Fires in the Pantanal: modeling and forecasting using multivariate analysis techniques
Datasets usually provide raw data for analysis. This raw data often comes in spreadsheet form, but can be any collection of data, on which analysis can be performed.
Abstract The occurrence of fires in Pantanal causes great damage to the local fauna and flora. Predicting these events is of great importance, enabling catastrophes in this ecosystem to be mitigated or even avoided. This study evaluated the occurrence of fires in Southern Pantanal associated with meteorological variables and created a predictive model using multivariate data analysis techniques. The environmental variables involved in this process were extracted from the database of the Center for Weather Forecasting and Climatic Studies of the National Institute of Space Research (INPE) and the meteorological database for teaching and research of the National Institute of Meteorology (INMET). It was observed that temperature, relative humidity and solar radiation have a close relationship with the occurrence of fires and the resulting correlations were considered satisfactory for the application of forecasting models. The Multiple Linear Regression technique presented an adjustment of 41% and the Integrated Averaging Analysis of Moving Averages presented an adjustment of 66.5% and a general performance of 68.4%, making it the most-recommended forecasting methodology.