Extraction by Mask-Seeking Guidance

[JohnNash]

My professor was unable to resolve this problem and suggested that I post the problem here. I am using ArcGIS Desktop and want to extract by mask precipitation data (IDW) just for the state of connecticut. I use the precipitation data as the input and the mask as the state of Connecticut and there is no error but it doesn't look quite right. It extracts the data for the most part but isn't very precise, especially the southwest corner of the state where the state juts out. Enclosed find the attached image. My professor said he is having the same problem as me and suggested I post my problem here. This is my first GIS class so any help would be greatly appreciated. Thanks

*updated image

[curtvprice]

The geoprocessing extent is set to the extent of your IDW tool inputs by default. If you want the IDW interpolation to cover a larger extent, you need to set the GP extent to include the entire state.

After you generate your surface, you can clip your surface to the state boundary using another tool such as Clip_management.

[JohnNash]

Once I have a larger extent, which I now do, should a extraction by mask (state of connecticut) work? I am trying to use that too but when I do I am no longer getting precipitation values. I am getting a raster with a max value of 3.4e^38 to -3.4^e38. Maybe it is just late and I am too tired 🙂

[JohnNash]

I changed the extent and then I got what is in the first attachment (Updated project). Now I will try and mask this data. I will use this raster as the input and then I use Connecticut Counties as the mask. When I do this it appears that the new raster created no longer is observing precipitation for some reason. It is showing a range of 3.4e23 to -3.4e23. I wasnt sure what this value was but when examining the updatedproject raster i noticed that this value (3.4e23) is "no value" under the source tab for the raster. Do you have any idea what is going on?

[curtvprice · Accepted solution]

It is showing a range of 3.4e23 to -3.4e23.

This means your output is corrupted. This can happen sometimes with raster operations when there is a bug involving bit depth, or you did something that produces invalid results. This kind of sounds like a bug.

Have you tried clipping your raster using Clip_management? That may not only work, but I learned last week is dramatically faster than the Extract By Mask tool. (Clip_management isn't coded to have to work with Map Algebra so it's a bit of a "shortcut".)

[JohnNash]

Thank You! This method is giving the desired results. I had never used this method before and was unfamiliar with it. In the future if the mask function isn't working I will be able to use this method, which is much quicker as well.

[JeffreySwain]

I wonder after your extract if the statistics were present on the raster or if there was a stretch that could be applied.  Depending on your settings in your mxd, you might need to generate the statistics and apply the stretch.