Sometimes when it comes to modeling it is necessary to compare layers from several sites and define which is the best. There are also cases in which we need layers from several sites because they are in one or the other. Is it possible to convert, for example, a climatic layer 5 arcminutes ~ 10km in .tiff to 2.5 arcminutes ~ 5km in .asc in R, or some other program?
Dear Asiel,
of course it is possible to increase the resolution of individual layers to match that of others available at higher resolutions. However, this process (downscaling) involves interpolation. Therefore, it would be better to find alternatives for those course data that are closer in resolution to the remainder of data. Imho, it might be preferable to go the opposite direction (upscaling) and decrease the resolution of all layers to that of the coarsest layer required to avoid interpolation.
Finally, a lot of global data is available at 1km2 which should be sufficient for most applications. For example check
EarthEnv -- www.earthenv.org/
and CHELSA -- http://chelsa-climate.org/
The former includes a large collection of topographic variables, but also proportional land cover, etc. The latter is one alternative to WorldClim and provides bioclimatic variables.
The R commands for your specific question are:
downscaling:
raster::disaggregate
upscaling:
raster::aggregate
You can also use raster::projectRaster to exactly match the extent, projection and resolution of an existing layer. Mind, that the projection should be an equal area projection if the study area encompasses a large latitudinal range.
Cheers,
R
- Badges
- Science topic
- Similar topics
- Engineering
- Modeling