Identifying the most effective input climate variables of Canadian Forest Fire Weather index system

Early warning systems in natural areas are one of the ways to prevent and manage fires. Many factors that affect fire occurrence can be divided into two categories climatic factors and human factors. Among the most important climatic factors; increasing the temperature, decreasing the rainfall, decreasing the humidity and increasing the wind speed can be mentioned. Thus,forest fire risk assessment and early warning systems have become an important component of land management in recent decades. The Canadian Forest Fire Danger Rating System is one of the most important early forest fire warning systems in the world. The Fire Weather Index subsystem provides relative numerical ratings of various aspects of wildfire potential based on four weather observations. The purpose of this study is to identify the most effective input variables of the system in predicting the probability of forest fire in Golestan Province.

In order to estimate the Canadian Forest Fire Weather Index System, the input variables to the system include four climatic variables: daily observations of maximum temperature, relative humidity, wind speed, and 24-hour precipitation for a period of 21 years (1997-2018) during the fire seasons (April to December). Were collected daily from synoptic and evaporative stations in the province. First, the Canadian Forest Fire Weather Index modeling system was calculated daily during the study period at each station. Then, the correlation test and Multi-Dimensional Scaling analysis between input Weather variables and system output index were performed using Spearman, Pearson, and Kendall correlation coefficients.

The results showed that the variable of maximum daily temperature with a correlation coefficient of 0.911 in the study period has the highest effect on the weather index of Canadian Forest Fires compared to other variables. After temperature, daily relative humidity, daily rainfall and daily wind speed with correlation coefficients -0.89, -0.79, 0.29 have the greatest impact on the output of the system of forest fires in predicting the probability of forest fires in this province. Also, in this study, it was found that Spearman’s correlation coefficient is better than other correlation coefficients.

Forest managers can use this system to predict the likelihood of fire if other variables are not available, using the order of importance of each of the climatic variables and rate the risk in different parts of the province.