Quantcast

Validation of the Meteosat Second Generation (MSG) fire monitoring product using ground observations over Turkey

Research paper by Ibrahim Sönmez, Ahmet Emre Tekeli, Erdem Erdi, Fatih Demir, Murat Arslan

Indexed on: 02 Aug '13Published on: 02 Aug '13Published in: Arabian Journal of Geosciences



Abstract

Validation of the active fire monitoring (FIR) product from Spinning Enhanced Visible and Infrared Imager sensor of Meteosat Second Generation satellite is performed over Turkey. Ground truth data from the Ministry of Environment and Forestry for the years 2007, 2008, and 2009 is used by taking the active fire months of April–September into consideration. The contingency tables are obtained on a monthly basis and categorical statistics of probability of detection (POD) and false alarm rate (FAR) are derived. The results are introduced for each subclass, namely probable, possible, and combined cases. The highest POD amounts of 8.5, 1.93, and 9.4 % are obtained for the possible, probable, and combined cases, respectively, with the lowest FAR amounts of 70.9, 4.5, and 86.1 % for the same categories. FIR product validation with respect to predefined burnt area thresholds is introduced in the second part of the study in order to investigate burnt area effect in active fire detection. Increasing POD values with respect to burnt area amount is obtained with relatively higher values in probable than possible type. On the other hand, FAR variation is observed to be not correlated with the burnt area amount.