An unusually deep trough dug into the east-central U.S. the last couple of days, causing considerable rains along the Gulf Coast and SE U.S. ECWMF forecasts anticipated the enhanced rainfall over this region reasonably well with good lead time. However, over southern and central FL the model's QPFs (hires and ensemble solutions) were uncharacteristically poor and the ECMWF's EFI (for total precip) didn't even register. Unfortunately, because of the model's poor performance (and because no other model handled this forecast scenario well) SFWMD operations failed to prepare for a rain event of this magnitude. At the moment it is unclear what contributed to the forecast failure in the model. Since the ECWMF ensemble system changed in May, we note the model has not anticipated the heavier rain days the SFWMD area of responsibility has seen thus far this wet season. More verification is needed, however, to determine if this is part of a trend. Several graphics are provided. The first two are the observed rainfall in inches over the SFWMD area of responsibility over the 24-hour period from 8 AM Tues, 3 Aug through 8 AM, 4 Aug and by basin (second graphic). ECMWF forecasts available in real time are also included. Comments on this post is welcome.
Thanks Todd Kimberlainfor this interesting post on the forum and for your insights into decision-making / preparation. This is certainly an interesting case for predictability, particularly as you mention that no other model handled this event well either. Your note on the lack of anticipation of heavier rain days since the model change is really useful. I would be keen to look into whether/how this changes with the next update to the model coming in October, and will provide an update here if we can make any conclusions on this soon. Thanks again and best wishes, Rebecca.
1 Comment
Rebecca Emerton
Thanks Todd Kimberlainfor this interesting post on the forum and for your insights into decision-making / preparation. This is certainly an interesting case for predictability, particularly as you mention that no other model handled this event well either. Your note on the lack of anticipation of heavier rain days since the model change is really useful. I would be keen to look into whether/how this changes with the next update to the model coming in October, and will provide an update here if we can make any conclusions on this soon. Thanks again and best wishes, Rebecca.