| The development and evolution of the "5·21" Mongolia wildfire in 2009 and the effect of fire-prevention isolation barrier along the China-Mongolia border were studied by the use of WRF-Fire. The study discussed the fire propagation through weather pattern, fuel model and terrain height. The research conclusions are summarized as follows:The weather pattern in "5·21" Mongolia wildfire in 2009: The fire propagation was affected by the development of the weather system. The area had high pressure and wind speed was small in May 22nd and 23rd, and it was a good time for the firefighting. However, the wind speed of region was increased because of trough development. It was difficult for firefighting. Then using MODIS data obtained by Terra satellite, using 7, 6, 5 RGB synthetic images, analysis the characteristics of the fire spread: The image did not showed the fire point at 03:30 21st May(UTC, the same below), then fire had been formed and eastward expansion at 02:35 22nd; Fire point mainly concentrated in the territory of Mongolia, and the fire area eastern boundary was consistent with the border. The obvious fire point in the boundary at 03:20 23rd; Fire extended about 1 latitudes in the north and south direction (-78 km)from 03:20 23rd to 03:05 25th, meanwhile, there were a few fire points in the Mongolia territory; The fire continued to spread north but this time had no fire point at 03:50 26th.Secondly, using the high resolution land cover data created a mapping with Anderson fuel model to determine the type of fuel. The types of area were composed of grassland,forest and noncombustible. According to the types of local land cover in the study area, we would correspond grassland to Anderson fuel model 1 (short grass); The fire types was mainly surface fires in the area, so it corresponded to Anderson fuel model 11 (closed timber). Therefore, the fuel model consisted of short grass, closed timber and noncombustible in the area.On the basis of determining the fuel model, WRF-Fire was used to simulate the fire propagation. The results showed that the simulation could reproduce the impacts of boundary winds on fire propagation. The simulation for fire area and propagation are shown to be broadly consistent with the observations: The first stage (12:00 21st to 22:00 21st), fire spread southwest slowly in the Mongolia because of low fuel loading on the near-bare surface with northeast wind. The second stage (22:00 21st to 11:00 22nd), wind direction turned to northwest, and fire spread southeast across flat and uniform grassland. Before the fire reached the border, the fire spread rapidly towards the isolation barrier because wind speed increased to reach around 8-10 m·s-1 and wind direction was consistent with slope direction. The third stage (11:0022nd to 20:00 23rd), the barrier and firefighting was effective in preventing the fire entering Chinese territory. Then, the fire line spread north after wind direction turned to south. The fourth stage (20:00 23rd to 12:00 24th), wind speed reached more than 16 m·s-1 and direction turned to northwest. However, fire spread south slowly due to the eastern isolation barrier and lake. When fire line reached the southern barrier, because of barrier and firefighting, the wildfire event ended at 08:00 26th.If there was no barrier, sensitivity test showed that the fire would across the border and spread southeast and northeast following wind direction. In this situation,the main fuel type changed from grassland to forest.Fire area was 1.65 109 m2 at 12:00, 24th (final simulation time), and it was about 70 times before entering the border. That would lead to a great loss of forest. |
PDF Full Text Request