Take the 2-minute tour ×
Geographic Information Systems Stack Exchange is a question and answer site for cartographers, geographers and GIS professionals. It's 100% free, no registration required.

As one of many processes within a workflow/script, I need to convert polygons to rasters. I am using the feature_to_raster tool in ArcGIS 10.1. This is done within a python script, but for the purpose of trouble-shooting I'm also trying these things in ArcMap.

For most features the process works fine, but there are a handful of features which result in empty rasters (i.e., only NoData cells). I do need to have at least one data cell created for each polygon; empty rasters are not acceptable.

I determined that the problems are very small polygons which don't intersect any cell center of my reference (snap) raster, since the cell size is relatively large in comparison to the polygon. If I don't specify the snap raster, I've found that it does successfully create a raster with a single cell that does have a data value. But I DO need to specify a snap raster so that all my outputs line up, and when I do this, I end up with the empty rasters.

I've tried using the alternate tool, polygon_to_raster, and specifying the cell assignment type to either MAXIMUM_AREA or MAXIMUM_COMBINED_AREA, but I still get the empty rasters.

Does anyone know of a workaround?

Thanks!

EDIT 1: I got polygon-to-raster to work when I added an arbitrary priority field.

share|improve this question
1  
Please edit your question to explain why you are making this conversion. After all, any workaround will be a compromise--it will likely bias many subsequent analyses--so we need to know how to weigh the pros and cons of possible solutions. –  whuber Jun 25 '13 at 19:17
    
I got the process to work as stated in my edit above. But to address your comment: The rasterized polygon is fed into the watershed process to generate the catchment area (which is later converted back to a polygon). I'm not sure whether it's better to underestimate the original polygon (as happens when it uses the cell-center option) or to overestimate it (as happens when using the maximum-area option). –  KHaze Jun 25 '13 at 20:43
    
The original is not underestimated: on average, the polygon areas are correctly represented. But I don't understand what you mean by the "watershed process" exactly: that covers a lot of ground! Regardless, given that (by definition) the missing polygons are tiny, does their omission make any material difference in your calculations? I am also curious as to why you need to create a separate raster for each polygon, because it sounds like they are intended not to overlap. –  whuber Jun 25 '13 at 20:45
    
I am totally new to this forum. How do I mark my question as resolved? –  KHaze Jun 25 '13 at 20:47
    
Okay, the cell-center option represents the polygon correctly, on average. But the max-area option will always overestimate it. And by "watershed process" I mean the watershed tool in ArcGIS spatial analyst. I have to create a separate raster for each polygon, because many polygons do overlap but they each need separate catchments calculated. I can't go into all the details, but this is necessary. –  KHaze Jun 25 '13 at 20:51
show 8 more comments

1 Answer

I got polygon-to-raster to work when I added an arbitrary priority field. It seems from documentation that it should work without that, but it didn't.

Important note: this works for my purpose because I'm only working on one polygon at a time, within a loop. Results when working with a feature class containing multiple polygons may or may not be as desired.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.