Fire risk associated with landscape changes, climatic events and remote sensing in the Atlantic Forest using ARIMA model

Abstract

The human influence on climate change has increased the occurrence of extreme events and made heat waves and droughts more frequent and severe, which leads to an increased number of fires. It is likely that in the past, the Atlantic Forest has responded to the occurrence of extreme events with fires that may have contributed to the reduction of forest cover in the state. Just as there is a hypothesis that in the future, with the greater recurrence of these events, the Atlantic Forest will be susceptible to greater risk of forest fires. This study aims to develop a model using the Autoregressive Integrated Moving Average Model (ARIMA) to evaluate the danger of forest fires for past and future climatic periods, in the State of Rio de Janeiro. Images from the Thematic Mapper sensor and the Enhanced Thematic Mapper sensor from 1985 to 2015 were used in order to classify them into anthropized and forest areas. A set of meteorological variables on a daily and monthly scale for the period 1985 to 2015 was used to calculate the F-index on a monthly scale. ARIMA was used to simulate the observed and future F-index data up to 2030. The results show higher values of Normalized Difference Fraction Index (NDFI) in areas to the south and southwest of the State, coinciding with the areas of greater predominance of Atlantic Forest. The most degraded regions are in the northeast and north, and the year 2000 showed the largest area of degraded forest. By analyzing the F-index for the past, it was possible to observe a gradual increase in fires, which were associated with extreme events, mainly La Niña. ARIMA modeling allowed us to identify a class change from high to very high in terms of fire danger in the past and future. In 2030, the minimum value of the F-index reached 2.98, which is considered very high in May and June. By analyzing the entire future period on a monthly scale, the highest fire danger values were found in August and September. It is important that actions be taken to minimize the effects of climate change, once that these changes causes an increase in the occurrence of extreme events, which favor a fire-prone climate, increasing the number of forest fires.