I created a hillshade for the newly projected DEM and went on to deriving a topographic slope in percent. The hillshade would allow the surface terrain to be more easily seen over the Los Angeles County area. Slope is important because fires spread quickly on steep slopes. Fires move upslope as they consume trees up to the canopy and spread to other vegetation as more heat and smoke rises and accumulates. I then reclassified the slope according to the hazard points based on the steepness assigned by the National Fire Protection Association (NFPA). So slopes with a range of 0-10 percent ranked a value of 0 (which is the lowest or smallest fire hazard). Slopes between 10-20 percent were ranked a value of 1, while slopes between 20-30 percent ranked a value of 2. A rank of 3 represents slopes between 30-40 percent and the highest rank of 4 is given to all slopes that are great than 40 percent.
I proceeded to load the vegetation or land cover shapefile and reclassified it by the cover types using points from 0-4 (where 4 represents the most hazardous fuel that is most likely to burn). I then converted the vegetation or land cover feature to raster to add it to the topographic slope in the raster calculator. By adding them, I was able to get a raster of the combined slope and fuel hazard of areas to determine where fires were most likely to break out. Such areas should contain the steepest slopes and the best fuels (cover types with the highest ranking hazard points). Then I went on to add the perimeter shapefiles for the days that the Station Fire occurred and overlayed them onto the reclassified fuels and reclassified slopes of the Los Angeles County area. The three perimeters of the area covered by the Station Fire occurred on August 29th, 30th, and 31st of 2009. These are shown on my maps (two below). On my first map, I overlayed the Station Fire perimeters over my combined slope and fuel hazard raster to create a fire hazard map. On my second map, I showed both reclassifications of the slopes and fuels for the Los Angeles County area with the Station Fire perimeters overlayed on top of them.
The two maps (below) indicate that fires are most likely to burn in areas with the steepest slopes and with the best fuels, which were mainly hardwood forests and conifer forests. These had high fuel rankings (or high fuel model codes), which made them big fire hazards. A mixture of these two produced the best fuel with the highest fuel ranking. Part of the vegetation in Los Angeles consisted of shrubs, which had a lower fuel ranking. However, the lowest fuel ranking went to herbaceous vegetation, which also composes a part of the land cover for Los Angeles County. The land cover types that received a fuel rank of 0 were agricultural, urban, and residential areas. This does not mean that fires cannot break out in these areas, but that they do not occur naturally. And if they do occur in these areas, people are highly inclined to prevent them or put them out. Other areas that had a fuel rank of 0 were areas that have water, are barren, or are covered by rocks or snow. This makes sense because these areas are least likely to burn and are not big fire hazards because they do not provide fuel for fires.
I encountered some difficulties in creating these two maps because at first I had a hard time overlaying the vegetation or land cover shapefile for Los Angeles County onto the topographic slope for the DEM covering this area. I realized that this was due to the fact that the DEM file is a raster file and does not originally have a projection. This same problem occurred when I tried to overlay the Station Fire perimeter shapefiles onto the topographic slope for the DEM in the beginning. So I had to go back and project the DEM (to UTM Zone 11 for the Los Angeles area) and then recreate a slope layer in percent. I also reprojected the shapefiles for the land or vegetation cover and the Station Fire perimeters to the same projection of the DEM so that these would match and overlay correctly with the slope layer of the DEM later. After that, I had to reclassify the vegetation or land cover types using hazard points on a smaller scale compared to the one assigned by the NFPA in the tutorial provided by ESRI. I found a website that gave me some indication of what cover types would be assigned with which hazard points or fuel model codes. Everything went smoothly after I overcame these challenges and the end products show interesting results.
Bibliography:
1) "FRAP - Fire and Resource Assessment Program." FRAP - Fire and Resource Assessment Program. State of California, n.d. Web. 20 May 2010.2) "Geospatial technology for the citizens of Los Angeles County." Geospatial technology for the citizens of Los Angeles County. Los Angeles County Enterprise GIS, n.d. Web. 20 May 2010.
3) Hart, Christopher. "Wildland Fire Lessons Learned Center." Wildland Fire Lessons Learned Center. Arkansas Firewise Communities, 2 Jan. 2005. Web. 20 May 2010.
4) Price, Mike. "ArcUser Online." ESRI - The GIS Software Leader | Mapping Software and Data. ESRI, n.d. Web. 27 May 2010.
5) "The National Map Seamless Server." The National Map Seamless Server. U.S. Department of the Interior | U.S. Geological Survey, 1 Feb. 2010. Web. 20 May 2010.
I proceeded to load the vegetation or land cover shapefile and reclassified it by the cover types using points from 0-4 (where 4 represents the most hazardous fuel that is most likely to burn). I then converted the vegetation or land cover feature to raster to add it to the topographic slope in the raster calculator. By adding them, I was able to get a raster of the combined slope and fuel hazard of areas to determine where fires were most likely to break out. Such areas should contain the steepest slopes and the best fuels (cover types with the highest ranking hazard points). Then I went on to add the perimeter shapefiles for the days that the Station Fire occurred and overlayed them onto the reclassified fuels and reclassified slopes of the Los Angeles County area. The three perimeters of the area covered by the Station Fire occurred on August 29th, 30th, and 31st of 2009. These are shown on my maps (two below). On my first map, I overlayed the Station Fire perimeters over my combined slope and fuel hazard raster to create a fire hazard map. On my second map, I showed both reclassifications of the slopes and fuels for the Los Angeles County area with the Station Fire perimeters overlayed on top of them.
The two maps (below) indicate that fires are most likely to burn in areas with the steepest slopes and with the best fuels, which were mainly hardwood forests and conifer forests. These had high fuel rankings (or high fuel model codes), which made them big fire hazards. A mixture of these two produced the best fuel with the highest fuel ranking. Part of the vegetation in Los Angeles consisted of shrubs, which had a lower fuel ranking. However, the lowest fuel ranking went to herbaceous vegetation, which also composes a part of the land cover for Los Angeles County. The land cover types that received a fuel rank of 0 were agricultural, urban, and residential areas. This does not mean that fires cannot break out in these areas, but that they do not occur naturally. And if they do occur in these areas, people are highly inclined to prevent them or put them out. Other areas that had a fuel rank of 0 were areas that have water, are barren, or are covered by rocks or snow. This makes sense because these areas are least likely to burn and are not big fire hazards because they do not provide fuel for fires.
I encountered some difficulties in creating these two maps because at first I had a hard time overlaying the vegetation or land cover shapefile for Los Angeles County onto the topographic slope for the DEM covering this area. I realized that this was due to the fact that the DEM file is a raster file and does not originally have a projection. This same problem occurred when I tried to overlay the Station Fire perimeter shapefiles onto the topographic slope for the DEM in the beginning. So I had to go back and project the DEM (to UTM Zone 11 for the Los Angeles area) and then recreate a slope layer in percent. I also reprojected the shapefiles for the land or vegetation cover and the Station Fire perimeters to the same projection of the DEM so that these would match and overlay correctly with the slope layer of the DEM later. After that, I had to reclassify the vegetation or land cover types using hazard points on a smaller scale compared to the one assigned by the NFPA in the tutorial provided by ESRI. I found a website that gave me some indication of what cover types would be assigned with which hazard points or fuel model codes. Everything went smoothly after I overcame these challenges and the end products show interesting results.
Bibliography:
1) "FRAP - Fire and Resource Assessment Program." FRAP - Fire and Resource Assessment Program. State of California, n.d. Web. 20 May 2010.
3) Hart, Christopher. "Wildland Fire Lessons Learned Center." Wildland Fire Lessons Learned Center. Arkansas Firewise Communities, 2 Jan. 2005. Web. 20 May 2010.
4)
5) "The National Map Seamless Server." The National Map Seamless Server. U.S. Department of the Interior | U.S. Geological Survey, 1 Feb. 2010. Web. 20 May 2010.
Wow! You did a great job with putting different pieces of spatial data together for analysis concerning the Station Fire. I'm very impressed with the level of analysis you've done. However, I don't fully understand the second map and it's classifications. Extra points for advanced analysis, but points docked for not having a reference map as well.
ReplyDeleteStill great job! 9.8/10