Thursday, May 16, 2013

Your Light Reading Of Today: Sandy Service Assessment

If you haven't read the 66 page assessment on Sandy, you should if you have some time to burn.  If you want the Cliff's Notes version of the assessment, here goes:

1) NOAA found that the forecast was pretty good on the whole.  Emergency managers, media, and commercial weather providers believed that forecasts from the NWS were handled pretty well, and the National Hurricane Center track error at three and five days was below the five year average.  On the whole from the media and emergency management angle, many thought they were prepared well in advance of the storm.

2) The area that was most in need of improvement was the forecasting lead time of storm surge and how it was communicated out to the public.

3) Regarding the whole hurricane warning/non warning issue, here's the Assessment Team's report:

"The Service Assessment Team found differences between many of the views of NOAA’s partners and public opinion research on whether issuing non-tropical watches and warnings for Sandy’s landfall influenced perceptions of severity and response. A University of Pennsylvania Wharton School telephone survey of people along the Mid-Atlantic coast (Virginia to New York) conducted during the Sandy event revealed a majority thought they were under a hurricane watch on the day Sandy made landfall (even though they were under non-tropical watches and warnings). A telephone survey of coastal residents conducted 5 months after the storm also found a significant proportion thought a hurricane watch or warning was in effect when Sandy made landfall (Gladwin, Morrow & Lazo, 2013). The level of concern they reported, however, did not differ substantially from respondents who thought a non-hurricane warning was in effect. Both of these academic surveys also found no statistical difference  between the proportions of coastal residents who planned to evacuate or in fact did evacuate for Sandy versus Hurricane Irene, even though one made landfall with a hurricane warning in effect and the other did not have one."

I found this part particularly interesting. While a majority of people thought they were under a hurricane watch, there was not much of a difference between the level of concern between those who thought there was a hurricane watch and those who did not.  This is a credit to the media and those in the public sphere who got the word out about the storm.  For all of the concern and froth about hurricane warnings and the lack thereof from a number of us out there, most residents still thought they were under a warning and most folks treated the storm with seriousness regardless.  That said, the decision to continue hurricane/tropical storm warnings going forward should a storm transition out of tropical status is a wise one.  Consistency in warnings is important.

That said, we (public, weather community) are overly fixated on titles in front of storms, whether it's a hurricane or not.  The last two tropical (or quasi-tropical) entities to come up the coast have had vastly different impacts.  While Irene was bad, its calling card for most of us was flooding rainfall, with wind as a secondary consideration.  Its track was different in that it was up the coast, as opposed to a direct landfalling system.  Sandy, in the process of transitioning away from tropical to nontropical, had the influence of a deepening mid-level trough to capture it and pull the storm back towards the US coastline and onshore, bringing the "perfect storm" of sorts where damage was widespread along the coast to the north of the landfall point.  The key is to educate the public to worry more about impact and what the storm's bottom line is going to be.  It's also important for media types and the weather community not to get hung up on technicalities -- the job is to educate what the storm is going to do (or what we think it's going to do).

Could better clarity in warnings help that?  Yes...and the service assessment includes that as a finding.  Whether it's consistency in warning continuation or explaining impact, both are needed...and all parties need to ensure that the communication is clear in explaining the impact of the storm, not just that "OMG, a hurricane is coming up the coast."

One last note, Gary Szatkowski, the meteorologist-in-charge at Mount Holly, was cited positively for his harshly-worded warnings in the run-up to Sandy.  This slide in particular (click below for the full version) was a great usage of urging the public to take heed by using past history as a guide.

All in all, the findings suggest better clarity, ease to access information, and consistency in explaining impact.  All of which are good ideas.  Let's be clear, Sandy was a once-in-a-generational type storm (think Hazel in terms of how it was "captured" by a trough) with impacts that were very unusual for this part of the world (how often do storms landfall at Ocean City tracking west-northwest), a storm that we hadn't seen before in the satellite weather era.  Forecasters did do a pretty good job with this storm.  They weren't perfect and there was room for improvement -- the good news is that some of the recommendations should help make unique situations like Sandy a bit easier to manage going forward.