Source: http://www.barefootbenny.com |
For a few days now it has been clear that Sandy would evolve into an exceptional cyclone as it moved into the mid latitudes. The track forecast has proven more difficult.
Below are track forecasts produced by the NCEP Global Ensemble Forecast System over the past four days. Note the wide range of possible tracks in the older forecasts (top figures), including some that took Sandy eastward and well away from the US. Over time, with decreasing forecast lead time, the GEFS ensemble converges toward a solution where Sandy curves westward and makes landfall somewhere between Maryland and southern New England.
Forecasts initialized 0000 UTC 24 Oct. Source: NOAA/ESRL |
Forecasts initialized 0000 UTC 25 Oct. Source: NOAA/ESRL |
Forecasts initialized 0000 UTC 26 Oct. Source: NOAA/ESRL |
Forecasts initialized 0000 UTC 27 Oct. Source: NOAA/ESRL |
Forecasts initialized 0000 UTC 27 Oct. Source: NOAA ESRL |
Ensembles are playing a critical role in forecasts for the Frankenstorm. They help to illustrate the most likely areas of landfall, but they also tell us that we can't precisely nail down the track and location of landfall at that time. That being said, this is a very broad storm and impacts will be widespread. Stay tuned to official forecasts at weather.gov.
On Friday, the operational ECMWF had Sandy making landfall in southern New Jersey. Increasingly this weekend, the GFS has moved toward that solution. Yet another case in which the ECMWF appears superior to the GFS. My personal feeling is that this gap in forecast skill is growing.
ReplyDelete