Abstract
This study develops a decision-oriented groundwater modeling framework for the Marvdasht-Kharameh aquifer (Fars Province, Iran) to diagnose long-term drawdown and salinity exposure. A 127-month hydroclimatic and groundwater-level record (Oct-2010-Mar-2021) underpins a transient MODFLOW model coupled with MODPATH particle tracking. The calibrated simulation (final RMSE ≈ 2.59 m) reproduces a basin-mean decline of ≈ 7.79 m and highlights eastern–southeastern hotspots where pumping concentrates. Advective pathways connect wetlands/surface waters to vulnerable cells; observed chloride increases are consistent with transport along high-conductivity corridors. Sensitivity analysis elevates horizontal hydraulic conductivity and storage as first-order controls, suggesting targeted abstraction curbs, managed aquifer recharge at high-decline sites, and operational oversight of river/wetland stages.
Keywords: Groundwater Resources; MODFLOW; MODPATH; Water Resources Management; Climate Change; Salinity Intrusion; Hydraulic Conductivity.
1- Introduction
Groundwater is the principal and, in many districts, the only reliable source for domestic and agricultural supply across Iran’s arid and semi-arid plains. Persistent over-abstraction and quality degradation-particularly salinization driven by surface-water–groundwater interactions-threaten this security. While numerous studies document declining heads, fewer provide an integrated, decision-relevant quantification of (i) decadal dynamics under observed stresses, (ii) the physical pathways by which salinity reaches production zones, and (iii) which parameters and processes most strongly condition management outcomes.
This study addresses that gap for the Marvdasht–Kharameh aquifer (Figure 1) by combining transient flow modeling and particle tracking with targeted sensitivity analysis. The objectives are to (1) reconstruct spatiotemporal drawdown over 2010–2021, (2) delineate advective salinity pathways from adjacent rivers/wetlands, and (3) identify leverage points for policy and operations.
2-Materials and Methods
Data and study window: This study assembled 127 monthly observations of groundwater levels alongside hydroclimatic inputs and hydrogeologic characterization. The archive exhibits a multi-year decline with regular seasonal oscillations (Table 1).
Numerical model: Transient flow was simulated using MODFLOW on a finite-difference grid (representative cell size on the order of a few hundred meters), with monthly stress periods, spatially variable recharge, and head-dependent boundaries representing river/wetland connectivity.
Calibration and validation: A split record was used for calibration/validation against observation wells. Performance was assessed via RMSE, MAE, and mean error; the final transient fit achieved RMSE ≈ 2.59 m. Calibrated statistics indicate horizontally heterogeneous conductivity and low storage, consistent with alluvial media.
Transport diagnostics: MODPATH particle tracking delineated advective pathways from identified saline surface sources into the aquifer to map exposure zones and likely travel directions under the calibrated flow field.
Sensitivity analysis: Structured perturbations to horizontal hydraulic conductivity, anisotropy, recharge, river–aquifer exchange, and storage were used to rank controls on heads and on particle-based exposure metrics.

Figure 1. Geographic location of the Marvdasht–Kharameh Aquifer study area

Table 1. Optimized Hydrogeological Parameter Values Across Calibration Stages for the Marvdasht–Kharameh Aquifer (MODFLOW–MODPATH)
Parameter Min Max Mean Median Standard deviation Unit
Horizontal hydraulic conductivity (Kx) 0.0376 139.9163 27.7992 12.5539 32.7247 m/day
Vertical hydraulic conductivity (Kz) 0.0151 98.2843 14.2547 9.0636 19.9783 m/day
Horizontal hydraulic anisotropy (Anisotropy) 0.0533 16.3287 1.7024 1.1892 1.9784 –
Storage coefficient (S) 0.0001 0.0053 0.0008 0.0005 0.0014 –
Transmissivity (T) 0.02 25.6832 5.7982 3.5097 7.8634 m²/day
Recharge (R) 0.0011 0.0452 0.0154 0.0102 0.0138 m/day
Vertical hydraulic anisotropy (Kz/Kx) 0.05 3.57 0.98 0.78 0.75 –

3- Results and Discussion
Decadal dynamics and spatial structure: The model reproduces the observed downward trend, yielding a basin-mean decline of ≈ 7.79 m over 127 months. Drawdown concentrates in the east and southeast, where abstraction density is high and transmissivity troughs sharpen gradients. Hydrographs show deeper summer depressions consistent with peak demand and evaporative losses (Figure 2).
Salinity exposure pathways: Particle tracking reveals hydraulic connectivity from wetlands/surface waters to production areas through high-conductivity corridors. Field-informed chloride increases-from ~70 to ~110 mg L⁻¹ in a southeastern focus area and ~50 to ~75 mg L⁻¹ in a central tract-corroborate the modeled exposure pattern, indicating that advective transport, rather than solely local evapoconcentration, is a credible driver of quality risk.
Process controls and management leverage: Sensitivity diagnostics consistently rank horizontal hydraulic conductivity and storage as first-order controls. Higher conductivity steepens local cones near pumping centers and accelerates advective travel from saline sources; higher storage damps seasonal oscillations and mitigates cumulative decline. River–aquifer exchange terms are influential along boundary reaches, underscoring that surface-water operations co-determine groundwater outcomes.
Implications: Three levers emerge: (i) targeted abstraction curbs in high-impact sub-basins; (ii) siting managed aquifer recharge (MAR) where modeled declines are greatest and infiltration capacity is adequate; and (iii) operational oversight of river/wetland stages and continuous salinity monitoring to interrupt pathways before chloride fronts intersect supply wells.

Figure 2. MODPATH-simulated salinity migration pathways and hydraulic connectivity to surface saline sources in the Marvdasht–Kharameh aquifer
4- Conclusion
An integrated MODFLOW–MODPATH framework, calibrated to a 127-month record, quantifies a substantial basin-mean decline (≈ 7.79 m) and delineates advective salinity pathways linking surface waters to vulnerable production zones. Horizontal hydraulic conductivity, storage, and river-aquifer exchange dominate system response, suggesting actionable interventions: redistribute and reduce pumping in identified hotspots, deploy MAR where it is hydraulically effective, and manage/monitor surface-water stages and salinity as part of a coupled operations plan. The framework is transferable to other arid and semi-arid aquifers facing coupled quantity-quality stresses and provides a defensible basis for adaptive management.