Assessment of Slope Instability Hazard Potential Using Logistic Regression Model in the Shahrchay Watershed, Urmia
Document Type : Original Article
Authors
1
Department of Geography Education, Farhangian University, P.O.Box 14665 -889, Tehran
2
Associate Professor, Department of Geology, Faculty of Sciences, Urmia University
3
master of Science, Technical and soil mechanics, makoo. iran.
10.22034/gmpj.2025.490123.1536
Abstract
Introduction
The present study aims to assess slope instabilities in the Shahrchai Basin of Urmia using the logistic regression model. Slope instabilities are significant natural hazards in mountainous areas, having substantial impacts on human lives and infrastructure. These instabilities can be caused by various factors such as geological conditions, topographic features, climatic changes, and human activities. Specifically, the Shahrchai Basin of Urmia, due to its geographical location and unique characteristics, is prone to slope instabilities. Accurate assessment of these instabilities and identification of high-risk areas are of particular importance for planners and crisis managers. In this study, factors such as elevation, slope, slope aspect, distance from fault lines, distance from rivers, distance from roads, lithology, vegetation cover, and land cover were examined as variables influencing the occurrence of instabilities. The significance of this research lies in its potential to contribute to better identification and management of natural hazards and to reduce the damages caused by them. Additionally, this research can serve as a model for other similar areas and be adapted to the specific conditions of each region.
Methodology
In the methodology section, after conducting preliminary studies and collecting necessary data from various sources, maps of instability points distribution were prepared using satellite images. Online data included Sentinel-2 and Landsat-9 (2023) satellite images and a Digital Elevation Model (DEM) with a 10x10 meter pixel resolution. Geological and topographic maps were also obtained from institutional sources. Using GIS tools and data analysis software, maps of slope, slope aspect, distance from faults, rivers, and roads, as well as maps of vegetation cover and land cover, were created. Then, the logistic regression model was applied to these data, and the results were analyzed. For data analysis and preparation of sensitivity maps, various tools in ArcGIS and IDRISI software were used. Slope and slope aspect maps were extracted from the DEM, and maps of distance from faults, rivers, and roads were created using the Euclidean Distance tool in GIS software. Moreover, vegetation cover maps were prepared using the NDVI index and satellite data. Finally, by inputting all data into the logistic regression model, the final analysis was conducted, and results were obtained.
Findings and Discussion
The results of the logistic regression model indicated that variables such as distance from roads, slope aspect, distance from faults, and vegetation cover have a positive and direct relationship with instabilities and were identified as the most significant influencing factors. The logistic regression equation based on the obtained coefficients showed a significant relationship between independent and dependent variables. The sensitivity zoning maps of the area demonstrated that areas with very high risk occupied the smallest percentage of the total area. The analysis of the coefficients obtained from the logistic regression model showed that reducing the distance from roads and faults, increasing the southward slope aspect, and decreasing vegetation cover significantly increase the risk of instabilities.
For preparing sensitivity maps, various classes of slope, slope aspect, elevation, and other factors were precisely analyzed. The results showed that instabilities occur more frequently in areas with slopes of 5 to 20 percent and in south and southeast directions. Additionally, instabilities were more observed within 0 to 1000 meters from faults and in areas with poor vegetation cover.
Conclusion
This research aims to assess the potential risk of slope instabilities in the Shaharchay watershed of Urmia using a logistic regression model. The results indicate that the logistic regression model is highly accurate in identifying the factors influencing slope instabilities and in producing susceptibility maps for the area.
The analysis of the relationship between various factors and the occurrence of instabilities shows that distance from roads, slope direction, distance from faults, and vegetation cover are the most significant factors contributing to increased likelihood of instabilities. The probability of ground instabilities near roads is elevated due to human activities, soil weakening, and construction operations. Slope direction plays a crucial role in land stability by affecting factors such as water saturation, soil erosion, and solar radiation intensity. Proximity to faults increases the risk of instabilities due to tectonic activities and geological changes. Furthermore, vegetation cover acts as a stabilizing factor, playing a vital role in reducing erosion and preventing landslides; its absence can significantly increase the risk of instability.
Statistical indices used in this study, such as ROC, Pseudo R Square, and Chi Square, confirm the validity and accuracy of the proposed model. The high Chi Square value and appropriate Pseudo R Square value indicate a good fit and high predictive power of the model. Additionally, the high ROC value (0.9223) demonstrates a strong correlation between the independent and dependent variables in the model.
The produced zonation maps indicate the distribution of areas with varying sensitivities to slope instabilities. These maps accurately identify high-risk areas and can greatly assist planners and crisis managers in making effective decisions to mitigate risks and improve natural resource management.
The findings of this research can serve as an effective tool for crisis management and reducing damages caused by slope instabilities in the Shaharchay watershed of Urmia. It is recommended that planners and local managers utilize these findings to develop and implement preventive policies and programs. Moreover, this study demonstrates that logistic regression can be an efficient tool for analyzing and predicting slope instabilities in similar areas.
The present study aims to assess slope instabilities in the Shahrchai Basin of Urmia using the logistic regression model. Slope instabilities are significant natural hazards in mountainous areas, having substantial impacts on human lives and infrastructure. These instabilities can be caused by various factors such as geological conditions, topographic features, climatic changes, and human activities. Specifically, the Shahrchai Basin of Urmia, due to its geographical location and unique characteristics, is prone to slope instabilities. Accurate assessment of these instabilities and identification of high-risk areas are of particular importance for planners and crisis managers. In this study, factors such as elevation, slope, slope aspect, distance from fault lines, distance from rivers, distance from roads, lithology, vegetation cover, and land cover were examined as variables influencing the occurrence of instabilities. The significance of this research lies in its potential to contribute to better identification and management of natural hazards and to reduce the damages caused by them. Additionally, this research can serve as a model for other similar areas and be adapted to the specific conditions of each region.
Methodology
In the methodology section, after conducting preliminary studies and collecting necessary data from various sources, maps of instability points distribution were prepared using satellite images. Online data included Sentinel-2 and Landsat-9 (2023) satellite images and a Digital Elevation Model (DEM) with a 10x10 meter pixel resolution. Geological and topographic maps were also obtained from institutional sources. Using GIS tools and data analysis software, maps of slope, slope aspect, distance from faults, rivers, and roads, as well as maps of vegetation cover and land cover, were created. Then, the logistic regression model was applied to these data, and the results were analyzed. For data analysis and preparation of sensitivity maps, various tools in ArcGIS and IDRISI software were used. Slope and slope aspect maps were extracted from the DEM, and maps of distance from faults, rivers, and roads were created using the Euclidean Distance tool in GIS software. Moreover, vegetation cover maps were prepared using the NDVI index and satellite data. Finally, by inputting all data into the logistic regression model, the final analysis was conducted, and results were obtained.
Findings and Discussion
The results of the logistic regression model indicated that variables such as distance from roads, slope aspect, distance from faults, and vegetation cover have a positive and direct relationship with instabilities and were identified as the most significant influencing factors. The logistic regression equation based on the obtained coefficients showed a significant relationship between independent and dependent variables. The sensitivity zoning maps of the area demonstrated that areas with very high risk occupied the smallest percentage of the total area. The analysis of the coefficients obtained from the logistic regression model showed that reducing the distance from roads and faults, increasing the southward slope aspect, and decreasing vegetation cover significantly increase the risk of instabilities.
For preparing sensitivity maps, various classes of slope, slope aspect, elevation, and other factors were precisely analyzed. The results showed that instabilities occur more frequently in areas with slopes of 5 to 20 percent and in south and southeast directions. Additionally, instabilities were more observed within 0 to 1000 meters from faults and in areas with poor vegetation cover.
Conclusion
This research aims to assess the potential risk of slope instabilities in the Shaharchay watershed of Urmia using a logistic regression model. The results indicate that the logistic regression model is highly accurate in identifying the factors influencing slope instabilities and in producing susceptibility maps for the area.
The analysis of the relationship between various factors and the occurrence of instabilities shows that distance from roads, slope direction, distance from faults, and vegetation cover are the most significant factors contributing to increased likelihood of instabilities. The probability of ground instabilities near roads is elevated due to human activities, soil weakening, and construction operations. Slope direction plays a crucial role in land stability by affecting factors such as water saturation, soil erosion, and solar radiation intensity. Proximity to faults increases the risk of instabilities due to tectonic activities and geological changes. Furthermore, vegetation cover acts as a stabilizing factor, playing a vital role in reducing erosion and preventing landslides; its absence can significantly increase the risk of instability.
Statistical indices used in this study, such as ROC, Pseudo R Square, and Chi Square, confirm the validity and accuracy of the proposed model. The high Chi Square value and appropriate Pseudo R Square value indicate a good fit and high predictive power of the model. Additionally, the high ROC value (0.9223) demonstrates a strong correlation between the independent and dependent variables in the model.
The produced zonation maps indicate the distribution of areas with varying sensitivities to slope instabilities. These maps accurately identify high-risk areas and can greatly assist planners and crisis managers in making effective decisions to mitigate risks and improve natural resource management.
The findings of this research can serve as an effective tool for crisis management and reducing damages caused by slope instabilities in the Shaharchay watershed of Urmia. It is recommended that planners and local managers utilize these findings to develop and implement preventive policies and programs. Moreover, this study demonstrates that logistic regression can be an efficient tool for analyzing and predicting slope instabilities in similar areas.
Keywords
اسحاقی، امیر؛ معتمد وزیری، بهارک؛ فیض نیا، سادات. (1388)، پهنهبندی خطر وقوع حرکتهای تودهای با استفاده از روش رگرسیون لجستیک (مطالعه موردی: حوزة آبخیز صفارود)، فصلنامه جغرافیایی سرزمین، 6 (24): 77 -67.
پیروان، حمیدرضا؛ شریعت جعفری، محسن. (1392). ارائه روشی جامع برای تعیین فرسایشپذیری واحدهای سنگشناسی با نگرشی بر زمینشناسی ایران. نشریه علمی – پژوهشی مهندسی و مدیریت آبخیز. 5(3):213- 199.
پیروتی اول، دیمن. (1392). بررسی و پهنهبندی خطر زمینلغزش در حوضه آبریز شهرچای ارومیه تا محل سد (پایاننامه کارشناسی ارشد). وزارت علوم، تحقیقات و فناوری - دانشگاه تبریز، دانشکده جغرافیا.
چپی، کامران؛ طالب پور، داود؛ شیرزادی، عطا الله. (1397). مقایسه مدلهای رگرسیون لجستیک و بیزین رگرسیون لجستیک به منظور پیشبینی مکانی حرکتهای تودهای استان کردستان. پژوهشهای ژئومورفولوژی کمّی. 7(2): 81-60.
حسینزاده، محمدمهدی؛ صدوق، سید حسن؛ متش بیرانوند، سعیده؛ اسماعیلی، رضا.(1396). تحلیل پایداری کرانه رودخانه در برابر فرسایش با استفاده از تنش برشی- مطالعه موردی: رودخانه لاویج. فصلنامه جغرافیای طبیعی. 10(35): 44- 29.
رحیمی، حسین. (1390)، مدلسازی زمانی- مکانی تغییرات پوشش زمین با تلفیق تحلیل زنجیره مارکوف، شبکهای عصبی مصنوعی، سلولهای خودکار، رستمزاده، هاشم، دانشکده جغرافیا، دانشگاه تبریز.
روستایی، شهرام؛مختاری، داود؛ جانانه، کریستینه. (1401). تحلیل کمی ناپایداریهای دامنهای در دامنههای مشرف بر اتوبان در حال احداث تهران- شمال (تهران- سولقان) با استفاده از روش رگرسیون لجستیک. جغرافیا و برنامهریزی. 26(80): 169-157.
روستائی، شهرام؛ خدائی، لیلا. (13975). ارزیابی روشهای تحلیل شبکه ( ANP ) و رگرسیون لجستیک در بررسی پتانسیل وقوع زمینلغزش در محدوده محور و مخزن سد، مطالعه موردی: سد قلعه چای. پژوهشهای ژئومورفولوژی کمّی. 5(3): 80-67
زندی، رحمان؛ امیراحمدی، ابوالقاسم؛ محمدنیا، ملیحه. (1397)، استفاده از مدل آنتروپی در ارزیابی مخاطره زمینلغزش در مسیر جاده پیشنهادی طرقبه – درود (مشهد- نیشابور). جغرافیا و روابط انسانی. 2: 58- 37.
صبوری، سعید محمد. (1402). بررسی ارتباط بین رخداد زمینلغزش با پهنههای گسلی در ایران. دو فصلنامه کواترنری ایران (علمی-پژوهشی). 9(2 و 1): 107-80.
صدیقی، حدیثه؛ قاسمی، احمد رضا. (1402). مدلسازی خطر وقوع زمینلغزش با استفاده از مدل رگرسیون لجستیک (مطالعه موردی: استان چهار محال و بختیاری). پژوهشهای دانش زمین. 14(4): 60-42.
عابدینی، موسی؛ یعقوب نژاد اصل، نازیلا. (1396)، پهنهبندی خطر وقوع زمینلغزش در استان تهران با استفاده از مدل فازی. دو فصلنامه علمی پژوهشی مدیریت بحران. 6( 1): 155- 145.
گلیپور، شیوا؛ حسینزاده، سیدرضا، پورعلی، ملیحه. (1401). شناسایی دامنههای مستعد زمینلغزش و طبقهبندی انواع آن با استفاده از مدل رگرسیون لجستیک و منطق فازی ( مطالعه موردی: حوضه قهرمانلو استان خراسان شمالی). پژوهشهای ژئومورفولوژی کمّی. 11(1): 228-209.
متولی، صدرالدین؛ اسماعیلی، رضا. (1391)، پهنهبندی خطر زمینلغزش با استفاده از اپراتور فازی گاما (مطالعه موردی: حوزه آبخیز طالقان). فصلنامه علمی- پژوهشی پژوهشهای فرسایش محیطی.2(8): 20-1.
متولی، صدرالدین؛ حسینزاده، محمدمهدی؛ اسماعیلی، رضا؛ درفشی، خهبات. (1397). ارزیابی دقت روشهای رگرسیون چند متغیره (MR)، رگرسیون لجستیک (LR)، تحلیل سلسله مراتبی (AHP) و منطق فازی (FL) در پهنهبندی خطر زمینلغزش حوضهآبخیز طالقان. پژوهشهای ژئومورفولوژی کمّی.4(1): 20-1.
معماریان، حسین؛ سیارپور، مرتضی. (1385). نقش پارامتر شیب دامنه در بروز خطا در پهنهبندی خطر زمینلغزش. نشریه دانشکده فنی. 40(1): 113-105.
همتی، فریبا؛ حجازی، سید اسداله.(1396). پهنهبندی خطر زمینلغزش با استفاده از روش آماری رگرسیون لجستیک در حوضه آبریز لواسانات. نشریه تحقیقات کاربردی علوم جغرافیایی. ۱۷ (۴۵) :24-7.
همتی، فریبا؛ مختاری، داود؛ روستائی، شهرام؛ زمانی قرهچمنی، بهزاد.(1398). تعیین حساسیت وقوع ناپایداریهای دامنهای در پیرامون گسل بناروان با استفاده از مدل رگرسیون لجستیک. فضای جغرافیایی. ۱۹ (۶۵) :۱69-151.
Al Zawbaee, O. M. M., & Tabra, H. M. (2020). Identifying factors influencing decision making using logistic regression. General Letters in Mathematics (GLM), 8(2),83-89.
Ayalew, L., & Yamagishi,H. (2005). The application of GIS-based logistic regression for landslide susceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan. Geomorphology , 65 (1-2), 15-31.
Can, T., Nefeslioglu, H. A., Gokceoglu, C., Sonmez, H., & Duman, T. Y. (2005). Susceptibility assessments of shallow earthflows triggered by heavy rainfall at three catchments by logistic regression analyses. Geomorphology, 72(1-4), 250-271.
Chau, K. T., Sze, Y. L., Fung, M. K., Wong, W. Y., Fong, E. L., & Chan, L. C.(2004). Landslide hazard analysis for Hong Kong using landslide inventory and GIS". Computers and Geosciences , 30 (4), 429–443.
Chouhan, S., & Mukherjee, M. (2023). Design and application of a multi-hazard risk rapid assessment questionnaire for hill communities in the Indian Himalayan region. Natural hazards and earth system sciences, 23(4), 1267-1286.
Cook, K., Andermann, C., Puri, B., Schwanghart, W., & Adhikari, B. R. (2022). The influence of river-hillslope interactions on fluvial hazards in the Himalaya (No. ICG2022-508). Copernicus Meetings.
Dai,FC., & Lee,CF. (2002). "Landslide characteristics and slope instability modeling using GIS -Hong Kong. Geomorphology, 42 (3-4), 213–228.
Dastrup R.A.) 2020(. Physical Geography and Natural disasters. Salt Lake Community College. Creative commons Attribution-Noncommerical-ShareAlike 4.0 International Licencse.
Greco, R., Sorriso-Valvo, M., & Catalano, E. (2007). Logistic regression analysis in the evaluation of mass movements susceptibility: the Aspromonte case study, Calabria, Italy. Engineering geology, 89(1-2), 47-66.
Ildromi, A. (2011). The study and modeling land slide hazards by using logistic regression. Case study: Syahdare watershed, Hamedan, Iran.
Jamir, I., Kumar, V., Ojha, A. K., & Gupta, V. (2022). An appraisal of Pre-and Post-failure regimes of a hillslope using stability, runout, and structural implications; A case study from the NW Himalaya (No. ICG2022-185). Copernicus Meetings.
Khan, M. Y., Shafique, M., Turab, S. A., & Ahmad, N. (2021). Characterization of an unstable slope using geophysical, UAV, and geological techniques: Karakoram Himalaya, Northern Pakistan. Frontiers in Earth Science, 9, 668011.
Menard, S. (2002). Applied logistic regression analysis. Thousand Oaks, Calif. Sage Publications, c2002. (2nd ed).
Mohammadi, M., Moradi, H. R., Feyznia, S., & Pourghasemi, H. R. (2010). Comparison of the efficiency of certainty factor, information value and AHP models in landslide hazard zonation (case study: part of Haraz Watershed).
Nefeslioglu, H. A., Gokceoglu, C., & Sonmez, H. (2008). An assessment on the use of logistic regression and artificial neural networks with different sampling strategies for the preparation of landslide susceptibility maps. Engineering Geology, 97(3-4), 171-191.
Pourghasemi, H. R., Mohammady, M., & Pradhan, B. (2012). Landslide susceptibility mapping using index of entropy and conditional probability models in GIS: Safarood Basin, Iran. Catena, 97, 71-84.
Riegel, R. P., Alves, D. D., Schmidt, B. C., de Oliveira, G. G., Haetinger, C., Osório, D. M. M., & de Quevedo, D. M. (2020). Assessment of susceptibility to landslides through geographic information systems and the logistic regression model. Natural Hazards, 103, 497-511.
Sharma, M., Upadhyay, R. K., Tripathi, G., Kishore, N., Shakya, A., Meraj, G., & Thakur, S. N. (2023). Assessing landslide susceptibility along India’s National Highway 58: a comprehensive approach integrating remote sensing, GIS, and logistic regression analysis. Conservation, 3(3), 444-459.
Sun, D., Wen, H., Zhang, Y., & Xue, M. (2021). An optimal sample selection-based logistic regression model of slope physical resistance against rainfall-induced landslide. Natural Hazards, 105, 1255-1279.
Van Westen, C. J., Castellanos, E., & Kuriakose, S. L. (2008). Spatial data for landslide susceptibility, hazard, and vulnerability assessment: An overview. Engineering geology, 102(3-4), 112-131.
