---

Background:The online and efficient information about the spatial distribution of wildfire susceptibility and occurrence has a major role in improving of fire prevention activities.
Materials and Methods: In this study a maximum entropy (MaxEnt) approach was used for modeling wildfire susceptibility in East Azerbaijan Province and a webGIS system called fire susceptibility webGIS system (FSWS) was developed to easily share and utilize data and facilities among local people and managers. The methodology was composed of three different phases. First, dependent and independent variables were produced by several methods includingimage processing technique, interpolation method and GIS analysis. Next, the wildfire susceptibility was analyzed by using a MaxEnt approach to predict the possibility of wildfire occurrence based on history of wildfire data and environmental variables (anthropogenic, topography, climate and vegetation datasets) during 2005–2015 and the model performs well in terms of accuracy, with an area under ROC curve (AUC) value of 0.909. Finally, the webGIS system was developed by up to date and proper information.
Results: This webGIS system was provided from the spatial database of variables, wildfire susceptibility map, fire occurrence layers and base maps. FSWS was set up based on ArcGIS component and provided the facilities and capabilities of a web application that would be used by any user even without any prior knowledge of the GIS field.
Conclusions: By FSWS, the environmental authorities will be able to design many operational plans to control the wildfires, supporting conservation managers in improving pre-fire management and raise the awareness among the local people.

---

Aims: This study assesses the impacts of natural and human factors on fire occurrences, identifies key contributors to fire susceptibility maps, and employs machine learning algorithms (MLAs) to enhance the spatiotemporal patterns of fire susceptibility maps.
Materials & Methods: Data were collected from 110 fire locations and 110 non-fire points spanning from 2001 to 2022 at annual scale. Various auxiliary variables, including climate data, terrain features, Normalized Difference Vegetation Index (NDVI), and distance to roads, were analyzed to model fire susceptibility. The study employed multiple MLAs, including Random Forest (RF), Support Vector Machine (SVM), and Gradient Boosting Decision Trees (GBDT), to generate the fire susceptibility maps.
Findings: About 70% of fires occurred within 2 km of roads, indicating significant human influence. Grasslands had the highest fire rates, with over 25% of fires from 2001-2022 due to flammable fuels. The RF and mean models identified 0.4% and 1.31% of the area as very high susceptibility (38,800 km² and 12,600 km²), while the GBDT and SVM models identified 2.42% and 1.86% (234,700 km² and 180,000 km²). The very high susceptibility class, though small in percentage, covers large areas.
Conclusion: This research highlights the importance of integrating environmental and human factors for predicting fire events in arid regions and developing comprehensive fire susceptibility maps, critical for protecting vulnerable ecosystems. These outcomes provide valuable tools for fire management and mitigation strategies within vulnerable ecosystems. Moreover, developing targeted fire management strategies focused on high-risk areas, such as juniper and broadleaf forests must be a priority.
 

---