Appears that it is somewhat common for FR hydroID catchments to contain multiple NWM flow lines. This forces the crosswalk process to choose one featureid for linking flow data. This can lead to situations where smaller NWM streams are selected over larger NWM streams for a given HydroID.
The root cause of this issue appears to be in the headwaters. I suspect that it's a bug in either aggregate_vector_inputs.py or reduce_nhd_stream_density.py
Appears that it is somewhat common for FR hydroID catchments to contain multiple NWM flow lines. This forces the crosswalk process to choose one featureid for linking flow data. This can lead to situations where smaller NWM streams are selected over larger NWM streams for a given HydroID.
Example near Fayetteville, AR (HUC 11010001) -->