بررسی هیدروژئولوژی و عدم‌قطعیت مدل آب زیرزمینی آبخوان دشت خرم‌آباد

Abstract
This study aims to evaluate the uncertainty associated with temperature and precipitation parameters influencing groundwater flow in the Khorramabad Plain aquifer. To achieve this objective, a Monte Carlo simulation was employed to investigate the effects of climatic variability on model performance. In this context, the MODFLOW groundwater flow model was utilized, and calibration and validation processes were meticulously conducted using a structured six-step approach. This methodology ensures the model’s compatibility with the hydrogeological conditions of the plain, thereby enhancing the reliability of the results. In light of climate change impacts on groundwater resources, temperature and precipitation projections were simulated based on emission scenarios from the Sixth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC), and groundwater level changes were forecasted for 20 years. The findings revealed that November exhibited the highest level of uncertainty in precipitation forecasts, with rainfall ranging from 0 to 262 mm and an estimated mean of 41.6 mm. Regional analysis indicated that Zone 5 is expected to experience the most significant fluctuations in groundwater levels, with a potential decline of up to 13.5 meters. The risk of aquifer depletion in this area is notably higher than in other regions. These results underscore the vulnerability of groundwater exploitation in Zone 5 and highlight the urgent need for precise and sustainable water resource management strategies. 
 
Extended Abstract
Background and Objective
Groundwater resources play a crucial role in meeting agricultural and domestic water needs, particularly in semi-arid and arid regions like the Khorramabad Plain, situated in western Iran. The intensifying overexploitation of aquifers, coupled with the unpredictable consequences of climate change, has raised critical concerns about the long-term viability of groundwater reserves in this region. Hydrological modeling serves as a robust analytical approach for simulating aquifer dynamics and guiding policy formulation; however, the intrinsic uncertainties of such models—arising from data limitations and variability in key parameters like precipitation and temperature—remain a significant challenge.
This study aims to quantitatively assess the uncertainties associated with groundwater level projections under future climatic conditions. To achieve this, the MODFLOW groundwater simulation model is integrated with a Monte Carlo-based uncertainty assessment framework utilizing the Generalized Likelihood Uncertainty Estimation (GLUE) methodology. Incorporating climate projections from the Coupled Model Intercomparison Project Phase 6 (CMIP6), this research endeavors to establish a scientifically grounded and reliable framework for strategic groundwater management in the Khorramabad aquifer system.
 
Methodology
The methodological framework of this study is centered on the integration of the MODFLOW groundwater simulation model and the Generalized Likelihood Uncertainty Estimation (GLUE) approach within the Groundwater Modeling System (GMS) environment. Initially, a three-dimensional groundwater flow model was constructed to simulate hydrological behavior over a decadal period (120 months), utilizing long-term observed hydrological and meteorological datasets. Model calibration was conducted using 75% of the recorded groundwater level data, while the remaining 25% was reserved for validation to ensure model robustness and predictive reliability.
To rigorously examine the impacts of climatic uncertainty on groundwater dynamics, a set of 810 synthetic climate scenarios was generated through a customized Python-based Monte Carlo simulation. This simulation stochastically varied three key climatic inputs—minimum temperature, maximum temperature, and precipitation—based on multi-model ensemble outputs derived from Phase 6 of the Coupled Model Intercomparison Project (CMIP6). These inputs were aligned with three distinct Shared Socioeconomic Pathways (SSPs), namely SSP1-2.6 (sustainability-focused), SSP2-4.5 (intermediate pathway), and SSP5-8.5 (fossil-fueled development).
The generated climatic variations served as dynamic forcings for the MODFLOW model, allowing the simulation of groundwater level responses under a wide spectrum of plausible future conditions. Subsequently, the GLUE methodology was applied to quantify uncertainty and characterize the probabilistic distribution of groundwater levels across scenarios. For spatially explicit analysis, the model domain was stratified into five hydrogeological zones, facilitating localized risk identification and enabling targeted groundwater management strategies.
 
Findings
The results of the uncertainty analysis indicated pronounced variability in groundwater level responses driven by fluctuations in climate input parameters. Among these, precipitation emerged as the most variable factor—particularly in November—exhibiting a wide range from 0 to 262 mm, with a monthly mean of 41.6 mm. In contrast, temperature variations (both minimum and maximum) were comparatively more consistent but still demonstrated notable seasonal dynamics that influenced groundwater recharge patterns.
Spatial analysis across the five delineated hydrogeological zones revealed Zone 5 as the most vulnerable to climatic variability. This zone exhibited water table depths ranging from 7.3 to 25.1 meters, with a 95% probability of substantial aquifer decline projected within the next two decades. The other zones showed relatively moderate or low sensitivity to climatic fluctuations, suggesting spatial heterogeneity in groundwater system resilience.
Among the climate models evaluated, the ACCESS-ESM1-5 simulation demonstrated the highest fidelity to historical climatic records, thereby being identified as the most suitable predictor for long-term groundwater projections in the study area. Notably, simulation runs that excluded anthropogenic groundwater withdrawals yielded relatively stable water level trends, underscoring the dominant role of human extraction in accelerating aquifer depletion and reinforcing the necessity of integrated demand management strategies in future water resource planning.
 
Conclusion
This research underscores the critical importance of incorporating uncertainty analysis into groundwater modeling, particularly in regions that are simultaneously exposed to climate change impacts and intensive anthropogenic water extraction. The integrated application of MODFLOW, the Generalized Likelihood Uncertainty Estimation (GLUE) method, and climate projections from Phase 6 of the Coupled Model Intercomparison Project (CMIP6) has enabled the development of a comprehensive and adaptable framework for assessing future groundwater dynamics in the Khorramabad Plain.
The analytical results identify Zone 5 as a high-priority area for targeted conservation measures, given its pronounced vulnerability to climatic fluctuations and declining aquifer levels. These findings underscore the necessity for policymakers to explicitly account for both climatic uncertainty and human-induced stressors in the formulation of groundwater management strategies.
Moreover, the methodological approach demonstrated in this study is readily transferable to other regions facing similar challenges, where uncertainties in input datasets and future climate trajectories complicate long-term planning. By emphasizing probabilistic forecasting rather than relying solely on deterministic outputs, the framework provides a more nuanced and realistic foundation for sustainable water resource management and policy development.