Tuesday, January 22, 2019

The NCEP Suite: An Update

Good afternoon and hope most of you East of the Rockies kept warm yesterday. We in Philly had our second big storm (first since Nov - second snowfall since last week) and it took the northernmost route possible. Now we look toward the weekend into next week for our next event. More info can be found on Facebook & Twitter toward the end of the week but a summary is posted on social media. I am writing for two reasons but really one big one: the shutdown.

First, to all the forecasters, great job forecasting these last events and issuing the watches and warnings. You have my deepest gratitude. The reason the NWS spends lots of time forecasting - for those that favor consolidation - is to prepare the warnings: most of us know that the NWS issues severe storm, tropical, flood and winter storm alerts but did you know they issue temperature-, marine-, aviation-, and wind-related alerts too? They are just as critical because all those types lead to lives threatened - exposed skin can become frostbitten and heat stroke requires hospitalization - while the others threaten both lives and property. It can be stressful when the models don't agree (Euro is doing well, GFS is ok but UKMET & CMC didn't do so well the last few days) and that's doesn't take into account the NWS's mission - add that and no paycheck and the stress levels go through the roof. Forecasting is not just monitoring the models but they can tell us what's coming in the next five days with great confidence and - if you follow them daily - sometimes seven or even ten days.

According to Capital Weather Gang (Washington Post's Weather Team), there is only one NCEP employee (no staff either, so thank you!) monitoring the models and while the government is shuttered, they are only authorized to prevent & fix crashes - like loss of data or frozen components. Any other changes are on hold, including the implementation of the 13-km GFS and constantly-running now-4D-VAR-EnKF-hybrid GFS Data Assimilation (initialization) System (GDAS). It was originally scheduled for last fiscal year but has been delayed to early FY2020. If the shutdown persists, it will be pushed back even further. In testing, the new GFS reduced error in TC Track and the skill score (unitless) at 500 mb/ 18,000 ft - a key layer of the atmosphere - improved by 0.01. Taking yesterday into account, that would have told us - earlier - that the storm would most likely turn NW over Arkansas.

Please, Mr. Trump, end the shutdown. These disasters need better predictions, especially hurricanes and tornadoes - it was a former hurricane (Sandy) that led to accelerating this effort. Oh yes, (1) this low produced a strong tornado in Alabama before coming up the mid-Atlantic and (2) damage and especially disruption can occur from any storm. Our predictability falls further behind Europe everyday and that is without the shutdown.

Good job to the forecasters. Good luck to the others once this fiasco ends.

No comments:

Post a Comment