Generation of District Level Deforestation Data

1990 and 2000 land use raster files with 30 meter resolution were made available from Dr. Keshav Bhattarai of the Central Missouri University Department of Geography.

I used raster calculator to generate two different raster files, giving each parcel of land a “vegetation value”: 2 for mixed mature forest or high altitude mixed forest, 1 for young vegetation, and 0 for all other types.

The difference between the two (1990-2000) yielded a deforestation raster file. Here, positive numbers (red) mean deforestation and negative numbers (green) mean forestation.

The high resolution of this file made it large and unnecessarily difficult and time-consuming to work with, so I simplified the file. I merged pixels in a 100:1 ratio; the output pixel is the mean of all the input pixels.

I converted this raster file to a polygon shapefile (in order to do this, I had to multiply the deforestation index by 100 and round to the nearest integer, since ArcGIS only accepts integer input values for this process):

I then identified this polygon deforestation file with a file showing Nepalese districts:

so that I could select deforestation polygons by district name. I manually did this for each Nepali district and jotted down deforestation index means for each district. This allowed the creation of a dataset with a mean deforestation index for each district: