SPATIAL MODELLING OF FOREST AND LAND FIRE RISK IN MOUNT CIREMAI NATIONAL PARK

• Main Author: Adesta Ramadhanti, Nadhira
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/51056
• Institution: Institut Teknologi Bandung

Forest and land fires are serious threats to conservation areas such as Mount Ciremai National Park (MCNP) because of the damage on biodiversity, as well as social and economic impact on local population. The intensity of forest fires in this area is mainly caused by human activities like unsupervised land clearing or as a natural result of long dry season. This phenomenon requires improved management and prevention. Fire prevention can be done by creating a risk map for forest fire potential in national park area. Therefore, this research was conducted to make a spatial modelling of MCNP fire risk map by determining the distribution of burnt areas from the fires in 2019. In order to spatially model the fire risk map, this research uses the Maximum Entropy (Maxent) software. The model requires data such as map of delineated burnt areas, environmental variables which consist of DEM map, satellite image, zonation map, hiking trail map, and map of natural tourism parks (ODTWA). These variables were ranked and selected by each of its importance using the jackknife test. The best model was selected by accuracy which was determined by ROC curve (receiver operating characteristics) and the AUC value (area under the ROC curve). The result shows the burnt area of MCNP throughout 2019 is 1083.78 ha, with 929.06 ha in Majalengka Regency and 154.72 ha in Kuningan Regency. The fire risks were classified into five levels; very low (6778.42 ha), low (2914.38 ha), moderate (4538.50 ha), high (295.14 ha), and very high (320.17 ha). Areas that are prone to fire are open lands, while areas with low risk are natural forests. The variables that intensify the risk of fire in MCNP based on Maxent result are land cover, distance from the hiking trail, and distance from tourist attractions with AUC accuracy value of 0.820 (very good).