Mapping Fire Risk
This Exercise show how to make a Fire Hazard Risk Map for the Area of Los Angeles County, considering the availability of different GIS Datasets and the powerful tools of ArcGis Spatial Analysis Extensions.
First, It has been collected three DEMs from the Seamless Server Dataset covering 34N117W; 34N118W and 33N118W. Have been unified into one single Raster Elevation Dataset as ESRI GRID.
The second step has been the calculation of relative slopes across the whole area.
The slopes of elevation have been calculated with the percentage mode, and divided into 9 significative ranges that represent strong differences in fires and gases behaves:
3%-6%-12%-20%-50%-85%-150%-300%
The average augment between one range and the following has been divided into 66% (four cases) and 100% augment (four cases) from the previous range.
This in order to distribute an exponential level of change within every class.
These ranges have then been reclassified considering the exponential range values into ten new classes, which combined for the from level 1 to level 8. This scale identify the advantage gained by fire in these areas, considering that steepness is a factor able to increase fires' power and speed, and that warmer gases tend to move in upper elevations of the atmosphere.
This values will weigh up to 8 points or 22.2% into the final hazard scale for the Station Fire Area, which has a total of 36.
The zone where Station Fire sprawled two years ago is an area of high elevation and steepness, which present no rivers or water land, differently form other areas in LA county.
(see map 1)
A second factor for the identification of Fire hazard risk factor has been the land coverage.
Using the scale produced by researchers from Greece (see Ref. 1), who determined risk indices for Mediterranean forest species based on flammability properties of Vegetation, and the Land Cover data from USGS Earth Explorer of Los Angeles County, the different coverages have been grouped into a classification scale with factors ranging from 0 to 25,
which will be added later to the original 1-8 scale for Slopes using Raster Calculation.
Also the presence of water in nearby areas has been taken into account for two reasons:
- First, wetlands act as natural obstacles to fire, and delimitate the sprawl with benefice consequences for the areas close to them.
- Second, in case of Firemen intervent, the availability of water ready to use to fight against fire in crucial areas not always reachable by airplanes and helicopters or when those are used in other zones, is an important factor.
Areas of Station Fire zone have been divided into four ranges corresponding to buffers of 5, 10, 15 and 20 kilometers from an available resource of natural water, from 1 to 4.
The fourth and last factor is the distance from available roads. Roads may act as cut fires,
can facilitate detailed intervents into particular areas, which otherwise necessitate the action of Canadair planes and helicopters. To this factor have been given values of risk augment from 0 (within 200 meters from an existing road) to 3 (more than 2 kilometers from an existing road).
Raster data Calculator allows analysts to weigh every factor easily, according to what other specific experts say regarding particular factors, such as the fuel potential of vegetation or the behave of fire with steepness factor.
The final map is the final result of these four different factors into a weighed, comprehensive map of risk.
The Station Fire occurred into an area matching with these factors, as is it possible to see in the last map. Many Red Squares are within its boundaries. Probably also factors of climate and predominant winds had an active role in determining the shape of the involved area. Cells with different color legend are missing one or more of the factors of calculation, for example are outside the buffer calculated for roads of wetlands, or still missing Land Coverage data or Slope of Elevation.
It is included again the road and wetland layers in order to show a geographic reference.
The power of GIS Analysis is extremely impressing.First, It has been collected three DEMs from the Seamless Server Dataset covering 34N117W; 34N118W and 33N118W. Have been unified into one single Raster Elevation Dataset as ESRI GRID.
The second step has been the calculation of relative slopes across the whole area.
The slopes of elevation have been calculated with the percentage mode, and divided into 9 significative ranges that represent strong differences in fires and gases behaves:
3%-6%-12%-20%-50%-85%-150%-300%
The average augment between one range and the following has been divided into 66% (four cases) and 100% augment (four cases) from the previous range.
This in order to distribute an exponential level of change within every class.
These ranges have then been reclassified considering the exponential range values into ten new classes, which combined for the from level 1 to level 8. This scale identify the advantage gained by fire in these areas, considering that steepness is a factor able to increase fires' power and speed, and that warmer gases tend to move in upper elevations of the atmosphere.
This values will weigh up to 8 points or 22.2% into the final hazard scale for the Station Fire Area, which has a total of 36.
The zone where Station Fire sprawled two years ago is an area of high elevation and steepness, which present no rivers or water land, differently form other areas in LA county.
(see map 1)
A second factor for the identification of Fire hazard risk factor has been the land coverage.
Using the scale produced by researchers from Greece (see Ref. 1), who determined risk indices for Mediterranean forest species based on flammability properties of Vegetation, and the Land Cover data from USGS Earth Explorer of Los Angeles County, the different coverages have been grouped into a classification scale with factors ranging from 0 to 25,
which will be added later to the original 1-8 scale for Slopes using Raster Calculation.
Also the presence of water in nearby areas has been taken into account for two reasons:
- First, wetlands act as natural obstacles to fire, and delimitate the sprawl with benefice consequences for the areas close to them.
- Second, in case of Firemen intervent, the availability of water ready to use to fight against fire in crucial areas not always reachable by airplanes and helicopters or when those are used in other zones, is an important factor.
Areas of Station Fire zone have been divided into four ranges corresponding to buffers of 5, 10, 15 and 20 kilometers from an available resource of natural water, from 1 to 4.
The fourth and last factor is the distance from available roads. Roads may act as cut fires,
can facilitate detailed intervents into particular areas, which otherwise necessitate the action of Canadair planes and helicopters. To this factor have been given values of risk augment from 0 (within 200 meters from an existing road) to 3 (more than 2 kilometers from an existing road).
Raster data Calculator allows analysts to weigh every factor easily, according to what other specific experts say regarding particular factors, such as the fuel potential of vegetation or the behave of fire with steepness factor.
The final map is the final result of these four different factors into a weighed, comprehensive map of risk.
The Station Fire occurred into an area matching with these factors, as is it possible to see in the last map. Many Red Squares are within its boundaries. Probably also factors of climate and predominant winds had an active role in determining the shape of the involved area. Cells with different color legend are missing one or more of the factors of calculation, for example are outside the buffer calculated for roads of wetlands, or still missing Land Coverage data or Slope of Elevation.
It is included again the road and wetland layers in order to show a geographic reference.
MAP 1: Rivers and Wetlands within Station Fire Extent.
MAP 2: Creating a Fire Risk Grid
MAP 3:
1) Determining hazard risk indices for Mediterranean forest species based on particle flammability properties
S. Liodakis, a, , I.P. Agiovlasitisa, T. Kakardakisa, N. Tzamtzisa, D. Vorisisa and E. Loisa
a Department of Chemical Engineering, National Technical University of Athens (NTUA), 9 Iroon Polytechniou Street, Athens 157 80, Greece
Received 9 February 2010; revised 20 June 2010; accepted 23 November 2010. Available online 24 December 2010.
No comments:
Post a Comment