Paper Title
Time Series Analysis and Forecasting of Forest Fire Weather

Abstract
Forest fire is one of the most dangerous natural hazards around the world which meteorological factors are important for conducting forest fire. This objective is to compare the performances between Autoregressive Integrated Moving Average (ARIMA) and Holt-Winters (HW) models for the forecasting meteorological data including average wind speed, precipitation, temperature pressure and relative humidity in northern Thailand. As performance measures of models by using the minimum Mean Absolute Percentage Error (MAPE). The result showed that the HW model obtained better model with precipitation, pressure and relative humidity data. However ARIMA model appear to the better model with average wind speed and temperature data. Then we fit time series models to forecast meteorological factors year 2015. Index Terms- Time Series Analysis, ARIMA, Holt-Winter