I want to compute the value of tree cover loss per year from 2001 to 2014, based on the "Global Forest Change 2000 to 2014" data by Hansen et al.

I downloaded the geotif "lossyear" from http://earthenginepartners.appspot.com/science-2013-global-forest/download_v1.2.html and have added it as a raster layer to my QGIS project (version 2.16.1).

Ultimately I want to compute the tree cover loss for each year in the (green) buffer zones around household locations that I have from a survey. Setting the band rendering under properties to "Singleband pseudocolor" with equal intervals the map with the household location and buffer zones looks like this:

Map of Ghana showing year of tree cover loss (in blue shades), 10km buffer zones in green and household location in yellow

I used "Raster/Zonal Statistics" before to compute the % tree cover loss from a different tif that shows a dot for locations with treecover loss between 2000 and 2014 (so no distinction by year). But how can I distinguish between the different colors i.e. years in this case here, given that its just in a singleband geotif?

  • Is it for example possible to split/decompose the tif "lossyear" into 14 separate tifs for each year? Then I could compute the tree cover loss using "zonal statistics". – Martin Mar 16 '17 at 13:33
  • Hi Martin, I believe that the raster you downloaded does not have the info you are looking for. From the description it says that it contains: "A disaggregation of total forest loss to annual time scales. Encoded as either 0 (no loss) or else a value in the range 1–13, representing loss detected primarily in the year 2001–2014, respectively". Therefore the raster just tells you in which year between 2001 and 2014 losses were "primarily" detected. It doesn't seem like the year by year losses you are looking for are available in this dataset (in any of the rasters offered). – Francisco Mar 17 '17 at 12:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.