Original Research Paper
Satellite Technology Eengineering
M. Nasiri Sarvi; M. Jalilian
Abstract
Background and Objectives: Optical navigation for spacecraft in low Earth orbit is increasingly valued as a primary or backup solution when radio-frequency positioning becomes unreliable due to interference, intermittent coverage, or mission constraints. This study addresses the need for a robust, low-overhead ...
Read More
Background and Objectives: Optical navigation for spacecraft in low Earth orbit is increasingly valued as a primary or backup solution when radio-frequency positioning becomes unreliable due to interference, intermittent coverage, or mission constraints. This study addresses the need for a robust, low-overhead processing chain that can estimate both state and position using only onboard imaging sensors. The approach intentionally combines two complementary sources of information: a star sensor, which stabilizes attitude estimation and improves the linearization needed for filtering, and an Earth-horizon sensor, which imposes a strong geometric constraint along the radial direction of the orbit. The overarching objective is to design and validate a non-synchronous fusion architecture that produces accurate and temporally well-behaved estimates without relying on external radio navigation. Specifically, the study aims to: develop a geometry-aware weighting strategy aligned with the radial, along-track, and cross-track frame; enforce principled statistical گیتینگ (ناحیه بندی) to ensure measurement quality; and apply selective post-processing to reduce short-period fluctuations in the along-track and cross-track directions while preserving the radial constraint provided by the Earth-horizon sensor. The intended outcome is a practical chain suitable for small satellites and missions with limited computational resources, and for operational contexts that are sensitive to transient estimation oscillations.Methods: The investigation is performed in a high-fidelity simulation of a representative low Earth orbit with truth data generated by an orbital propagator and environmental models suitable for that regime. Two complementary measurement models are employed. The star sensor provides direction vectors that primarily stabilize the attitude solution and the associated linearization of the dynamics. The Earth-horizon sensor provides limb observations that yield a strong constraint on radial position. Because the sensors operate at different update rates, fusion is event-driven: measurement updates are processed whenever new data arrive, while state predictions evolve continuously according to the orbital dynamics and disturbance models. Temporal alignment across the two streams is handled through state interpolation. Measurement quality is controlled by statistical گیتینگ (ناحیه بندی) based on the Mahalanobis distance to reject outliers without discarding informative data. To respect the physics of the orbit geometry, an elliptical weighting scheme is formulated in the radial, along-track, and cross-track frame so that information is emphasized where each sensor is most informative. After filtering, a Rauch–Tung–Striebel smoother is applied selectively to the along-track and cross-track components, leaving the radial estimate unchanged to avoid weakening the Earth-horizon constraint. Performance is evaluated across multiple noise regimes and viewing conditions. Error behavior is characterized using root-mean-square and mean-absolute measures, together with time-domain analyses of innovations, acceptance rates for measurement updates, and qualitative inspection of position-component traces in the orbital frame.Findings: The non-synchronous fusion of star and Earth-horizon measurements yields a clear and consistent reduction in overall position error relative to a filter-only baseline. The radial component exhibits rapid convergence and remains tightly constrained throughout, reflecting the strong geometric information provided by the Earth-horizon sensor. The selective post-processing smooths the along-track and cross-track components, attenuating short-period oscillations without introducing noticeable bias or drift, and doing so while intentionally leaving the radial component unaffected. Analyses of the innovations and the evolution of their orientation confirm that temporal alignment is effective and that the geometry-aware weighting is well tuned. Acceptance rates for measurement updates remain steady across scenarios, indicating that statistical گیتینگ (ناحیه بندی) is neither overly permissive nor excessively conservative. Under more challenging noise conditions, the chain maintains stable behavior: convergence persists, error growth is bounded, and the largest variability continues to appear in the along-track and cross-track directions, where the smoother delivers the most visible benefit. Visual inspection of the component-wise time histories corroborates these conclusions, showing consistent damping of fluctuations in the orbital plane and a preserved, physically plausible trajectory along the radial direction. Computational demands remain modest, supporting deployment on resource-limited platforms.Conclusion: The proposed fusion chain—built on complementary sensors, event-driven filtering, geometry-aware weighting, principled گیتینگ (ناحیه بندی), and selective post-processing—offers a practical and deployable framework for optical navigation in low Earth orbit. Beyond improving accuracy, the method delivers temporally orderly estimates that are well suited to threshold-based decision making in flight operations. The approach is especially relevant for small satellites and missions that must tolerate intermittent or degraded radio-frequency positioning. Nevertheless, several factors warrant further investigation before routine operational use. Because the present results are obtained in simulation, progression to ground and hardware-in-the-loop testing with real sensors is essential, together with careful calibration of installation matrices. Sensitivity to Earth-horizon sensor bias, scattered light, extended eclipse periods, and short maneuver segments should be quantified. From an estimation standpoint, exploring alternative variants of the Kalman filter family, stronger constraints on the radial channel during smoothing, higher-fidelity environmental and dynamical models, and more precise handling of non-synchronous timing may yield additional gains. Taken together, these directions chart a clear path toward an operational, image-only navigation capability that can function independently of external radio infrastructure while meeting the accuracy and stability expectations of contemporary space missions.
Original Research Paper
Geodesy
L. Zamany; Y. Djamour; A. Milan
Abstract
Background and Objectives: The phenomenon of ground surface displacement—particularly subsidence and horizontal movement—is one of the most significant environmental and geotechnical threats in many regions of the world, including Iran. Therefore, accurate monitoring and measurement of these ...
Read More
Background and Objectives: The phenomenon of ground surface displacement—particularly subsidence and horizontal movement—is one of the most significant environmental and geotechnical threats in many regions of the world, including Iran. Therefore, accurate monitoring and measurement of these displacements are of great importance. In recent years, the application of remote sensing technologies such as radar interferometry (InSAR) and offset tracking has made remarkable progress. Additionally, optical imagery and UAV (drone) data have enabled the reconstruction of 3D models and the extraction of surface deformation patterns. However, most previous studies have utilized only one of these techniques, and few have focused on integrating radar and optical data. Accordingly, the main objective of this study is to conduct a comprehensive investigation of ground displacement fields through the integration of radar and optical imagery.Methods: Five villages-Bolbol, Joko, Tazehkand Nasirpour, Bali Qaya, and Lalehzar, located in the southeastern region of Maragheh County, were selected as study areas to allow the assessment of displacement behavior across different land types (residential/urban and flat/rural areas). Ground displacement was analyzed using two distinct approaches: vertical displacement (subsidence/uplift) and horizontal displacement. The analyses were based on average deformation values within the two major land categories. Three complementary techniques were employed: radar interferometry (InSAR), offset tracking, and optical photogrammetry using 3D model reconstruction. In the interferometry technique, seven Sentinel-1 radar images with approximately three-month intervals were collected and processed. Similarly, seven radar images were processed for offset tracking, resulting in a total of twelve radar-based analyses. Furthermore, four series of drone images (Mavic 3) were captured and processed at approximately four-month intervals.Findings: In the vertical displacement analysis, the greatest subsidence occurred in Lalehzar Village, where ground elevation decreased by approximately 13–15 cm in residential/urban areas and 12–14 cm in flat terrains. This level of subsidence is likely related to excessive groundwater extraction and the unique geological conditions of the region. Conversely, the lowest subsidence was recorded in Bali Qaya Village, with changes of about 2–3 cm in residential areas and around 7 cm in flat regions. Regarding horizontal displacement, the highest value was observed in Bolbol Village, with movements of about 5–6 mm in flat terrains and 1.8–2.1 mm in residential areas. The lowest horizontal displacement was found in Bali Qaya Village, where horizontal shifts were less than 1.5 mm in residential areas and about 4 mm in flat terrains.Conclusion: Overall comparison between the methods indicates that, although the displacement trends observed by both approaches are nearly identical—and both successfully identify the same regions with maximum and minimum displacement—the numerical values differ slightly in some cases. These discrepancies can be attributed to each method’s sensitivity to different displacement types (vertical or horizontal), imaging conditions, and processing algorithms. In general, the combined use of radar and optical approaches provides a more comprehensive understanding of surface deformation patterns and significantly enhances the accuracy of decision-making in resource management, subsidence monitoring, and urban planning.
Original Research Paper
Remote Sensing
M. Hasanlou; Z. Roodsarabi; P. Hasan Teymori
Abstract
Background and Objectives: Large-scale wildfires, through the destruction of vegetation, increased soil instability, and disruption of ecosystem functioning, have become one of the most serious environmental challenges of the modern era. Accurate post-fire burn-area delineation is essential for damage ...
Read More
Background and Objectives: Large-scale wildfires, through the destruction of vegetation, increased soil instability, and disruption of ecosystem functioning, have become one of the most serious environmental challenges of the modern era. Accurate post-fire burn-area delineation is essential for damage assessment, restoration planning, and risk management. Satellite optical data—particularly Sentinel-2 imagery—combined with widely used spectral indices provide a powerful basis for mapping burned areas; however, their performance depends strongly on the choice of indices and classification models. The objective of this study is to evaluate and compare the effectiveness of a classical statistical classifier, three machine learning algorithms, and two deep learning architectures for burned-area detection, using a combination of Sentinel-2 spectral bands and spectral indices in the Kenneth wildfire in Los Angeles.Methods: Following selection of the post-fire Sentinel-2 imagery and cloud masking, eight core spectral bands (visible, near-infrared, red-edge, and shortwave infrared) along with five commonly used indices related to burn severity, vegetation condition, and moisture content were extracted, forming a 15-variable input image for model development. Binary reference labels (burned/unburned) were derived from the official wildfire incident database, and spatially random sampling was used to create training (70%) and testing (30%) subsets. All features were normalized using min–max scaling. Subsequently, a classical Maximum Likelihood Estimation (MLE) classifier, three machine learning algorithms—Adaptive Boosting (AdaBoost), Random Forest (RF), and Support Vector Machine (SVM)—and two deep learning models—Convolutional Neural Network (CNN) and Multilayer Perceptron (MLP)—were trained. Model evaluation was performed using confusion-matrix metrics including Accuracy, Precision, Recall, F1-score, and Intersection over Union (IoU). Feature importance was also calculated for each algorithm.Findings: All models successfully distinguished the general burn pattern from the unburned background; however, they differed substantially in numerical accuracy and spatial noise. The MLE classifier, although yielding nearly 98% accuracy, showed the lowest reliability due to a high rate of misclassified unburned pixels (FP) and scattered artifacts around burn perimeters. Among machine learning methods, RF exhibited the best performance, achieving ~99.67% Accuracy, ~97% F1-score, and the highest IoU, with the lowest FP and FN values. SVM also showed stable and competitive performance with an F1-score exceeding 96%, though slightly more boundary-related errors than RF. AdaBoost improved notably over the statistical classifier but, due to sensitivity to difficult samples, produced higher FN values. Both deep learning models (CNN and MLP) generated smooth, low-noise burn maps and achieved Accuracy, F1-score, and IoU values closely matching RF. Feature-importance analysis indicated that shortwave infrared bands (SWIR-1, SWIR-2) and burn/vegetation indices—particularly NBR, NDVI, and SAVI—were the most influential predictors, whereas visible bands contributed less to model decisions.Conclusion: The results demonstrate that integrating Sentinel-2 infrared bands with vegetation and moisture indices, combined with machine learning and deep learning models, provides an accurate and robust framework for post-fire burn-area mapping in heterogeneous landscapes. RF, followed by CNN and MLP, emerges as the most effective set of models for operational implementation, while MLE and AdaBoost serve better as baseline methods. Key limitations include reliance on a single wildfire event and single-date post-fire data; thus, extending the framework to multiple fire regimes, diverse vegetation types, and more complex topographic conditions, as well as incorporating multitemporal data and radar/altimetry sensors, is recommended for future research. The findings support the development of operational wildfire monitoring systems, prioritization of restoration zones, and sustainable resource-management planning in fire-prone regions.
Original Research Paper
Transportation
S. Bolouri; F. Sheidaei
Abstract
Background and Objectives: The vast country of Iran is always exposed to various land hazards. These hazards include earthquakes, subsidence, floods, extreme cold and heat, etc., which can cause serious damage to infrastructure, especially the railway network, and cause a lot of costs to the country's ...
Read More
Background and Objectives: The vast country of Iran is always exposed to various land hazards. These hazards include earthquakes, subsidence, floods, extreme cold and heat, etc., which can cause serious damage to infrastructure, especially the railway network, and cause a lot of costs to the country's railways, while also causing life-threatening hazards to passengers. Therefore, it is important to study these natural hazards in the railway, especially the Zagros railway region, which is located in a mountainous region and has seen many accidents. The aim of this study is to study the impact of some hazards on the railway network.Methods: For this purpose, all layers of slope, elevation, population, rainfall, river, geology, land use, soil type, and fault are analyzed after being reclassified using the AHP method and performing an overlay analysis in GIS.Findings: After overlapping the fuzzy layers, the research findings indicate high to medium hazard in the study area.Conclusion: The results show that this area is one of the most critical areas because about 740 square kilometers of the railway network within a 2-kilometer radius is at moderate to severe hazard and requires better management and more attention from the authorities. The results of past hazards also confirm the performance of this method.
Original Research Paper
Remote Sensing
F. Rabiee; S. Adibzadeh; A. Sharifi; S. Sadeghian
Abstract
Barckground and Objectives: Rapid urbanization has made land use change studies essential for urban planning. Parand, a new town near Tehran, has undergone significant development over the past two decades. This study aims to quantitatively monitor land use changes, focusing on the expansion of built-up ...
Read More
Barckground and Objectives: Rapid urbanization has made land use change studies essential for urban planning. Parand, a new town near Tehran, has undergone significant development over the past two decades. This study aims to quantitatively monitor land use changes, focusing on the expansion of built-up areas and reduction of barren land, compare land use maps for two periods, assess classification accuracy, and analyze the spatial pattern of Parand’s urban growth and its implications.Methods: A multi-sensor approach was applied to analyze urban development, using Landsat-7 (2000) as the baseline and Sentinel-2 (2024) as the contemporary dataset, with Sentinel-1 radar data to improve urban texture separation. Data processing was performed in Google Earth Engine. Four classes—vegetation, built-up, barren land, and roads—were classified using Random Forest. Training samples were derived from visual interpretation and ancillary data. Validation involved an error matrix, overall accuracy, and the kappa coefficient. Outputs included binary built-up maps, classified land use maps, and statistical tables for both periods.Findings: The results of the analyses indicated that the area of built-up land in Parand increased substantially during the study period. In 2000 the built-up area was approximately 3.66 square kilometers, whereas in 2024 it exceeded 14.31 square kilometers. This growth represents an increase of more than fourfold over two decades. At the same time, barren land, which covered 146.66 square kilometers in 2000, decreased to 134.50 square kilometers in 2024, indicating the conversion of a large portion of these lands into urban areas. Spatial analysis of the changes showed that urban growth was mainly concentrated in the eastern and southern sectors of Parand, shifting from an initially dispersed pattern to a more cohesive and organized structure. Comparison of classification performance also showed that the 2024 map, with an overall accuracy of 95.97 percent and a kappa coefficient of 0.9444, had higher quality than the 2000 map, which had an overall accuracy of 89.06 percent and a kappa of 0.8201. This difference is attributed to the higher spatial resolution of the newer data and the fusion of optical and radar sources. Vegetation and roads showed little change, and their relative stability is consistent with the more permanent nature of these classes.Conclusion: Findings show that combining optical and radar data with machine learning in a cloud environment is effective for long-term urban monitoring. Rapid built-up expansion, especially in eastern and southern Parand, highlights the need for revised land use policies, infrastructure development, and agricultural land protection. This approach supports sustainable urban planning and investment prioritization, though challenges such as radar noise, limited field data, and class separability remain.
Original Research Paper
Geo-spatial Information System
A. A. Alesheikh; K. Kheirkhah; F. Rezaie; A. Jafari; M. Panahi
Abstract
Barckground and Objectives: Iran's northern provinces (Golestan, Mazandaran, and Gilan) lie along the Caspian Sea and combine humid climate, steep Alborz topography, intensive groundwater extraction for agriculture, and rapid land-use change, making them simultaneously exposed to floods, landslides, ...
Read More
Barckground and Objectives: Iran's northern provinces (Golestan, Mazandaran, and Gilan) lie along the Caspian Sea and combine humid climate, steep Alborz topography, intensive groundwater extraction for agriculture, and rapid land-use change, making them simultaneously exposed to floods, landslides, and land subsidence. Despite this co-occurrence, most prior studies have addressed each hazard in isolation, providing limited insight into where these threats overlap spatially and how they should be jointly managed at the regional scale. Existing susceptibility models often rely on single machine learning algorithms without systematic hyperparameter tuning and treat their predictions as opaque outputs, limiting their value for evidence-based decision-making and undermining stakeholder trust. This research aims to develop an optimized and interpretable framework that produces joint susceptibility maps for all three hazards across this region, supporting integrated risk planning and informed land-use policy.Methods: A hybrid model combining Support Vector Regression with the Two-phase Mutation Grey Wolf Optimizer (SVR-TMGWO) was built separately for each hazard. Twenty years of recorded flood and landslide events (2000–2020) were compiled from the Geological Survey of Iran and the national watershed-management authority, while subsidence locations were derived from Sentinel-1 InSAR analyses. Non-event points were generated randomly outside a 2000-metre buffer around recorded events at a 1:1 ratio to preserve class balance, and the dataset was partitioned into 70% training and 30% testing within K-means clusters (with the optimal K determined via Gap and Silhouette analyses) so that the feature-space distribution was balanced across both subsets. Eighteen conditioning factors spanning topographic, hydrological, environmental, anthropogenic, and soil-physical variables were prepared for each pixel. The TMGWO algorithm simultaneously tuned the three SVR hyperparameters (C, ε, γ) for each hazard, producing distinct configurations that reflect the differing physics of each process. Model accuracy was evaluated using AUC-ROC, RMSE, R², and adjusted R² on independent test data. To overcome the black-box limitation of machine learning, Shapley Additive Explanations (SHAP) were computed at both global and local scales, exposing the contribution of each conditioning factor to the prediction at the regional level and at individual locations. Finally, the three single-hazard maps were combined into an eight-class multi-hazard map identifying areas of single, dual, and triple hazard overlap.Findings: TMGWO consistently improved test-phase AUC-ROC over the base SVR model, reaching 0.8404 for flood, 0.9329 for landslide, and 0.9642 for subsidence, while narrowing the training-test gap and indicating reduced overfitting compared with the unoptimized baseline. SHAP identified elevation as the leading driver across all three hazards but revealed contrasting secondary controls per process. The multi-hazard map showed that triple-overlap zones were spatially restricted and concentrated along the mountain-plain transition.Conclusion: The proposed SVR-TMGWO framework, combined with SHAP interpretability, produced spatially explicit and physically meaningful susceptibility maps for three coexisting hazards in northern Iran. The resulting eight-class multi-hazard product highlights priority areas for integrated land-use planning, infrastructure protection, and early-warning prioritisation, while the SHAP outputs make the model's reasoning transparent to non-expert stakeholders. Future research should incorporate temporal dynamics through multi-temporal InSAR and climate projections, compare deep learning alternatives such as convolutional neural networks, and extend the framework to full risk assessment by including exposure and vulnerability components.
Original Research Paper
Geo-spatial Information System
A. Rezaie; M. Ghasemi; S. M. Tafreshi
Abstract
Background and Objectives: The rapid growth of digital banking, electronic payment systems, and online financial services has generated vast amounts of transactional and customer-related data. These developments have created new opportunities for data-driven decision-making in the banking sector, particularly ...
Read More
Background and Objectives: The rapid growth of digital banking, electronic payment systems, and online financial services has generated vast amounts of transactional and customer-related data. These developments have created new opportunities for data-driven decision-making in the banking sector, particularly in demand forecasting, service planning, and branch network optimization. Because banking demand is influenced by spatial, demographic, and socioeconomic factors, integrating geospatial analysis with artificial intelligence techniques can provide valuable insights for strategic banking decisions. Despite the increasing availability of banking data, the practical use of spatial data mining and machine learning for demand analysis and branch planning remains limited. Therefore, this study aims to develop and evaluate an integrated framework that combines spatial analysis, machine learning, deep learning, and explainable artificial intelligence to forecast banking service demand and support spatial decision-making for branch network optimization.Methods: The study was conducted using customer and branch data obtained from Bank Sepah in Markazi Province, Iran, covering the period from 2021 to 2023. Markazi Province was selected as the study area because of its diverse urban, industrial, rural, and service-oriented characteristics, which provide a suitable environment for investigating heterogeneous patterns of banking demand. The proposed framework consists of four main stages. First, transactional, spatial, and socioeconomic datasets were integrated and processed through data cleaning, preprocessing, and feature engineering procedures. Second, spatial demand patterns were investigated using exploratory spatial data analysis, including Moran’s I index, Local Indicators of Spatial Association, and density-based spatial clustering methods. Third, predictive models were developed using Random Forest, Extreme Gradient Boosting, Long Short-Term Memory networks, and Spatio-Temporal Graph Neural Networks. Finally, model outputs were interpreted using feature importance analysis and Shapley Additive Explanations to identify the key drivers of banking demand and support decision-making processes. Model performance was evaluated using standard statistical measures, including the coefficient of determination, precision-recall analysis, calibration assessment, and classification performance metrics.Findings: The results revealed significant spatial heterogeneity in banking service demand across the study area. Spatial analysis produced a Moran’s I value of 0.71, indicating strong positive spatial autocorrelation and confirming the existence of clustered demand patterns. The identified hotspots were primarily concentrated in economically active and densely populated counties. Comparative evaluation of predictive models demonstrated that tree-based machine learning algorithms outperformed deep learning approaches for the available dataset. Random Forest achieved the highest predictive performance with an R² value of 0.95, followed closely by Extreme Gradient Boosting with an R² value of 0.94 for forecasting monthly customer visits. In contrast, the Spatio-Temporal Graph Neural Network model achieved a lower performance level (R² ≈ 0.51). Explainability analysis showed that spatial accessibility, customer socioeconomic characteristics, and historical transaction behavior were among the most influential factors affecting demand patterns. The results further demonstrated that the integration of spatial features substantially improved predictive accuracy and enhanced the interpretability of the generated models.Conclusion: The findings indicate that combining spatial analytics, machine learning, and explainable artificial intelligence provides an effective framework for understanding and forecasting banking service demand. Beyond model comparison, the proposed framework supports practical decision-making related to branch network development, service allocation, and digital banking expansion. The identified spatial demand patterns can assist banking managers in prioritizing investment locations, improving service coverage, and optimizing operational resources. However, the limited temporal depth of the available dataset restricted the effectiveness of advanced spatio-temporal deep learning models. Future studies may benefit from longer time-series datasets and real-time transactional information to further enhance predictive performance and spatio-temporal modeling capabilities. Overall, the proposed framework offers a practical and scalable approach for developing intelligent decision-support systems in data-driven banking environments.
Original Research Paper
Geo-spatial Information System
M. Minaei; H. Aghamohammadi; M. H. Vahidnia; A. R. Neshat; S. Behzadi
Abstract
Background and Objectives: Rapid population growth, industrial and agricultural expansion, and the limited availability of surface water resources have significantly increased dependence on groundwater in arid and semi-arid regions of Iran. Groundwater serves as a vital source of water for domestic consumption, ...
Read More
Background and Objectives: Rapid population growth, industrial and agricultural expansion, and the limited availability of surface water resources have significantly increased dependence on groundwater in arid and semi-arid regions of Iran. Groundwater serves as a vital source of water for domestic consumption, agricultural irrigation, and industrial development, particularly in regions where alternative water resources are scarce. However, groundwater quality has been increasingly threatened by various pollutants, particularly dissolved ions such as sodium, sulfate, and total dissolved solids (TDS), originating from both natural hydrogeochemical processes and anthropogenic activities. Elevated concentrations of these parameters can reduce water suitability for drinking and irrigation purposes and may have adverse environmental and socioeconomic impacts. Understanding the complex relationships among groundwater quality parameters is therefore essential for effective water resource management and long-term sustainability. In recent years, artificial intelligence-based techniques have emerged as powerful tools for modeling environmental systems characterized by nonlinear and uncertain relationships. The primary objective of this study is to model and analyze the interactions among major groundwater quality pollutants in Markazi Province using the Adaptive Neuro-Fuzzy Inference System (ANFIS), with the aim of identifying temporal variations in groundwater quality and providing a scientific basis for sustainable groundwater management.Methods: In this study, groundwater quality data including sodium, sulfate, and TDS were collected from monitoring wells distributed across Markazi Province for three time periods corresponding to the years 2017, 2020, and 2023. Prior to modeling, several preprocessing procedures, including data screening, normalization, and outlier removal, were applied to improve data quality and ensure reliable model development. The ANFIS modeling framework was implemented in the MATLAB environment to capture the nonlinear relationships among the selected groundwater quality parameters. Different fuzzy inference structures, including Grid Partitioning, Fuzzy C-Means (FCM), and Subtractive Clustering, were evaluated in combination with hybrid and backpropagation optimization algorithms. The dataset was divided into training (20%) and testing (80%) subsets to assess model generalization capability. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and the coefficient of determination (R²).Findings: The modeling results demonstrated that ANFIS is capable of effectively identifying and reproducing complex nonlinear relationships among groundwater quality parameters. The evaluation metrics indicated satisfactory predictive performance for most of the investigated models. In general, models based on Grid Partitioning combined with hybrid optimization produced more reliable results in the earlier study periods, exhibiting lower prediction errors and higher coefficients of determination. The results also revealed that model performance varied among the investigated years, reflecting temporal changes in groundwater quality conditions and data characteristics. While positive R² values and relatively low RMSE and MAE values confirmed the effectiveness of ANFIS in most cases, some models developed for the 2023 dataset showed weaker performance, indicating increased complexity and variability in groundwater quality patterns. Nevertheless, the overall findings confirmed the capability of ANFIS to model groundwater quality parameters with acceptable accuracy.Conclusion: Overall, the findings of this study confirm that ANFIS is a reliable and efficient tool for groundwater quality modeling and can be applied as an intelligent decision-support system for regional groundwater monitoring and management. The results highlight the importance of selecting appropriate fuzzy inference structures and optimization algorithms to achieve accurate predictions. Considering the critical role of groundwater resources in supplying drinking water and supporting agricultural activities in Markazi Province, the application of ANFIS-based models can enhance management strategies, assist in identifying vulnerable and critical areas, and contribute to the prevention of further groundwater quality degradation. Furthermore, the use of similar intelligent modeling approaches in other regions with comparable hydrogeological conditions can improve groundwater resource management, support sustainable development goals, and contribute to long-term environmental sustainability.
Original Research Paper
Geo-spatial Information System
S. A. Nobakht; R. Ali Abbaspour; A. Chehreghan
Abstract
Background and Objectives: Positioning in indoor spaces and environments without Global Navigation Satellite System signals is one of the major challenges in the development of location‑based services in smart cities. Among various approaches, infrastructure‑independent methods such as smartphone‑based ...
Read More
Background and Objectives: Positioning in indoor spaces and environments without Global Navigation Satellite System signals is one of the major challenges in the development of location‑based services in smart cities. Among various approaches, infrastructure‑independent methods such as smartphone‑based Pedestrian Dead Reckoning are considered a reliable solution, especially in critical conditions and when infrastructures are damaged or destroyed. This research aims to present an innovative approach by providing a two‑dimensional model for simulating and analyzing specific human movement activities in critical environments, including normal walking, crawling on hands and knees, and belly crawling.Methods: The proposed process consists of data collection, signal segmentation, gait cycle detection, step counting, movement type recognition, step length estimation, direction determination, and internal coordinate calculation. Data was collected in two phases: training and testing. The signals were segmented using the Pruned Exact Linear Time algorithm and the Functional Pruning Optimal Partitioning algorithm. Subsequently, movement type was recognized by employing Dynamic Time Warping, and step length was estimated using the Weinberg model.Findings: In the proposed approach, the accuracy of path estimation under different conditions reached an average of 0.17 percent. Furthermore, the direction of movement was determined by fusing data from the accelerometer, gyroscope, and magnetometer using the Gradient Descent Algorithm. The results showed that in the primary Pedestrian Dead Reckoning method, for a path with an average length of 238.226 meters, the mean error was approximately 2.97 ± 1.04 meters. Subsequently, the use of a particle filter based on map matching reduced the error to 0.35 ± 0.07 meters.Conclusion: The results of this study indicate that designing an intelligent and self-adaptive processing chain for analyzing motion signals can significantly enhance the accuracy and robustness of the Pedestrian Dead Reckoning method in indoor environments and critical conditions. Furthermore, the findings confirm that the smartphone sensor‑based Pedestrian Dead Reckoning method can provide accurate and reliable positioning even when the type of movement or the way the device is carried changes.
Original Research Paper
Geodesy
S. A. Adnani,; M. Arian; A. Solgi; M. Shirazian; R. Heidari
Abstract
Background and Objectives: Accurate monitoring of crustal deformation in transpressional tectonic systems is a fundamental challenge in geodynamics and seismotectonic assessment because strike-slip faulting and thrusting occur simultaneously and mechanically interact across adjacent fault segments. In ...
Read More
Background and Objectives: Accurate monitoring of crustal deformation in transpressional tectonic systems is a fundamental challenge in geodynamics and seismotectonic assessment because strike-slip faulting and thrusting occur simultaneously and mechanically interact across adjacent fault segments. In the Central Alborz, oblique convergence between the Arabian and Eurasian plates has resulted in the partitioning of deformation among the Mosha–Fasham, North Tehran Thrust, Purkan–Vardij, Emamzadeh Davud, and Taleghan fault systems. Consequently, the assumption of uniform fault slip cannot adequately represent the spatial and depth-dependent heterogeneity of the deformation and stress fields. This study aims to develop an integrated framework based on remote sensing, geodesy, and fault mechanics to reconstruct the interseismic deformation field, partition the slip rate into stri ke-slip and dip-slip components, and calculate Coulomb failure stress changes throughout the fault network of the Central Alborz.Methods: : C-band synthetic aperture radar data acquired by Envisat from 2003 to 2012 and Sentinel-1A and Sentinel-1B from 2014 to 2022 were used to generate differential interferograms and retrieve surface displacements along the satellite line of sight. Interferometric synthetic aperture radar processing was performed in SARscape and included selecting image pairs with suitable spatial baselines, removing the topographic phase using the Shuttle Radar Topography Mission digital elevation model, applying Goldstein and adaptive filters, unwrapping the interferometric phase, and correcting orbital errors using ground control points. To assess geodetic consistency and impose boundary conditions, the eastward and northward horizontal velocity components, together with their one-sigma uncertainties, were obtained from 45 Global Positioning System and Global Navigation Satellite System stations in a Eurasia-fixed reference frame. The radar, geodetic, seismic, and fault-geometry datasets were spatially co-registered in the International Terrestrial Reference Frame 2000 using a seven-parameter Bursa–Wolf transformation. The velocity field was subsequently transformed between reference frames using the Euler pole of the Central Alborz. Twenty-two fault segments were represented in Universal Transverse Mercator Zone 39 North and discretized into regular (1 × 1)-km elements. The final model comprised 8,248 elements and 16,496 unknown parameters corresponding to the strike-slip and dip-slip components. The model was solved using the boundary element method and dislocation Green’s functions for a homogeneous, isotropic elastic half-space. A crustal shear modulus of 30 gigapascals and an effective friction coefficient of 0.4 were adopted, and Coulomb failure stress changes were calculated at depths of 5, 10, and 15 km. To evaluate the consistency of the model outputs, the radar-interferometry-derived deformation field was compared with ground-based observations in areas of overlapping data coverage, while the modeled stress pattern was compared with the spatial distribution of filtered earthquake events extracted from homogenized seismic catalogs.Findings: The slip distribution on the modeled fault surfaces does not follow a symmetric or elliptical pattern and is strongly controlled by local variations in fault dip, strike, depth, and the geometric linkage between adjacent segments. Along the eastern section of the Mosha–Fasham fault, the left-lateral strike-slip component reached a maximum rate of approximately 4 mm/yr at shallow depths and exceeded the reverse dip-slip component, which attained a maximum of 2.4 mm/yr. At depths of 13–14 km, these components decreased to approximately 2 and 1.1 mm/yr, respectively. In contrast, the reverse component was dominant along the central section of the Mosha–Fasham fault. Along the northwestern branch of the North Tehran Thrust, the maximum reverse dip-slip rate was 1.5 mm/yr, whereas the strike-slip component did not exceed 0.2 mm/yr. The maximum positive Coulomb stress change was +0.092 bar and occurred at a depth of 10 km in the northern Damavand area. In the Fasham–Mosha zone, a correlation coefficient of approximately 0.87 was observed between the spatial pattern of stress changes and the deformation field derived from interferometric synthetic aperture radar. Furthermore, the positive stress lobes exhibited substantial spatial overlap with areas characterized by high concentrations of microearthquakes and aftershocks recorded in the homogenized seismic catalog.Conclusion: The integration of multitemporal interferometric synthetic aperture radar observations, Global Positioning System and Global Navigation Satellite System velocities, and three-dimensional boundary-element modeling demonstrates that subsurface fault geometry and segmentation are the principal controls on the heterogeneous partitioning of slip and the concentration of interseismic stress in the Central Alborz. Compared with the assumption of uniform slip, the one-kilometer fault discretization enables localized variations in fault-motion components and zones of stress concentration to be resolved more effectively. Because the model assumes a homogeneous elastic half-space and spatially invariant mechanical parameters, the results should be interpreted as relative patterns of fault loading and mechanical interaction rather than deterministic predictions of the timing or magnitude of future earthquakes. The proposed framework provides a quantitative basis for prioritizing geodetic monitoring and improving assessments of seismic potential in Tehran and the surrounding regions.
Original Research Paper
Geo-spatial Information System
F. Babayi; A. Vafaeinejad; A. Sharifi
Abstract
Background and Objectives: With the growing global population, increasing pressure on natural resources, climate change, and the rising importance of food security, accurate crop yield prediction—particularly for wheat—has become a crucial research topic. In recent years, the integration ...
Read More
Background and Objectives: With the growing global population, increasing pressure on natural resources, climate change, and the rising importance of food security, accurate crop yield prediction—particularly for wheat—has become a crucial research topic. In recent years, the integration of remote sensing data, climatic information, and soil properties with machine‑learning algorithms has created new opportunities for estimating crop yields. However, many previous studies have either focused on a single data source or have been conducted at limited spatial and temporal scales. Consequently, a gap still exists in conducting comprehensive and comparative evaluations of models using multi‑source datasets. The aim of this study was to evaluate the capability of different machine-learning algorithms to predict wheat yield using climatic variables and the remotely sensed Normalized Difference Vegetation Index (NDVI), and to assess the generalizability of these models across different spatial and climatic conditions.Methods: This study was conducted using multi-source data comprising climatic variables and the remotely sensed Normalized Difference Vegetation Index (NDVI). The study covered nine regions in Germany over a 22-year period, thereby enabling the assessment of spatial and temporal variations in wheat yield. The XGBoost, Random Forest, and Ridge Regression algorithms were employed to predict wheat yield. Among the investigated variables, NDVI and vapor pressure deficit (VPD), as an indicator of atmospheric evaporative stress, were identified as important predictors in the model analyses. Model performance was evaluated in terms of predictive accuracy and error using the coefficient of determination (R²), root mean square error (RMSE), and mean absolute error (MAE). The study was designed not only to analyze the data but also to enable a regional comparison between conventional and advanced predictive models.Findings: The results showed that the XGBoost algorithm outperformed both Random Forest and Ridge Regression in predicting wheat yield. XGBoost was able to reconstruct spatial and temporal yield patterns with lower error and higher explanatory power. The findings also indicated that NDVI and vapor pressure deficit (VPD), as an indicator of atmospheric evaporative stress, were important predictors of wheat yield patterns. Higher model accuracy was observed in more homogeneous and high‑yielding regions, whereas prediction errors increased in areas with greater environmental heterogeneity. Nevertheless, the overall trend demonstrated that machine‑learning techniques—especially gradient‑boosting models—have strong capabilities in capturing the complex relationships between environmental factors and crop yield.Conclusion: Overall, the findings of this study demonstrate that integrating remote-sensing and climatic data with advanced machine-learning algorithms can provide an effective tool for predicting wheat yield and supporting decision-making in smart agriculture. The superior performance of XGBoost compared with the other models indicates its strong capability to capture the nonlinear and multidimensional relationships inherent in agricultural data. Nevertheless, this study had several limitations, including the lack of farm-management data, such as sowing dates, cultivar types, irrigation levels, and fertilizer application rates, as well as the absence of soil data with sufficient spatial and temporal resolution for inclusion in the final models. Moreover, the use of more complex models may increase the risk of overfitting when data are limited. Therefore, future studies are recommended to employ more comprehensive datasets, hybrid spatiotemporal models, and advanced approaches such as recurrent neural networks and convolutional neural networks. Such approaches could improve predictive accuracy and model generalizability while enhancing the practical application of yield-prediction models in sustainable agricultural management and food-security planning.
Original Research Paper
Photogrammetry
A.M. Moazezi Mehr-e-Tehran; M.R. mahmoodi ghouzhdi
Abstract
Background and Objectives: In cultural heritage documentation, the first and most fundamental step involves producing accurate records and maps of the current condition of heritage assets. Recent scientific advancements have enabled the application of innovative tools and methodologies that not only ...
Read More
Background and Objectives: In cultural heritage documentation, the first and most fundamental step involves producing accurate records and maps of the current condition of heritage assets. Recent scientific advancements have enabled the application of innovative tools and methodologies that not only reduce the time required for documentation but also significantly enhance its accuracy and quality. Among the approaches that have attracted increasing attention in recent years is photogrammetry. In Iran, in parallel with global developments and growing international interest in digital heritage documentation, several of the country’s most significant cultural monuments have been documented using photogrammetric techniques. Nevertheless, despite the increasing scholarly and practical attention devoted to this field and the widespread global adoption of photogrammetry, its application within the domain of Iranian cultural heritage—particularly UAV-based photogrammetry—remains limited. Given the vast extent of Iran’s cultural heritage and the growing importance of digital heritage documentation, greater attention to heritage photogrammetry, both theoretically and operationally, is essential. Accordingly, the present study aims to elucidate the process of employing UAV photogrammetry for the surveying and three-dimensional modeling of historical and cultural monuments in Iran.Methods: Due to the outstanding universal value of the Dome of Soltaniyeh in Zanjan and the absence of prior digital documentation using photogrammetric methods, this UNESCO World Heritage monument was selected as the case study. This applied, case-based research was conducted using stereophotogrammetry (stereo-pair photogrammetry) supported by UAV photogrammetry through sequential stages of data acquisition, image processing, and three-dimensional modeling.Findings: Practical implementation of the study demonstrated that an image overlap ratio of 60–80%, as emphasized in similar studies, is essential during data acquisition and aerial imaging to prevent gaps and voids in the three-dimensional model of the Soltaniyeh Dome. Furthermore, due to the dome’s curvature and large scale, employing multiple flight platforms with varying UAV gimbal angles resulted in improved image mosaicking and model reconstruction. In addition, the appropriate spatial distribution of control and check points across different elevation levels ensured satisfactory accuracy in both horizontal and vertical dimensions. The achieved errors—approximately 2 cm in the horizontal dimension and 0.83 cm in the vertical dimension—confirm the reliability and precision of the adopted methodology. Ultimately, highly accurate two-dimensional and three-dimensional digital records of the Soltaniyeh Dome, capable of continuous updating, were produced for the first time, thereby facilitating the conservation and change management of this World Heritage property.Conclusion: The findings demonstrate the considerable capability of UAV photogrammetry for the accurate documentation and modeling of heritage monuments, despite the geometric complexity characteristic of traditional Iranian architecture. Consequently, this method can—and should—be widely employed to produce reliable metric documentation of immovable cultural heritage assets. As a non-destructive technique that does not require physical contact with the monument, UAV photogrammetry provides updatable digital documentation that supports future knowledge-based conservation, monitoring, planning, and management of heritage properties, including the World Heritage Site of Soltaniyeh. Given its accessibility, cost-effectiveness, and compatibility with other documentation techniques, widespread and substantial adoption of this method is anticipated in the near future. Furthermore, integrating heritage digitization requirements into national guidelines for heritage registration dossiers and conservation projects will accelerate the adoption and dissemination of such effective documentation tools. Enhanced interdisciplinary collaboration among specialists in architecture, conservation, and geomatics sciences will likewise contribute significantly to the proper implementation of this methodology and to the production of accurate and reliable heritage documentation.
