|
GIS-based Forest Fire Susceptibility Analysis Using Frequency Ratio, Statistical Index and Weighting Factor Techniques |
|---|---|
| รหัสดีโอไอ | |
| Creator | Worawit Suppawimut |
| Title | GIS-based Forest Fire Susceptibility Analysis Using Frequency Ratio, Statistical Index and Weighting Factor Techniques |
| Contributor | Baromasak Klanreungsang, Ratchaphon Samphutthanont |
| Publisher | Thammasat University |
| Publication Year | 2564 |
| Journal Title | Thai Journal of Science and Technology |
| Journal Vol. | 10 |
| Journal No. | 6 |
| Page no. | 660-676 |
| Keyword | geographic information system, forest fire, forest fire susceptibility map |
| URL Website | https://www.tci-thaijo.org/ |
| Website title | THAIJO |
| ISSN | 2286-7333 |
| Abstract | Forest fire is one of the most common natural hazards occurring in northern Thailand, causingthe loss of forest area and producing hazardous air pollution. The objective of this study was to apply a geographic information system for analyzing forest fire susceptibility area in Chom Thong District, Chiang Mai Province using frequency ratio (FR), statistical index(SI), and weighting factor(WF)techniques. The location data of 745and 306 hotspots were used as training and testing data. This study used eight conditioning factors for the analysis,namely elevation, slope, aspect, topographic wetness index (TWI), normalized vegetation index (NDVI),rainfall, stream density, and land use. All conditioning factors and the training data were analyzed for the rating and weighting scores. The forest fire susceptibility area was classified intofive susceptible levelsnamely very high, high, moderate, low, and very low susceptible levels. The result of FR, SI, and WF, revealed that12.88, 20.39, and 21.34 percent of the total area were identified as a very high susceptible areas whichwere mostly found in the central of the study area. The WF result revealed that the three most influencing factors were elevation, rainfall, and land use.Lastly, thevalidation results using the AUCmethodshow that the success rates were 83.73, 80.90, and 70.50 for WF, SI, and FR methods respectively while the prediction rates of WI, SI, and FRwere84.41, 80.77, and 70.62 respectively. |
