surface to pressurized flows, addressing flow regimes [5].

The monitoring still remains the real mandatory “password” when dealing with karst water resources. Using natural, artificial or any kind of tracer has been the key word (since the beginning of this discipline) to solve problems of hydrogeological nature through hydro-geochemistry, to formulate conceptual models, combining quantitative and qualitative analysis. The increasing diffusion of accurate field instrumentation allowed all over the world the use of fluorescent tracers, isotopes, microbiological indicators and sensors to continuously monitor parameters like electrical conductivity, temperature, and discharge [6]. Recently, also technologies as drones, LIDAR, and satellite imagery, born in a predominantly topographical context, have been used to map surface karst features. Remote sensing allows mapping surface collapse zones and dolines, while geophysics often helps to unveil subsurface conduit networks. Nowadays these methods are often merged with hydrologic data to obtain what is called “holistic aquifer modeling” [7].

Artificial intelligence (AI) and machine learning methods are gaining ground in every scientific field and also in the daily life of all of us. This is mainly thanks to the huge amount of data that we have recorded and store over the last 20 years. In the field of karst water management, the most far-sighted water utilities and managers have been able to obtain at a low cost historical series of data (ranging from water quality, to piezometric levels and spring flow rates). These are now proving to be very useful for training neural network models capable of simulating and, in some cases, even predicting with excellent accuracy many of the hydrological phenomena that the same physical models struggle to represent and understand [8,9]. Also in defining and simulating the conduit network AI is getting a leading gradually taking the place previously occupied by stochastic methods [10,11]

In sustainable water resource management the focus is mainly directed on karst aquifers as primary water sources in many regions (e.g., Mediterranean, Balkans) [12,13]. Many studies are integrating several different approaches and disciplines to protect aquifers both from overexploitation and contamination. This latter is always the major threat, especially for drinking purposes. The analysis of intrinsic aquifer vulnerability using historical methods has been integrated with new methods or updating the most famous ones like COP, EPIK, PI, DRASTIC, adapting to new constraints or specific areas [14]. After COVID-19 the research is now also focusing on new and rapid contaminant transport mechanisms (involving nitrates, microplastics, pesticides, viruses).

Climate change and related extreme events are increasing vulnerability drivers for karst water resources. Their short and long-term effects on evapotranspiration, recharge, droughts, and floods need further studies involving climate projections. In this sense, several interdisciplinary and international projects such as COST Actions or Horizon Europe are trying to solve these issues and raise awareness among people [15]. Studies at snow-influenced central European sites combine hydrograph analytics with lumped-parameter models to assess future discharge under IPCC scenarios, demonstrating robust, site-specific responses to global warming [16].

Last but not least the exploration and speleogenesis are research fields full of novelties and potential future implications on karst management [17,18,19]. The conduit formation mechanisms and hypogenic speleogenesis (conduits formed by ascending CO₂- or H₂S-rich fluids) require deeper investigations and new young willing researchers to fill the gaps currently existing.

Hence, karst hydrogeology, long defined by its inherent complexity, is undergoing a methodological transformation worldwide. The convergence of open-source hydrological codes, AI-driven network synthesis, advanced tracer analytics, integrated remote sensing, and climate-responsive monitoring frameworks is equipping scientists and water managers with an unprecedented toolkit [20,21].

However, several challenges are still there and new ones are now matter of research work conducted by karst scientists (Table 1). The most important ones include:

• Scalable modeling: Can we couple high-resolution site models with regional or even continental frameworks?
• Data equity and integration: How can tracer data, remote sensing outputs, and climate projections be fused in a new holistic approach?
• Operational vulnerability indices: Can we create adaptive tools for real-time risk assessment under dynamic climate conditions?

Multidisciplinary collaborations (fusing hydrology, computer science, geochemistry, geophysics, and remote sensing) are becoming essential to solve issues in a comprehensive way. Equally, investments in sustained monitoring across diverse karst systems and standardized tracer protocols will ensure models to be robust, methods to be transferable so that water managers can anticipate and mitigate risks in a changing climate scenario.

Table 1. Synthesis & Future Research Directions.

Challenge	Emerging Solution	Research Gap
Flow & transport heterogeneity	Hybrid models + Open-source tools (openKARST)	Integration of real-time data & AI-enhanced calibration
Network geometry uncertainty	Graph-based generative modeling	Need for larger, diversified network datasets
Tracer-based vulnerability insights	Environmental tracers + process models	High-resolution seasonal vulnerability assessments
Climate-impact quantification	Long-term spring monitoring + scenario modeling	Scaling from site-specific to regional/global assessments
Surface-subsurface mapping	UAV, LiDAR, GPR	Cost-effective integration into routine monitoring

Table 1. Synthesis & Future Research Directions.

Challenge	Emerging Solution	Research Gap
Flow & transport heterogeneity	Hybrid models + Open-source tools (openKARST)	Integration of real-time data & AI-enhanced calibration
Network geometry uncertainty	Graph-based generative modeling	Need for larger, diversified network datasets
Tracer-based vulnerability insights	Environmental tracers + process models	High-resolution seasonal vulnerability assessments
Climate-impact quantification	Long-term spring monitoring + scenario modeling	Scaling from site-specific to regional/global assessments
Surface-subsurface mapping	UAV, LiDAR, GPR	Cost-effective integration into routine monitoring

This Editorial refers to the Special Issue “Recent Advances in Karstic Hydrogeology, 2nd Edition”. Following the successful first volume of the Special Issue “Recent Advances in Karstic Hydrogeology”, the second version of this Special Issue have collected the most recent and advanced research studies on this topic to overcome issues related to karst water resources. The Special Issue highlighted new opportunities and challenges with seven papers accepted for publication included (six articles and one protocol).

Contribution 1 investigates how the geometry of karst conduits affects the dispersion and transport dynamics of artificial tracers, with a focus on the often-observed dual-peaked breakthrough curves (BTCs). The goal is to systematically link conduit design parameters to tracer responses. CFD simulations have been conducted using COMSOL Multiphysics 6.1: Simulations model tracer transport under turbulent flow conditions in synthetic conduit networks.

The study highlighted the importance of the conduit network topology in tracer test interpretation. BTCs alone may be misleading unless geometry is properly accounted for, aligning with broader karst hydrology findings. Geometrical complexity (pools, branching) strongly affects mixing, storage zones, and residence time distributions.

Contribution 2 focuses on a specific study area: the Sete Lagoas, Minas Gerais, in Brazil, where extensive groundwater pumping from a karst aquifer has caused several issues (drying lakes, land subsidence, sinkholes, and urban instability). Authors used a 3D geological model covering about 146 km² of Precambrian carbonates overlain by unconsolidated sediments. They developed and calibrated a numerical groundwater flow model integrating historical hydraulic head and flow measurements for more than 250 monitoring points, hydrostratigraphic properties for the epikarst, karst zones, and limestone matrix. By the 1980s, intensive pumping generated a ~30 km² cone of depression and between 1940 and 2020, 20 collapse events occurred within zones where limestone outcrops intersected this cone. Using this model authors found that a reduction of extraction by 40% is necessary to stabilize or reverse land subsidence risk.

Contribution 3 described the use of a fully connected neural network (FCNN) to simulate daily discharge patterns of six karst springs in Umbria Region (Central Italy). Authors used 20 years of daily rainfall and spring discharge data and the proposed model performed two sequential steps: gap-filling for missing discharge data and simulating spring responses to rainfall inputs. Cross-correlation analysis was applied to understand the influence of rainfall on karst spring discharge patterns whereas the FCNN demonstrated robust simulation capability, successfully modelling complex and different karst hydrological systems with limited process-based data. The study offers a pragmatic tool for water managers to estimate spring discharge using readily available rainfall records.

Contribution 4 analyze how lakes within the Neoproterozoic karst landscape of Lagoa Santa (central Minas Gerais, Brazil) connect hydrologically to different aquifer systems (karst conduit, fissure-karst, and porous) by assessing 129 lakes over 36 years (1984–2020). To do so a time-series analysis of satellite imagery (Landsat) and annual rainfall from 1984 to 2020 has been conducted by the authors. They used a supervised classification (Maximum Likelihood in QGIS) to measure seasonal and long-term changes in lake perimeters, identifying contractions >5% as significant. Different statistical trend tests (Pettitt, Mann–Kendall) on rainfall series have been finally evaluated to identify dry/wet periods and transition periods. Authors found that studied lakes act as dynamic interface zones, switching roles seasonally or under climatic stress. This result helps identifying aquifer-lake connectivity in the study area, but also proves to be useful for resource management in semi-arid karst contexts worldwide.

Contribution 5 is a study protocol which is the main result of the DY.MI.CR.ON initiative (funded via PRIN 2022/Next Generation EU). This project integrates geology, hydrogeology, microbiology, hygiene, and agronomy to develop an interdisciplinary framework for understanding, monitoring, and mitigating microbial contamination of groundwater in the Earth’s critical zone. The study areas are in Apulia (Carpignano Salentino, Lecce) and Sicily (Scicli, Pozzallo) in Southern Italy. Authors conducted core-flow experiments (to simulate microbial transport), mathematical modeling, molecular and culture-based detection and Quantitative Microbial Risk Assessment (QMRA). Outputs will guide monitoring protocols, refine public health policies, and improve agricultural safety standards regarding irrigation practices. The project is still going on and the core-scale experiments simulating unsaturated flow are extended through innovative modeling and complemented by field observations.

Contribution 6 evaluated the availability, water quality, and vulnerability of the Jebel Zaghouan karst aquifer (historically vital for Carthage and Tunis) under the stresses of overexploitation and climate change. Authors used APLIS and COP methods, supported by remote sensing and GIS, to estimate infiltration rates and creating vulnerability maps. They analyzed droughts and climate change using Standardized Precipitation Evapotranspiration Index (SPEI) and Standardized Groundwater Index (SGI). The water quality has been monitored tracking major ions in time. The most important findings are that the aquifer is chronically overexploited, with insufficient recharge to offset withdrawals. This fact is posing a serious risk to long-term sustainability of groundwater in the study area. The results of COP classified the area predominantly as moderately vulnerable (about 80%), indicating susceptibility to contamination which mainly consist of nitrates peaks in springs, especially during wet years, highlight that the risk is coming from agriculture activities.

Contribution 7 investigated the geochemistry and mineralogy of two deep cave environments in the Croatian Dinaric karst, specifically the Slovačka jama and the Velebita cave systems. It connects cave mineralogy with environmental and paleoenvironmental implications in deep subsurface settings. Authors used several methods such as XRD (mineral identification), elemental chemistry (ICP-MS), magnetic susceptibility (MS) measurements and factor analysis for grouping elemental patterns. As expected calcium is predominant, but clay-rich samples, especially from deep sedimentary zones, exhibited elevated magnetic susceptibility tied to Fe–Mn enrichment. Heavy-metal anomalies have been detected at depths of 300–400 m in Slovačka jama system. Major findings coming from this study are that deep cave geochemistry is shaped by a mix of geological sources, atmospheric deposition and biogenic and hydrothermal contributions.

Speleothems and sediments serve as archives, capturing both natural processes and subtle anthropogenic inputs, making them valuable for paleoenvironmental reconstruction. MS emerges as a fast, cost-effective tool for detecting subsurface trace element hotspots in karst systems.

The contributions’ references are listed in temporal order below:

• Rabah, A.; Marcoux, M.; Labat, D. Effects of Geometry on Artificial Tracer Dispersion in Synthetic Karst Conduit Networks. Water 2023, 15, 3885. https://doi.org/10.3390/w15223885.
• Galvão, P.; Schuch, C.; Pereira, S.; de Oliveira, J.M.; Assunção, P.; Conicelli, B.; Halihan, T.; de Paula, R. Modeling the Impact of Groundwater Pumping on Karst Geotechnical Risks in Sete Lagoas (MG), Brazil. Water 2024, 16, 1975. https://doi.org/10.3390/w16141975.
• De Filippi, F.M.; Ginesi, M.; Sappa, G. A Fully Connected Neural Network (FCNN) Model to Simulate Karst Spring Flowrates in the Umbria Region (Central Italy). Water 2024, 16, 2580. https://doi.org/10.3390/w16182580.
• Pacheco Neto, W.; de Paula, R.; Galvão, P. Karst Hydrological Connections of Lakes and Neoproterozoic Hydrogeological System between the Years 1985–2020, Lagoa Santa—Minas Gerais, Brazil. Water 2024, 16, 2591. https://doi.org/10.3390/w16182591.
• Verani, M.; De Giglio, O.; Caputo, M.C.; Cassiani, G.; Milani, M.; Carducci, A.; Federigi, I.; Pagani, A.; Angori, A.; Triggiano, F.; et al. Study Protocol of Predictive Dynamics of Microbiological Contamination of Groundwater in the Earth Critical Zone and Impact on Human Health (DY.MI.CR.ON Project). Water 2025, 17, 294. https://doi.org/10.3390/w17030294.
• Gargouri-Ellouze, E.; Slama, F.; Kriaa, S.; Benhmid, A.; Taupin, J.-D.; Bouhlila, R. Comprehensive Assessment of the Jebel Zaghouan Karst Aquifer (Northeastern Tunisia): Availability, Quality, and Vulnerability, in the Context of Overexploitation and Global Change. Water 2025, 17, 407. https://doi.org/10.3390/w17030407.
• Paar, D.; Frančišković-Bilinski, S.; Buzjak, N.; Maldini, K. New Insight into Geochemistry and Mineralogy of Deep Caves in Croatian Karst and Its Implications for Environmental Impacts. Water 2025, 17, 1001. https://doi.org/10.3390/w17071001.

In conclusion, this special issue has been characterized by very different articles, which have addressed most of the topics currently at the forefront of the field of karst hydrology and presented at the beginning of this editorial. In particular, the modelling of the conduit network and its influence on tracer tests (1), the overexploitation of the karst aquifer and its consequences also in the urban context (2), the use of innovative methods such as artificial intelligence and machine learning on long-term datasets for the simulation of spring flow and gap filling (3), the use of remote sensing for the identification and classification of karst landforms and the interaction between lakes and the karst system (4), the study and definition of protocols for microbiological characterization and transport in the saturated and unsaturated zone in karst environments (5), the holistic approach to the assessment of karst aquifers in semi-arid environments considering climate change, vulnerability and water quality (6) and finally the mineralogical, geochemical and speleological study to distinguish natural processes and anthropogenic inputs for paleoenvironmental reconstruction. This demonstrates the success that the special issue “Recent Advances on Karstic Hydrogeology, 2nd edition” has achieved, highlighting how the topic of karst hydrogeology is of considerable interest to the scientific community and the world as a whole.

The special issue, together with the broader literature, underscore a shift in karst hydrogeology toward:

• Multidisciplinary methods combining field data, remote sensing, geochemistry and geophysics
• AI and ML tools for flow pattern recognition, vulnerability classification, and system resilience forecasting
• Ecosystem-level evaluation, particularly focused on microbiology in Mediterranean and coastal contexts.
• Upscaling frameworks that link local hydrological understanding with regional and global resource management.

This synergy places the field to better anticipate climate impacts, manage water quality, and protect ecosystems in karst landscapes.