ANNOUNCEMENTS
‘Forest fire risk zones are regions where fire is likely to occur and spread to other regions’ (Jaiswal et al, 2002). Showing regions under different risk zones such as very low, low, moderate, high and very high, such information helps in the identification of priority for resource allocation, planning and management funds and interventions to prevent fire risks. In this study forest fire risk zones are determined for the district of Pauri Garhwal in Uttarakhand using a multi criteria decision making approach of Analytical Hierarchy Process. In order to achieve this several relevant factors contributing towards forest fire risks were considered and a tree species classification map was generated as an important variable using a machine learning approach. A multilayer perceptron neural network is used for carrying out the tree species classification. The classification result show overall accuracy of 87% with a substantial kappa value of 0.752. Weights for each factor were calculated through AHP with integration of GIS. The result of Forest fire risk zone (FFRZ) mapping showed a high correspondence with the fire incidences that had occurred in Pauri Garhwal from 2018 to 2021. About 88% of these incidences occurred in moderate to high risk zones where as only 12% occurred in low or very low risk zones. This study reaffirms the need of Forest Fire Risk Zone mapping as an important intervention measure to mitigate the worst impacts of Forest Fires.
