As awful as Frankenstorm, Superstorm hurricane-Nor-easter hybrid Sandy was, it was not even the worst that can happen in the New York area. There have been more powerful storms that hit the area in past history but at a "kinder gentler angle" so they were less destructive.
A worst case scenario would have a storm with an even more southeast to northwest angle but with more powerful winds funneling into NY Harbor and environs storm surge waters 20 feet high topped with waves.
The east coast risk will remain elevated for at least a few more years, or until we exit the current multi-decade cycle in the Pacific and Atlantic oceans and the quiet sun cycle.
Similar to the 1930 and 1950s both the solar cycle sunspot activity and the ocean temperatures in the Pacific and Atlantic favor the kind of heightened threat we've seen the past couple hurricane seasons.
This does not mean it has to happen every year or that we will necessarily get another east coast storm as bad or worse than Sandy, just that we are in a cyclical pattern that ups the ante.
This year most of the storms formed fairly close to the United States, next year if history holds true to past weather, the so-called "Cape Verde" storms off the "African wave-train" will be the ones that grow into monsters and roll toward the Caribbean, with possible impacts on the US Gulf or east coast.
Meanwhile, the patterns of storm tracks and temperatures in September and October thus far are re-enforcing the idea that this winter will be a colder and more snowy one in the nation, especially between the Continental divide and the Appalachians.
Short-term there will be a burgeoning warm spell to above-normal and maybe in some cases record warmth from the Plains to the Southeast for a couple days by Saturday before it turns cold again next week, then another such warm-up may occur by the following Veterans Day weekend and the following week, then another cold shot. Sounds like the back and forth up and down we saw in October doesn't it? Perhaps another signal for the coming winter.
- 33 deaths in US, more than 100 deaths total
- $20 billion in damage, $30 billion in lost business (may end up among the top 5 costliest US hurricanes)
- lowest pressure ever recorded for a hurricane north of the Carolinas
- Sandy was the 8th Atlantic hurricane so far this year
- Wind gust to 90 mph at Islip, NY and 89 mph in Surf City, NJ
- Nearly 8 million without power during the height of the storm
- NYSE closed for consecutive days for the first time since 1888
- more than 3 feet of snow in parts of West Virginia
- 32.5 ft wave in New York harbor which breaks previous record by more than 6 feet
- Record high water level of 13.88 ft at Battery Park in Manhattan
- gusts to 69 mph in Michigan City, Indiana
- 21.7" wave registered in southern Lake Michigan foot shy of all-time record
- Lake Michigan's water level climbed up to 11" above normal thanks to wind piling and low pressure.
The American forecast model failed to foresee Sandy track inland until last few days before landfall. The superiority of the Euro as we call it in forecasting circles is certainly old news to us insiders, although it was not the only foreign model to not curve Sandy out to Sea like the US model did, the Japan and Canadian and the US Navy models also had good forecasts, In fact, the Canadian had it before the Europe model if memory serves.
But the problem as always is just because model X performs best on storm A does not mean model X will perform best on storm B. They all have there moment in the sun so to speak, part of the headache of modern forecasting.
Its mainly that the US model is most often the worst of them all!!, especially in the 5-10 day range.
(and guess which one feeds the point and click free forecasts on the web, hand-held devices, and car screens and is followed most closely by TV people!) You guessed it, the GFS or as I call it the good for shX$
and not just because the European using 4D data integration, it has just as much to do with lousy internal physics and thermodynamic handling combined with medium range truncation of the equations to a more coarse mesh that occurs with the US model and not the others.
But the story is more complicated of course, because the ensemble version of the US model did OK, it was the deterministic version that stunk until within 3 days or so. Interestingly, WITHIN the 5 day period SOME of the metrics we use to measure accuracy showed both the Euro and the American had near record performance on Sandy's global patterns.
We are way behind in super computers in the US, not just in weather but in general!