z3222640_unsw61

Calculating burn frequency

Discussion created by z3222640_unsw61 on Jun 21, 2012
Hi guys,

I have been supplied with a brief whereby I have to develop a biodiversity survey and sampling scheme with reference to the impact of fire on threatened vegetation communities since 1980.

Data sources:

  • Aerial photographs for every year with 5 m res

  • Polygon data of fire extents for each year

  • Polyline data of roads and fire trails

Data conversion:

  1. I have decided to convert the fire extents to raster such that each cell that falls in a fire polygon has a value of one and all other cells will be assigned no data

  2. Calculate the frequency each site has been burnt by summing the rasters created

  3. Calculate the time since the last burn by setting the cell values in each raster to be the year of the burn, taking the maxiumum and then subtracting from 2012

Burn frequency:

  1. Cells that have been burnt 5 times since 1980 considered regularly burnt. Less than 5 infrequently burnt. Recently burnt are those burnt since 2000

  2. Classified rasters converted to polygon to use in sampling

  3. Polygon boundaries simplified in the conversion process

My question is: is it appropriate to use the contains operator when determining which cell falls within a fire polygon or would it be more accurate to use the intersect operator to capture cells that have some degree of fire damange when converting fire extent to raster? Any other criticisms would be greatly appreciated.

Cheers,

Nigel.

Outcomes