Fire Risk Map for Southeast Asia from AVHRR Imagery

The Local Maximum Fitting (LMF) method uses the curve fitting and time-series filtering technique. The trigonometric functions are adopted in this method to extract seasonal patterns from multi-temporal satellite data and decide parameters for each pixel.

The LMF-processed data have very little influence on noises, such as cloud, haze, observation angle, etc. On these images, seasonal changes of vegetation condition are very clear, and the fire spread risk is evaluated from the dryness and amount of inflammable materials of each pixel.