how to calculate cost path across multiple quadrangles

[MichaelMudd]

When my start and destination points are in the same topo quadrangle, I am able to calculate cost path. However, when they are in different quads, it shows an error. I have tried downloading a county mosaic but they do not come in DEM (only DRG) so the slopes are wonky. I don't know how else to create a "seamless" set of maps using USGS 24k topos and none of the ArcGIS online map data (terrain, DEM 30m, etc.) will download to my map.

[ChrisDonohue__GISP]

Assuming your project site is in the United States, one option is to download the data for the quads you need, combine them so you have full coverage in one dataset, then run the Cost Path Analysis.  I believe you can find DEMs here:

The National Map

Also, it was possible in the past to get a seamless DEM from the site seamless.usgs.gov, but that now directs one to the same National Map link above.  Since the changeover I have not tried to get a seamless DEM, but that functionality may exist in the National Map, so poke around a bit.

Chris Donohue, GISP

[MichaelMudd]

Thank you, Chris. I guess my real problem is how to combine the DEM data. I have plenty of DEM map options but they are separated by quadrangle, which prevents me from creating a path across quad boundaries.

[ChrisDonohue__GISP]

Do you have access to the Spatial Analyst extension?  If so, there are several options available to combine the DEMs.  Here's two I can think of:

1.  DEM to Raster each DEM, then Mosaic to New Raster.

2.  Use Raster Calculator to combine them.

I believe there are several other options available, but on this I am not so knowledgeable.  Let me add some folks who know way more than I do about this:

Darren Wiens

Dan Patterson

Chris Donohue, GISP

[DanPatterson_Retired]

mosaicing them together isn't a problem since the tools are readily available.  A problem that I have seen arise is that a gap can appear along the seam if there is not a slight overlap in the tile extents.  This sliver effectively produces a nodata stripe making traverse impossible or worse... it becomes a preferential travel route.  Filling gaps in dem files is well covered in many threads on GeoNet.  One common option is to filter using a kernel wide enough to cover any gaps.  Another is to produce a seam and use Con to use the values in the good portions of the dem and use a focal mean across the seam. 

So, if you have nodata stripes, you have to deal with them or find a seamless data source.

Mosaicing doesn't appear to be the problem here, but if it is, then the options identified by Chris should work with no problem.

I don't know what "won't download to my map" means?  Data normally has to be downloaded, perhaps unzipped, then added, so some clarification or confusion arises here... and hopefully this is being done in ArcMap and on in AGOL or someother environment

[MichaelMudd]

Thanks to all for the help. The National Map GIS data has a DEM the entire county so merging quads won't be necessary!