14 November 2017

Attribution Study Shows Hurricane Harvey Floods Becoming Much More Likely

Posted by Dan Satterfield

Check out this paper in PNAS today. When the subject of hurricane experts comes up, there are two names that come into the mind of nearly every meteorologist: Kerry Emanuel and Chris Landsea. There are a lot of others, but these two have published a lot of research that gets sourced frequently. Today Kerry Emanuel has what can be called an attribution study on Hurricane Harvey. These studies look at how likely a given event is and how that may change as the Earth warms.

Emanuel found that the odds of Harvey’s biblical rain totals were about 1% for the 20 year periods from 1981-2000, but by 2100 the odds will be nearly 20% over a period of two decades. The entire paper is here

Image from Meteorologist Brian McNoldy, Senior Research Assoc. at the Univ. of Miami.

Hurricane Forecasts are Improving

In related news, the National Hurricane Center  skill in predicting where a hurricane will be up to 5 days in the future is their best ever so far this year, with an increase in skill in every time period. The cone of uncertainty you see when a hurricane is approaching the coast is based on this error. The width of the cone at any given time period is equal to the average forecast error in previous forecasts. The success of these steely-eyed hurricane forecasters means the cone will be a little smaller in the 2018 hurricane season.

The graph on the right was posted by my friend Brian McNoldy at the Univ. of Miami, and the question is why such a jump in forecast skill? Could it be the increased number of drop-sondes, which give the numerical weather models a better initial analysis? Is it the recently improved European model which often has the best forecast? I’m thinking both, (plus other factors) and Brian is working on a write-up that I’ll share when he posts them. (Update: Read what Brian wrote for Capital Weather Gang here.)

Look at the trend, and you see a steady improvement in track errors. The tough nut to crack however is the intensity forecasts which remain problematic. 

From the NOAA/NHC.

The intensity errors (below) show only marginal improvement:

Trouble Ahead
Real progress is being made, and with climate change increasing the threat, it must continue. Unfortunately, proposed budget cuts threaten future progress, and that’s before you consider this: The new NOAA administrator nominee runs a company that posts ridiculously inaccurate 45-90-day weather “forecasts” (horoscopes?) to a gullible public. I think we need to stick with science and not horoscopes.