A serious bug regarding vegetation fraction has been discovered in the RUC. The partitioning between surface sensible and latent heat fluxes is strongly controlled by vegetation fraction, specified through high-resolution monthly values from NESDIS. The code in the RUC that linearly interpolates between monthly values to the current Julian date, we've discovered, sets the weight for both months to zero on the Cray due to an integer/real ambiguity. The code running at FSL for MAPS runs has been and still is correct.

This bug means that the vegetation fraction in the RUC-2 since its inception has been zero, meaning that evapotranspiration is also set to zero. This is not too big a deal in winter, when vegetation fraction is low anyway, but it is extremely important near the surface in the warm season.

Some of you (maybe most) have noticed the warm sfc bias in the RUC that John Brown and Phillip Bothwell wrote about in the Forum last week. This bug is responsible for that problem, as well as the dry soil moisture in the RUC, occasional mega-CAPE, and other RUC vs. MAPS differences that have been noted by several folks. (Thanks to all of you.)

We've spoken to NCEP and a bug fix is likely next week. We'll post again as soon as we know more. As when the significant-level bug was discovered 2 months ago, I regret that we did not catch the bug earlier but also look forward to improved RUC performance when it gets fixed. Thanks for your comments and criticism.

Stan Benjamin