volkcj
commented
7 years ago Exclude the channel units with cover estimate totals <90. when doing this, exclude these channel units from the total channel unit area used to normalize the substrate estimates.
e.g. if units 9 and 11 have substrate estimate totals of 80 and 55, do not include the channel unit areas of units 9 and 11 as part of the 'total channel unit area'.
This should also fix fish cover.
It appears that the VisitID's that have problems are the ones where the field "SumFishCover" isn't equal to 100%. There's a line in the method stating "Criteria: Sum of Fish Cover for site is >= 90%". I'm not sure what that means - Sitka still has the metrics for cases where one or more Channel Unit violates that criteria. Ignoreing any results where one or more Channel Units violates the criteria, I get the same results as Sitka (except for round off error).
Should we worry about this and/or try to find out what Sitka is/was doing for these fields? (It's also worth noting that, with only a single exception, the only VisitID's where this looks like a problem are low numbered VisitIDs - 436 or less, so these are all 2011 metrics except a single one that's from 2014).
nahorniak email on 1/18/17