ارزیابی خطر سیلاب با استفاده از سیستم اطلاعات جغرافیایی رویکردی نوین در مدیریت بحران در شهرستان نیشابور

Floods, as one of the most destructive natural disasters, cause significant human and financial losses annually across various regions. Due to its unique geographical location, diverse topography, and land use changes, Neyshabur County has consistently been vulnerable to devastating floods. This study aims to assess flood risk in this area by integrating Geographic Information Systems (GIS) and the Gradient Boosting algorithm. The analysis utilized seven key indicators (elevation, slope, slope aspect, precipitation, distance from waterways, geology, and land use) derived from multiple sources, including Digital Elevation Models (DEM), satellite imagery, and meteorological data. The results indicate that topographic factors (77% importance) and precipitation (61%) had the highest impact in the flood risk prediction model. Furthermore, flood hazard zoning was conducted across five classes, with approximately 20% of the county's area classified as "critical risk." These high-risk zones are primarily located in low elevations, near waterways, and areas with weak vegetation coverage.The model's performance evaluation, based on ROC-AUC (0.91) and RMSE (0.19), confirms its high accuracy in predicting flood-prone regions. The findings of this study can assist urban planners and policymakers in identifying high-risk areas, developing risk mitigation strategies, and optimizing land use management to minimize flood-related damages. This research also highlights the effectiveness of machine learning and GIS-based hybrid approaches in natural hazard assessment.

Floods, as one of the most destructive natural disasters, cause significant human and financial losses annually across various regions. Due to its unique geographical location, diverse topography, and land use changes, Neyshabur County has consistently been vulnerable to devastating floods. This study aims to assess flood risk in this area by integrating Geographic Information Systems (GIS) and the Gradient Boosting algorithm. The analysis utilized seven key indicators (elevation, slope, slope aspect, precipitation, distance from waterways, geology, and land use) derived from multiple sources, including Digital Elevation Models (DEM), satellite imagery, and meteorological data. The results indicate that topographic factors (77% importance) and precipitation (61%) had the highest impact in the flood risk prediction model. Furthermore, flood hazard zoning was conducted across five classes, with approximately 20% of the county's area classified as "critical risk." These high-risk zones are primarily located in low elevations, near waterways, and areas with weak vegetation coverage.The model's performance evaluation, based on ROC-AUC (0.91) and RMSE (0.19), confirms its high accuracy in predicting flood-prone regions. The findings of this study can assist urban planners and policymakers in identifying high-risk areas, developing risk mitigation strategies, and optimizing land use management to minimize flood-related damages. This research also highlights the effectiveness of machine learning and GIS-based hybrid approaches in natural hazard assessment.

Floods, as one of the most destructive natural disasters, cause significant human and financial losses annually across various regions. Due to its unique geographical location, diverse topography, and land use changes, Neyshabur County has consistently been vulnerable to devastating floods. This study aims to assess flood risk in this area by integrating Geographic Information Systems (GIS) and the Gradient Boosting algorithm. The analysis utilized seven key indicators (elevation, slope, slope aspect, precipitation, distance from waterways, geology, and land use) derived from multiple sources, including Digital Elevation Models (DEM), satellite imagery, and meteorological data. The results indicate that topographic factors (77% importance) and precipitation (61%) had the highest impact in the flood risk prediction model. Furthermore, flood hazard zoning was conducted across five classes, with approximately 20% of the county's area classified as "critical risk." These high-risk zones are primarily located in low elevations, near waterways, and areas with weak vegetation coverage.The model's performance evaluation, based on ROC-AUC (0.91) and RMSE (0.19), confirms its high accuracy in predicting flood-prone regions. The findings of this study can assist urban planners and policymakers in identifying high-risk areas, developing risk mitigation strategies, and optimizing land use management to minimize flood-related damages. This research also highlights the effectiveness of machine learning and GIS-based hybrid approaches in natural hazard assessment.

Floods, as one of the most destructive natural disasters, cause significant human and financial losses annually across various regions. Due to its unique geographical location, diverse topography, and land use changes, Neyshabur County has consistently been vulnerable to devastating floods. This study aims to assess flood risk in this area by integrating Geographic Information Systems (GIS) and the Gradient Boosting algorithm. The analysis utilized seven key indicators (elevation, slope, slope aspect, precipitation, distance from waterways, geology, and land use) derived from multiple sources, including Digital Elevation Models (DEM), satellite imagery, and meteorological data. The results indicate that topographic factors (77% importance) and precipitation (61%) had the highest impact in the flood risk prediction model. Furthermore, flood hazard zoning was conducted across five classes, with approximately 20% of the county's area classified as "critical risk." These high-risk zones are primarily located in low elevations, near waterways, and areas with weak vegetation coverage.The model's performance evaluation, based on ROC-AUC (0.91) and RMSE (0.19), confirms its high accuracy in predicting flood-prone regions. The findings of this study can assist urban planners and policymakers in identifying high-risk areas, developing risk mitigation strategies, and optimizing land use management to minimize flood-related damages. This research also highlights the effectiveness of machine learning and GIS-based hybrid approaches in natural hazard assessment.