Research on improving the forecast method and the forest fire risk warning software in Vietnam
2009
Be Minh Chau; Vuong Van Quynh
The reality indicates that the forecast method and the forest fire risk warning software currently applied in Vietnam still have some limitations. These issues decrease the accuracy of information and forest fire forecast results. This research applied a multivariate statistical method to determine the relationship among the meteorology elements (e.g., temperature, humidity, raifall) and geographical factors (e.g., latitude, longitude, elevation) at 158 national meteorology stations; Analysed the relationship between daily data series of the forest fire risk categories based on the current forest fire forecast formula and the series of forest fire events in reality; Analysed the relationship between the meteorology conditions and forest types with the forest fire risks. The research results indicated that in order to increase the forecast accuracy, it is necessary to interpolate the meteorology condition, to take into account the weather condition and forest types when conducting a forest fire risk forecast. It is possible to divide the forest types into 3 groups corresponding to the different forest fire risk categories. The rainfall, applied in the forecast fomula, appropriate to almost all localities, is at a threshold of 7 mm. This research also created a forest fire risk warning software for the whole country, overcoming the main limitations of the previous software.
Показать больше [+] Меньше [-]