Issue with segments

[MatthewMeier]

Didn't know exactly where to report this Sara but there are some issues that I have been seeing with the network building for idaho and the ones you have uploaded recently that need to be addressed. I am seeing that not including the artificial segments indeed saves time with canals but enters in the possibility of discontinuity due to small segments missing in the network. These are difficult to see as you are editing and can result in the following issues where overlapping nhd areas occur or culverts under roads. I think this is a prime candidate for a tweak in the code and after speaking with wally is pretty important to remedy. examples include... Maybe this could be an extending line segment to connect to the vertex of another in the segmenting network tool.

[joewheaton]

Chiming in late here. At this point, 'connectivity' is less of an issue as we don't yet do anything with topology. However, I see us using topology soon. I'm still unsure about why we use the Network Prep tool at all. The old logic was 'Trim down NHD network to smallest chunk of network and only run tool on that. ' However, I thought we now pretty much run it on everything (after segmenting) and then break out the perennial and main focal areas (or maybe those intermittents above some threshold capacity), to be our 'primary' output layer, but still have the 'rest' (i.e. intermittents and ephemerals, and canals) calculated everywhere? Can you guys bring me up to speed?

[wally-mac]

To bring everyone up to speed Joe is correct that now we pretty much run BRAT on everything (after segmenting) and then break out the perennial and main focal areas (or maybe those intermittents above some threshold capacity), to be our 'primary' output layer, but still have the 'rest' (i.e. intermittents and ephemerals, and canals) calculated everywhere. The issues that Matt Meier was facing was a artifact of the old logic 'Trim down NHD network to smallest chunk of network and only run tool on that. '

[bangen]

@wally-mac @joewheaton

Quick question regarding the perennial network in the context of BRAT. Yes, we are now running BRAT for the entire network, but are only 'showing' output for the perennial network. So this still requires techs to create a perennial network shapefile. Unfortunately we can't just select streams with an FCODE of perennial since this missing larger rivers. I see 2 options to building the perennial network:

Quick and Dirty Approach

• Select all NHDFlowlines with perennial FCODE ('46006')
• Select all NHDFlowlines within the NHDArea polygons

More Thorough Approach

• Run NHDFlowlines through the NBT
• Run cleanNetwork script (commit bbf3d53) which will resolve the issue shown in Matt's screenshots
• Run checkNetwork script (commit c9d720fbf7eca38a4045228d294053e227a95bd9) which flags potential gaps in the perennial network. See video I put together for Matt here.
• Manual clean-up based on the above

Thoughts?

[wally-mac]

I like the more thorough approach.