Table 2: Satellite-based monitoring of forest fire.
Reference, Year |
Purpose |
Data |
Classifier |
Performance |
Sifakis, et al. [85] |
Hot spot detection |
MSG-SEVIRI geostationary data |
Image Processing |
85% accuracy |
Giglio, et al. [87] |
Hotspot detection |
36 spectral bands (MODIS) |
Image Processing |
False Alarm Rates less than 2% |
Mallinis and Koutsias [91] |
Burned Area mapping |
Landsat Thematic Mapper (TM) imagery |
SVM and 9 others |
More than 93% Accuracy |
Mohler and Goodin [92] |
Burned Area mapping |
spectral bands (MODIS) |
Supervised minimum distance classification |
90% Accuracy |
Pourghasemi, et al. [98] |
Forest fire susceptibility map |
OLI and MODIS satellite images |
Boosted regression tree |
AUC is 0.89 |
Gibson, et al. [99] |
Fire Severity Mapping |
Sentinel2 satellite imagery |
Random Forest |
98% Accuracy |