Analiza korelacji pomiędzy wskaźnikami stanu roślinności a wskaźnikami intensywności ognia na przykładzie pożarów leśnych w Grecji w sierpniu 2007 roku

Forest fires influence significantly on the natural environment, forest management and economy. Therefore, it is very important to predict the susceptibility of the object to damage caused by fire. There are a lot of vegetation indexes obtained from the satellite images which are used to forecast the fire hazard. The aim of the study is to verify which vegetation index is the most appropriate to foresee the fire intensity. The study was carried out for the fi res which occurred in Greece in August 2007. The analyzed area had 190 000 ha. MODIS data were used in the study. The paper presents the results of the correlation analysis for some of vegetation indexes: NDVI – Normalized Difference Vegetation Index, NDII - Normalized Difference Infrared Index, NDWI – Normalized Difference Water Index, GVMI – Global Vegetation Moisture Index, SRWI - Simple Ratio Water Index, SIWSI(6,2) – Shortwave Infrared Water Stress Index and indexes of fi re intensity: BAI – Burned Area Index and NBR – Normalized Burn Ratio. The analysis shows that the vegetation condition is only one of the many factors which control the forest fire intensity. The correlation between vegetation and fire intensity indexes varies depending mainly on weather condition; especially maximum wind speed and the content of water vapour and secondarily on the type of vegetation cover. The most suitable vegetation indexes to predict the fi re intensity resulted to be the NDII and SIWSI(6,2), which present Pearson correlation of 0,66. The correlation is usually better for the BAI fi re intensity index then for the NBR. The only exception is the case of forest.