search

dailyerosion / dep

Iowa Daily Erosion Project (version 2)
MIT License
4 stars 7 forks source link

Climate File Version Debacle #83

Closed akrherz closed 2 years ago

akrherz commented 3 years ago

An audit of our DEP climate files revealed an unfortunate quirk whereby setting the first line of the file to 4.30 (cliver) results in a 0.7 factor being applied to the computed rainfall intensity (Le Sigh, Le Pain). On a daily basis, this can have a large impact in runoff calculated, but does not seem to impact soil erosion values at all?!? In bulk over time, the differences seem strangely small too.

Well, anyway. We had better document this and fix things up.

akrherz commented 3 years ago

Additionally, it was found that we could use 4 decimals for time precision such that exact seconds can be computed and thus have true two minute data.

akrherz commented 3 years ago

"Not so fast!" - Lee Corso, who is in Ames this weekend... There may be other bugs at play here than this 0.7 factor, debugging more.

akrherz commented 3 years ago

After a number of days of debugging and kind exchanges with the WEPP developers, we have come to the conclusion that a different bug was at play here, but did not impact the erosion numbers in any meaningful way. The bug was that a calculation to compute new snow depth based on dew point was omitted and a simple 10x relationship was assumed. While this could dramatically impact runoff during snowmelt on any given day, the overall impact summed for longer periods of time was near zero given the conservation of water. Whew, disaster averted.

akrherz commented 2 years ago

This issue can be closed with the following summary. 1) Using 4.3 as the climate file version is proper 2) Our results were not impacted by this issue. 3) We await a released version of WEPP with a proper fix to the snow depth calculation