Hi,
I've been using DIVA to make up some bioclim variables for future climate scenarios, and it works really well.
I do have one concern though. When I get to the stage of making the .clm file all the input asciis (prec, tmin, tmax) are uniform with the no_data value = -9999. Once the climate information is extracted from them the scale of the data rage appears to change for the output bioclim .grd / .gri. When exported to ESRI ascii, some are nodata= 0, some are -999.somethin and others are -25037 (or something like that).
Now, every data set i produce with DIVA gives me the same range for a particular variable. so comparing DIVA-made bioclim with DIVA-made bioclim layers works fine. I can build models and project.
The only thing is, my other layers are all uniform with the -9999 for no data etc. When I project to the BIOCLIM layers from my other layers, im getting no results.
So, i see there are 2 possibilities for this;
1.) Is there a way I can produce uniform scale ranges for my new bioclim layers?
2.) Do I go back and build all my layers with DIVA (a bit annoying since I have quite a few ready to go that were downloaded)
Cheers,
Matt.
nodata reclass
I am not sure if I fully appreciate your question. But first of all, nodata values should, after importing from the ascii foles, not be "-9999" or some other number. If they are, you should first reclass the grid (e.g. from -99999 to -999 becomes nodata). Leaving in these values could completely mess up your analysis.
Robert
nodata values
This is from following the Julian Ramirez and Aaron Bueno-Cabrera tutorial.
I'm talking about the values in the base ascii files (opened up in notepad) downloaded from worldclim. The prec, tmin, tmax asciis (once converted from .bil) are all uniform, all have the same header and a no_data value of -9999.
Once these files are loaded into DIVA to create .clm and then bioclim variables, the headers are all the same except that no_data value can be a range of different numbers (checked in notepad).
I'm wondering if this will cause problems in the long run. The no_data value from bioclim variable 1 is the same in the present climate scenario as it is for a future climate scenario, but it is different from bioclim variable 2.
I'm a little bit ill at the moment, so having trouble conveying my problem.
nodata values
The value used to flag "nodata" really should not matter to you, as long as the program that you use recognizes it. For example if you export from diva and then import to another program you must make sure that the nodata value is recognized and interpreted as such.
Robert
Thanks for your help, might
Thanks for your help, might use the reclass function to "tidy up", but appears that datasets should be all good for use.
Hey, I have had similar
Hey,
I have had similar problems with Idrisi and no data values, I use a reclass back to my desired value for them, Diva seems to recognise them fine though. When you do operations in Idrisi with the data you may have to use a binary mask, as it seems to have real trouble recognising the no data values