# Merging agricultural rasters using ArcGIS Spatial Analyst?

I am working in ArcMap 10.3 and I would like to merge two rasters, one of wheat harvest area, the other of corn harvest area. Looking at the corn rasters table shows the amount of corn harvested in that pixel. I would like to combine the two and get a value for the highest. I am thinking something like set corn to 1 and wheat to 2. Whenever corn is higher than wheat assign a 1 and so on. The output raster then have every pixel showing a 1 or 2 to represent which of the two was higher in that location. Both rasters have the same spatial extent

Can this be done?

I have been reading through the tool list and have not found any tools that look like they will do the job.

• Do you have the Spatial Analyst extension? – FSimardGIS Sep 23 '16 at 17:08
• Yes I do have this. – Jeffkrop Sep 23 '16 at 17:23

You can use the Raster Calculator. Something like this should get you what you want:

``````Con("corn.tif" > "wheat.tif", 1, 2)
``````

I would do this in a two step process. Take all 5 rasters and run a Cell Statistics. Set the Overlay Statistics value to Maximum. This will give you the largest value per cell out of the 5 rasters. Now use raster calculator to do the following:

``````Con("rasterA.tif" == "cs.tif", 1, Con("rasterB.tif" == "cs.tif", 2, Con("rasterC.tif" == "cs.tif",3, Con(rasterD.tif" == "cs.tif",4,5))))
``````
• This works, I just had to remove the zero from the data. – Jeffkrop Sep 23 '16 at 17:26
• Ok so now how do I do it with 5 different crops? – Jeffkrop Sep 23 '16 at 17:57
• @Jeffkrop, edited the answer for 5 rasters. – Brian Sep 23 '16 at 18:31
• There is pick function or similar in sa – FelixIP Sep 23 '16 at 18:42
• After taking the weekend to look over this I now see that the expression Con("corn.tif" > "wheat.tif", 1, 2) shows ONLY where corn is greater than wheat and leaves the rest of where corn is off. So if there is a place where corn grows and wheat does not it will not show anything. The Cell Statistics with the long Con does something similar, but with that one if you put in wheat before corn before soybeans before rice it shows something then if you change the order it shows something else. – Jeffkrop Sep 26 '16 at 17:07

The correct tool for this sort of analysis is Highest Position

You don't need to do cell statistics prior to using it. It accepts list of raster and determines on a cell-by-cell basis the position of the raster with the maximum value in a set of rasters, e.g.