September 02, 2021

Disaster Hits the New York Metropolitan Area. We Need to Do Better.

More people died last night in the New York area from the "remnants" of Hurricane Ida than over Louisiana and Mississippi as one of the most powerful hurricanes of the century made landfall on a low-lying coastal zone.  (Current count:  about two dozen in the NY area, about 8 in Louisiana)

Think about that.    And as I will suggest in this blog, we can do much better,  both in terms of forecasting and communicating serious weather threats.  There are investments and new policies that are needed.

Dramatic video of what happened last night

Heavy rain, flooding, and even tornadoes struck in a relatively narrow band stretching southwest-northeast from Pennsylvania and New Jersey, through New York, Connecticut and Massachusetts, as shown by a recent precipitation analysis by the NOAA/NWS Weather Prediction Center for the 48-h ending 6 AM EDT  this morning (see below).

A fairly large area of more than 6 inches of rainfall, with some locations hitting 8-10 inches. 

 Much of that rain fell during short intense bursts of precipitation associated with thunderstorms.  Newark, NJ, experienced 3.24 inches in one hour, with NY Central Park hit by 3.15 inches in one hour.  Both are hourly records for those sites.

The intense convective (thunderstorm) nature of the rain is illustrated by a weather radar image at 9:50 PM EDT last night, with the red colors indicating extraordinarily high rainfall rates.

After landfall, Hurricane Ida transitioned into a tropical storm and then underwent extratropical transition in which it took on the characteristics of a midlatitude cyclone.  The National Weather Service sometimes calls the resulting storm a Post-Tropical Cyclone.

But there was a real danger in this transition.

Extratropical cyclones have strong upward motion, often associated with frontal zones, where temperatures and winds change rapidly.  And tropical storms undergoing transition often entrain large amounts of tropical moisture that can result in heavy precipitation as the moist air is forced to rise by the storm circulation.  This moisture can be converted to rain very rapidly in strong thunderstorms/convection.

This is exactly what happened last night.

Below is a map of sea level pressure and atmospheric moisture (called precipitable water) for 10 PM PDT last night; you can see the low-pressure center and the plume of moisture (green colors) moving in from the southwest.

To the east of the low center, there was a warm front, as indicated by the National Weather Service analysis for 8 PM EDT (indicated by the black half circles)

The warm front had warm, southerly winds on the south side and cooler, easterly winds on the north side, with the warm, moist, unstable air to the south forced to rise by the front, resulting in heavy convective showers to the north of the line.  That is why the intense precipitation paralleled the front.

The Forecast

My colleagues at the National Weather Service had warnings out much of yesterday for heavy precipitation and the potential for flash flooding, with a flash flood watch out more than a day ahead of time.

Our models were useful but had some issues in both intensity and position.  High resolution is critical for this kind of forecast for many reasons, including the convective (thunderstorm) nature of the heavy rainfall, and the sharp frontal boundary that helped produce the rain. 3-4 km grid spacing is a minimum that is viable.

The highest resolution model run by the National Weather Service several times a day is the HRRR model....the High-Resolution Rapid Refresh Model, run with 3-km grid spacing.  Its total precipitation for a run starting at 8 AM on Wednesday for precipitation for the next day showed the band, but it was displaced a bit north and somewhat underplayed the precipitation intensity.

The National Weather Service also runs a small (7-8 member) ensemble of several high-resolution simulations (around 3-km grid spacing), called HREF  A statistically enhanced mean of these runs (starting at the same time) is shown below.  Better.

Everyone involved in numerical weather prediction knows what this country needs to get such forecasts correct:  a relatively large (30-50 member) high-resolution (3-km or better for the grid spacing) ensemble of forecasts that are carefully calibrated to give good probabilistic/uncertainty predictions.  

Committee after committee, workshop after workshop, have recommended this.  The National Weather Service's own modeling experts say the same thing.  

But the investment is never made to do this.  This means acquiring the necessary computer resources and building the modeling/statistical post-processing system.   Very, very frustrating that this critical capability is pushed off into the future.  (Senator Cantwell please take this on!)


We obviously failed last night.  Around a dozen people died in flooded basements.   Many people took to the roads that were flooded out. Abandoned cars were everywhere.

There are two stages to the warning process based on time.  Yesterday, hours before heavy rain, it was clear that a serious event was in store and people had to be warned in the strongest terms to stay off the road and to prepare.

By late in the afternoon it was clear that a severe event was about to take place, and we needed to get the message to folks not only to stay off the road but to get out of low-lying basement apartments.  That did not happen.

Newark Airport

There is no reason why so many people had to die last night from an event we knew was coming and which we could watch unfold with weather radar and surface observations.


  1. Basements are often remodeled into living space without awareness of the need for secondary safe exits in case of fires or floods. While the codes are good on this, enforcement is not able to prevent this sort of remodeling. I think more news coverage and feature stories might be effective.

    1. Most of those basement apartments in that region are not legal. The residents are poor and flying under the radar of authorities, so feel they have no recourse to report shoddy construction and poor conditions. The mayor of NYC even acknowledged that the city knows about them, but does nothing about them, because without those spaces the people living in them would be homeless and therefore the city’s problem.

  2. While I agree it would be great to have a model with the parameters you described, I really wonder how much difference it would actually make in terms of saving lives? Let's face it: Most people aren't weather nerds and pay only casual attention to the weather forecast. Especially in a place like the northeast coast where you only get severe weather once in a while, you could probably scream at the top of your lungs that some disastrous event was about to occur, but because people there rarely need to concern themselves with the weather and don't regularly need to do things like evacuate the way people in, say, the Gulf Coast, frequently have to do, a bunch of weather forecasters screaming at the top of their lungs probably isn't going to get many people's attentions in a place like that and people will largely go about their regular daily lives with little to no adjustment.

    1. My thoughts exactly. So many people do not trust any sort of media, science, or "experts," and will completely blow it off and even go beyond that to call it fake. It's a pitiful state of affairs in the U.S.

  3. Even if the prediction of this event was perfect, the death toll would probably have been only *slightly* less. There will always be morons or machismo that think they can defy the weather(who end up dead) but in this case it was just people caught in flash flooding while undertaking their daily routines which they must undertake. This is almost an inland tsunami caliber amount of water in motion. Many were just stuck in traffic or in their homes when a wall of water hit.

    Its doubtful employers will be sympathetic to those who want to stay home because 10 inches of rain *might* occur in an hour, as per the weather prognosticators....who are notoriously wrong in addition to being broadcast by the media overplaying its hand. It might be a case where less detail to an employer is more, due to that overarching lack of trust in the accuracy of the weather forecast. Its also doubtful politicians will risk an evac that could potentially be for naught, other than a feeding frenzy for lawsuits. Plus, evac to where? There is no plan for this. We are talking about New Jersey, not Florida.

    Businesses are understaffed and politics is already more than toxic enough so its pretty obvious its going to be a dash to the tall grass as far as accountability goes. The Libs will blame ignorance/denial of climate change and the more rugged individualist Ayn Rand GOP types will victim blame individuals for not prepping on a personal level.

    Unfortunately, the Republicans/Libertarians have a more tangible argument. The USA has some of the most extreme weather occurrences per capita on Earth. That needs to be respected and planned for by individual citizens. Some weather events are also akin to being killed by a train. Its obvious where the train is and you know its an easily avoided risk. Still, people get hit and killed by trains all the time. Granted, Ida's flooding might be more like being hit by a Washington State Ferry surfing the crest of a tsunami while in the middle of the Mojave.

    1. You are right....better forecasting is necessary, but not enough.
      We need better communication.
      More proactive governments and warning.
      And improved infrastructure (NYC drainage is only designed for a once in five year storm)

    2. Prefacing this with the blunt truth that any life needlessly lost is a tragedy:

      It can't be understated that given how many people were in harms way of Ida, such a low death toll is a win for climate science and civil engineering. Is there room for improvement? YES! Is hope lost in terms of Humanity versus Climate change? NO!

  4. New York Magazine as of Thursday evening has the total Northeast death toll at 43, and likely to climb. Interestingly reader comments from the locals about this news often decried the lack of the public listening to the well broadcast Weather Service warnings. Add to this the exceptional rainfall from Henri one week before and you would think people would be on alert. The videos of people trying to drive on roadways that look like rivers is incomprehensible.

  5. Not all areas got slammed.....My brother in Bay Ridge, Brooklyn, reports no flooding or damage. All the feral cats in the community backyard are doing fine.

  6. The Major Deegan Expy from Van Courtland South to the Fordham Road turned into a canal completely blocked with stalled tractor tailors. You'd think key arterial roads would have traffic lights to close down the traffic. WSJ now reporting at least 41 dead in the North East from IDA remnants.

  7. Hi Cliff

    You've written a lot recently about how many like to blame climate change for these disasters.

    I found this interview on NPR interesting (from

    "Would a storm like this have been possible to imagine 20 years ago?
    A storm like this would have been exceptionally rare 20 or 50 years ago. But we have to start thinking about it becoming the norm as the climate warms.
    It's very simple physics: As the atmosphere warms, it can hold more moisture — and that means more fuel for rainfall."

    and was curious to get your take on this. If the atmosphere is 2 degrees or whatever warmer now than normal how much additional moisture (or fuel) does that mean the atmosphere can hold? Does that affect the amount of rain that can fall in an event like this?

  8. xlm.... this kind of simplistic "analysis" is problematic. Storms "like this" certainly happened many times before....just in different places. There is a long history of heavy rain associated with transitioning tropical storms. ...cliff

  9. Hi Cliff, just wondering whether the rotation of the earth has anything to do with west coast hurricanes spinning off towards the northwestern Pacific before they reach more northerly latitudes?

  10. What is "normal?" Certainly weather (and climate) are not. They've been changing ever since there has been a planet called Earth.


Please make sure your comments are civil. Name calling and personal attacks are not appropriate.

More Wildfire Misinformation at the Seattle Times

 The Seattle Times continues to shamelessly exaggerate and hype the regional effects of climate change. This week they really went overboard...