Gaps in thinned raster flowlines

[AndrewMiller3]

Hi,

I am working with raster NHD flowlines over a large spatial extent in ArcMap 10. In order to reduce processing time and allow a variety of calculations tied to a DEM, land use, etc., I am using the thin tool to reduce the flowlines to a width of one cell. For some reason, I am encountering gaps throughout the network, and these are preventing further analysis. After thinning, I made all non-flowline cells NoData, but subsequently discovered the gaps in the thinned flowlines. I haven't experienced this problem with coarser resolution data, and was wondering if that may be leading to the issues I am having. My extent is approximately 48,000 x 48,000 cells, with a cell size of 10m. Throughout the thinned stream network, there are approximately 800 gaps. My input settings for the thin tool (I've tried a few combinations) include background values of Zero or NoData, no filter, round corners, and a range of maximum thicknesses of input linear features. Experimenting with the max thickness of input linear features has resulted in the relocation of gaps within the network, but they still remain unfortunately. I believe my environment settings are correct. I have been unable to find any discussion of similar issues related to this tool. Does anyone else have experience with this issue or know what might be leading to the errors and how to fix them?

Thanks!

[curtvprice]

Did you do the vector to raster conversion from multiple feature classes? Are the gaps at where those input datasets meet?

If so, it is possible introduce gaps in the rasterization process due to resampling -- especially if you didn't specify a snap raster.

[AndrewMiller3]

Thanks Curt. I had originally selected the vector flowlines in groups from the same feature class. I did this because there were too many objects (nearly 200,000) present in the vector layer to covert to a shapefile and rasterize simultaneously. I believe it took six separate rasters to cover the entire extent. The gaps are not located where these rasters meet, however. I will have to look into my old notes more closely to make sure a snap raster was properly specified. At the time I was working with 30m cell sizes, but resampled to 10m cells when using the mosaic tool to join the rasters. Perhaps an issue occurred during that process. I hope to figure it out soon and will follow up.