This paper aimed to assess land subsidence in the Marand Plain using Interferometric Synthetic Aperture Radar (InSAR) and to investigate the factors influencing its occurrence. Processing of radar images over the period 2020–2024 revealed that subsidence rates in the plain ranged from zero to a maximum of 3.745 cm over four years, with an average of approximately −1.723 cm. The spatial distribution of subsidence follows a heterogeneous but continuous pattern, with more than 70% of the plain’s area falling within moderate to high subsidence classes. Analysis of the driving factors indicated that groundwater level decline is the primary trigger of subsidence, reaching over 10 meters in the southeastern and western parts of the plain. Additionally, aquifer characteristics such as considerable thickness (up to 282 m), dominance of fine-grained clay and silt sediments (especially in the western half), reduced natural aquifer recharge (due to climatic and hydrological droughts), gentle terrain slopes, and human activities such as irrigated agriculture and land-use change, are other factors exacerbating subsidence in the region. Accuracy assessment of the method using the ROC curve, Area Under the Curve (AUC = 0.949), and Gini coefficient (GC = 0.898) indicates excellent model performance. The findings highlight the critical condition of land subsidence in the Marand Plain, particularly in its western half, emphasizing the urgent need for management strategies such as controlling groundwater extraction, enhancing artificial aquifer recharge, revising cropping patterns, and implementing efficient irrigation practices.
Introduction
Land subsidence is defined as the downward vertical displacement of the ground surface, which typically occurs as a result of susceptible geological conditions and under the influence of various environmental, hydrogeological, and economic changes and factors (Batubara et al., 2023). This phenomenon may have natural or anthropogenic origins and is consistently associated with significant concerns due to the reduction in aquifer storage capacity (Jiang et al., 2023). As one of the major environmental and engineering challenges in arid and semi-arid regions of the world, land subsidence mainly results from declining groundwater levels caused by excessive extraction, climate change, and geological factors. Iran, due to its heavy reliance on groundwater resources and exposure to prolonged droughts, has been increasingly affected by this phenomenon. The Marand Plain, as one of the important and densely populated plains of East Azerbaijan Province, is at risk of land subsidence due to excessive groundwater withdrawal, reduced natural recharge of the aquifer, and susceptible geological characteristics. This study aims to analyze the mechanisms and identify the factors influencing land subsidence in the Marand Plain using an integrated approach.
Material and Methods
This paper was conducted using an integrated and multidisciplinary methodological framework. The core of the methodology was based on Interferometric Synthetic Aperture Radar (InSAR) technology. For this purpose, a series of Sentinel-1 radar satellite images covering a four-year period from 2020 to 2024, with a four-month revisit interval, were utilized. Image processing was carried out in a specialized software environment and involved several steps: (1) generation of interferometric pairs through precise geometric registration and co-registration of master and slave images; (2) removal of topographic phase effects using a high-resolution Digital Elevation Model (DEM); (3) correction of atmospheric delay using atmospheric model data; (4) phase filtering to reduce noise and enhance the signal-to-noise ratio; (5) phase unwrapping; and finally, (6) conversion of relative satellite line-of-sight (LOS) displacement into absolute vertical ground surface displacement. The output of this process consisted of annual and cumulative land subsidence rate maps with millimeter-level accuracy.
To complement and validate the satellite-based analyses, a Geographic Information System (GIS) was employed as an integrated platform for data integration, spatial analysis, and visualization of various information layers. Hydrogeological data, including groundwater table levels from piezometric wells over a ten-year period, groundwater contour maps and water-level change maps, aquifer thickness layers, and quantitative sedimentary characteristics of the aquifer environment, were imported into and processed within the GIS environment. In addition, field observations using a three-frequency GPS device were conducted at control points across the plain to validate the InSAR results. The accuracy of the InSAR method was quantitatively evaluated using the Receiver Operating Characteristic (ROC) curve, calculation of the Area Under the Curve (AUC), and the Gini coefficient.
Results and Discussion
The results of satellite monitoring showed that land subsidence has extensively affected the entire Marand Plain, with varying degrees of intensity. Over the four-year period, subsidence rates ranged from zero to a maximum of −3.745 cm over four years, with an average subsidence of approximately −1.723 cm for the entire plain. Subsidence zonation analysis indicated that only about 9.5% of the plain falls within the very low subsidence class, while the moderate, high, and very high classes together cover nearly 70% of the plain’s total area. The most high-risk zones are concentrated in the western half and parts of the southeastern sector of the plain, and major cities such as Marand, Koshksaray, and New Bonab are located within areas experiencing moderate to high subsidence rates.
_{The main driving factor of land subsidence was identified as the severe and continuous decline in groundwater levels. The average groundwater table drop over the past decade was −4.67 m, reaching more than 10 m in critical hotspots in the southeastern and western parts of the plain. The significant spatial overlap between areas with the greatest groundwater decline and zones with the highest subsidence rates confirms a direct causal relationship between these two phenomena. In addition, intrinsic aquifer characteristics - including considerable thickness (up to 282 m) and fine-grained sediments (especially clay and silt layers) - together with reduced natural aquifer recharge and human activities such as the expansion of water-intensive irrigated agriculture, have contributed to the intensification of subsidence in the western half of the plain.}
Conclusion
_{This study demonstrates that the Marand Plain particularly its western and southeastern areas—is in a critical condition with respect to land subsidence. The continuation of current groundwater exploitation trends and the lack of effective management intervention could lead to irreversible damage to the environment, infrastructure, and human communities. Accordingly, immediate and comprehensive action is required, including integrated and sustainable groundwater resource management, implementation of artificial aquifer recharge programs, fundamental revision of agricultural and irrigation patterns, incorporation of hazard considerations into development planning, and continuous monitoring of subsidence-prone areas. These measures are proposed as key strategies to mitigate this crisis and move toward environmental resilience in the Marand Plain. The methodology employed in this research, with an AUC value of 0.949 and acceptable agreement with field observations, confirms the high reliability of the results.}