Tuesday, October 29, 2013

Sandy's Accurate Forecasting In Two Snapshots

In the wake of today's "anniversary" of Sandy, it goes without saying that the storm was accurately forecast by several and by the European computer model. We briefly touched on the Euro's track record last week...and it doesn't go without repeating how good of a job it did in comparison to the GFS with handling the storm several days out.

This was the forecast a week prior to Sandy. Euro 1, GFS...(click the link to get an idea of where I'm going with it).

Modeling from Tuesday 10/23/2012 -- graphic from Storm Vista Weather Models.

Granted, the GFS did improve as we closed in but the Euro's relative accuracy in some type of East Coast hit was the calling card.

The other accurate forecast was that of the National Hurricane Center regarding the storm's track, its "cone" of impact, and where they projected the storm to head.  Beginning with the Thursday 10/25/12 11 AM outlook, their forecast track rarely wavered over the succeeding five days, staying largely within a swath of real estate between Bethany Beach and Long Beach Island.

Source: Corey Pieper

Given we're talking about a forecast of a storm that had just cleared Cuba and had about 1000 miles of real estate yet to travel, that's pretty darn accurate forecasting from the National Hurricane Center.  The impacts that the National Weather Service communicated out were awesomely accurate as well -- and the words of caution were very wise.

Tomorrow, we'll talk about the "lesson" of Sandy -- where we needed to improve communication in this business and whether or not the weather industry and weather community has figured it out yet. Not sure that's the case universally, yet, but I think there has been some progress on this.