Dale,
Thanks for this very helpful information. It sounds like you have noticed a couple of problems. One is a washing out of sfc boundaries that show up in the analyses but dissipate somehow in the fcsts. The second is too large a lake breeze from the Great Lakes.

Do I have it correct that the analyzed winds are OK, but it is the fcst model that washes out the smaller boundaries? If so, then this tells me that we probably do not analyze these boundaries thru a sufficiently deep layer in the analysis (probably only the lowest 25 mb), stronger winds aloft without the boundary mix down, and presto, the boundary is gone.

On the precip in the Great Lakes region, the RUC2 experimental runs at FSL have certainly shown a fair amount of precip over all of Michigan for the month of August in an accumulation of the 6-9h fcsts. I will check to see if we are using the same lake sfc temps still in the experimental run vs. the operational RUC run at NCEP.

We will post some info early next week on a set of changes going into RUC2 later next week (depending on hurricanes, etc.). One of these changes could help maintenance of the boundary layer convergence zones some.

Thanks again, Dale, and we look forward to your future comments.

Stan