Automating climate envelope

[PedroGiminez]

Hello.

I am creating a project to show the change in distribution of an insect species in coming decades in the amazon.
I have projected variables such as rainfall and tree cover % for the Amazon. I have used max and min funtions in zonal statistics to give me the overall ranges that the insect is found within.

From there I created a CON function in the raster calculator and input all these variables , e.g. if treecover > 4 but < 67 and so on...
However, I was wondering how I can automate/make this last step easier? Presumably there is an easier way than me having to manually read and input these numbers, potentially making mistakes along the way!

Gracias! 🙂

[PedroGiminez]

(ArcGis Version 10)

[PedroGiminez]

Does anyone have any ideas about the most efficient way to do this?

Many thanks.

[GerickeCook]

Two options here.  You could set up the same functions using ModelBuilder: http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//002w00000001000000

Basically, your workflow is saved and you can always open one of the "bubbles" to inspect or change the values you are entering into the query statements. The logical statements that you're using in Raster Calculator is also available in the Spatial Analyst toolbox, under Math. The only thing I dislike about Model Builder is that it creates many temporary, intermediate raster layers during the workflow.

I personally would just go the Geoprocessing menu, select "Results" and expand the geoprocessing results run from Raster Calculator. You can double-click on the previously run tool to reopen the dialog with all the arguments you specified before. This assumes that you don't turn off the logging option for Geoprocessing.

[JeffreyEvans]

As mentioned, your best approach is to automate this in Molder builder or Python. That said, I would highly recommend reading some of the primary literature on niche modeling. You are misinterpreting the intent of an "envelope model". Using min-max values to define an ecological domain will likely be very misleading and not relevant in future novel climates. Envelope methods are intended to elucidate the multivariate statistical volume representing a species niche. If you are uncomfortable specifying a formal statistical model take a look at OpenModeller (http://openmodeller.sourceforge.net/index.php?Itemid=5&id=10&option=com_content&task=view). It has a nice GUI interface and several statistical models available.