Not specifying basin areas correctly can lead to bad results/not be mass conservative

[arbennett]

bmorph relies on an undocumented assumption when using SCBC that flows coming out of the local bias correction are volumes and need to have a specific modification to the network topology basin areas so that when rerouting mizuRoute interprets the flows correctly. In the case of daily flows the basin areas must be set to 86,400,000 which is the number of milliseconds in a day to work appropriately. If the basin areas are left as is, the flows will be converted to incorrect units, but simply run through mizuRoute. Here's a comparison - scaled is the correct version and unscaled is if you use the actual physical areas:

[bartnijssen]

🤯

[blounsbury-usbr]

Just a note that this is documented in issue #54 as well.

[arbennett]

🤯 - thanks @blounsbury-usbr, hopefully I won't file another issue outlining the same thing in another two years 😓

[bartnijssen]

It's probably not simply milliseconds/day, but a combination of seconds/day and some other unit conversion (like mm /m). Obviously the end result is the same, but it may provide a hint in tracking this down in the code.

[bartnijssen]

I should have read Issue #54 first. As stated there it is m/km.