Why was Ida so catastrophic…far from the landfall point? Several factors came together.
On September 1, 2021, a devastating series of extreme flash flooding occurred over the Northeast, stretching from western Maryland through southeastern Pennsylvania, New Jersey, southern New York (including New York City), Connecticut, Rhode Island and Massachusetts. A large swath of rainfall totals of 6 to 11 inches (locally higher) resulted in widespread flash flooding and river flooding. In New York City at Central Park, which has reliable weather data since 1869, it was the most rainfall ever recorded in such a short amount of time. Numerous tornadoes also touched down, including one in southern New Jersey that was rated EF-3 (the second such tornado in the Philadelphia region this year; see also my article from July 29 on such threats in the Northeast). Sadly, at least 46 people lost their lives in the Northeast alone, a number that is likely to continue to rise.
This disaster was certainly related to the remnants of Hurricane Ida, but the entire circumstances are more complicated.
The first feature involved was Hurricane Ida, or what was left of it. Ida made landfall on August 29 in southeastern Louisiana as a category 4 hurricane with 150 mph winds. That resulted in all of the hurricane’s extreme hazards — intense winds, devastating storm surge and very heavy rain. While the levees in New Orleans held, the wind and rain damage and power outages have resulted in major power outages. As it moved inland, heavy rain and winds advanced farther north through the Southeast as it weakened to a tropical storm and later a tropical depression, eventually working its way into the Mid-Atlantic as a post-tropical low. While there was flooding in the inland Southeast, it was nowhere near as intense as in the Northeast (the rainfall totals were generally 2 to 3 inches with locally higher totals)— suggesting that Ida alone could not be attributed.
As the remnants of Ida moved into the Northeast, a front was also advancing eastward out of the Midwest and had moved into Atlantic Canada. That meant that there were two weather systems working in tandem at the same time. In addition, the frontal boundary was relatively stationary throughout the day across the Northeast as it was largely oriented east to west, prolonging the impacts. Stationary fronts are notorious for producing flood disasters when a cyclone moves along it, which was the case here. Along and north of that frontal boundary, heavy rainfall was the story, while just south of the front, tornadoes were a major threat as the wind shear was magnified in the warm, moist air mass.
The front increased lift in the upper atmosphere, enhancing rainfall rates beyond what Ida alone would have provided. That explains why rain was falling at intensities that were virtually unheard-of in the region during the afternoon and evening of September 1. That was also one factor during Hurricane Harvey in southeast Texas in 2017 to magnify its rainfall totals there (although that lasted for a much longer duration).
Three other tropical cyclones or their remnants have also moved through the Northeast over the summer of 2021 — Elsa, Fred and Henri. Combined with other frontal systems, those have led to a very wet summer in the region. Soil moisture levels, particularly in the Tri-State region around New York City, were in the 90–95th percentile for the time of year, almost comparable to 2011 and 2018. (Note that the previous heaviest rain in New York City in a one hour period occurred ahead of Henri, but Ida obliterated that record). That meant that the ground was very saturated and there was little ability for the soil and ground to absorb any of the rainfall that fell during the remnants of Ida. Rainfall was going to end up using the built environment (such as roadways) to flow or stand in the already saturated grass.
Finally, the area that was affected is very urbanized — in fact, it is the most densely populated region in the United States. On top of an already saturated ground, much of the area is built up with concrete, asphalt and other man-made materials. Unlike grass or wetlands, those surfaces are not permeable and do not allow for much of the water to seep into the surface, instead resulting in the rainwater running off down streets, tunnels and into built-up areas. Mitigation of such is required through drainage ditches and sewers, but no system can even come close to managing the rainfall totals that fell as a result.
How well could we have prepared?
Weather forecasts throughout the week suggested that those factors would coalesce for a significant, if not historic, heavy rainfall event. There were warnings throughout the week and many meteorologists were sounding the alarm. Nonetheless, the general response was that it could not have been predicted, or was far worse than expected. Not until the flash flood emergencies were issued (the first ever for many of these areas) was it clear that this was going to be a devastating, if not catastrophic, event. It didn’t hit without warning and weather offices did an excellent job predicting the event.
Some areas, such as central New Jersey, have relatively recent experience with major flooding from Hurricane Floyd in 1999 and Hurricane Irene in 2011, while Pennsylvania has experience from the remnants of Tropical Storm Lee, also in 2011. However, for New York City, it was unprecedented to have such flash flooding. As a result, there was no knowledge on what such could do. Unlike Hurricane Sandy in 2012, the water was from the sky and not from the ocean, so lessons from Sandy would have limited effectiveness. Several people drowned in basement apartments. The public can read the forecasts and hear the alarm, but without any past knowledge, it can be difficult for them to listen — it’s human nature.
Some possible solutions that may exist:
- Strategizing for emergency management. Mapping out rainfall flooding areas may help identify areas that frequently flood and sorting them out by zones, particularly in urban areas. That way, evacuations can take place for them ahead of the weather event. Those would be similar to hurricane evacuation areas except built around rainfall flooding instead of storm surge. In addition, basements need to consider the water table using updated standards — many may have been built in places not properly mapped. If a similar event was forecast, that would allow for targeted evacuations (i.e. basement apartments and low lying areas) and possibly for partial closures of low-lying and underground roads and transit systems. It can also allow for alerting of motorists to stay off the roads and strategic road closures.
- Improving urban drainage systems. That will allow for more water to be pumped out during heavy rainfall events to reduce the impacts or likelihood of major urban flooding. There are limited capacities in terms of rainfall pumping that can take place. While the most extreme events will be difficult to manage no matter what, at least they could help mitigate the threat by reducing the overall impact and preventing moderate events from becoming major events.
- Better floodplain management. Too often, we have built in the floodplains and many mapped flood areas are from decades ago. We see every year somewhere that there is catastrophic flooding, often in places that have never even come close to flooding before. They need to be adjusted for extreme events too. If we leave areas prone to flooding as parkland and have larger flood water management facilities, we can greatly reduce the risk of flooding downstream.
- More permeability in the built environment. We have started to improve our buildings, roads and other man-made structures on a limited basis to help make them absorb water better (and also reduce the urban heat island with grass-covered roofs and more natural surfaces). Accelerating such into urban and suburban areas will help manage future heavy rainfall events which can take many forms — tropical systems, Nor’easters, training thunderstorms and stationary fronts, to name a few. The one serious downside is cost, as such is expensive, and it would likely take decades to adjust.
- Planning for the worst. It is certain that a similar — or more extreme — event will happen in the future. That is a common argument made in the climate change realm too (an issue that goes beyond the scope of this article). Any stalled tropical system has the potential to drop rainfall measured in feet, not inches, with excessive rainfall rates. Continuous thunderstorms can do the same in a blocking pattern or with a stalled front even in places not normally expecting such (including at higher latitudes). As a result, the levels seen from this event should be seen as a starting point, not a worst case scenario. Future infrastructure — buildings, roads, bridges, transit systems, ports, electrical grids and the like — should be built and fortified against events even more extreme than what we saw.
What we saw on September 1 was historic and devastating. Unfortunately, the loss of life was very significant. We can learn from the incredible images and horrible flooding to prevent such from happening again. A 500-year flood does not mean it will take 500 years for the next one to happen. It could happen again in very short order. Floods are one of many hazards we face and rainfall is one of the most common reason. Just as the Northeast learned to fortify after Hurricane Sandy, this should be another learning experience and that planning needs to go to even greater levels.
— Meteorologist Craig Ceecee —
