Was the horror of Harvey properly communicated?
Yes, the intensity forecast for Harvey could have been better earlier in the week prior to landfall. The intensity picture didn’t spin up well until Thursday, when the National Hurricane Center predicted Harvey would come onshore as a major hurricane, which begins at Category 3. Here is the National Hurricane Center/NHC damage scale on the five categories:
The intensity forecast became reasonably accurate about 36-42 hours ahead of final approach to landfall. But the intensity scale pertains only to the types of damage produced by hurricane winds, which are tied to structural damage and storm surge produced by hurricane wind intensity.
When Harvey came onshore, its strongest winds targeted more thinly populated portions of the Texas coast. That held down casualties, and the worst of the storm surge also reached these areas, reducing storm surge damage to coastal cities such as Galveston and Corpus Christi. It was known for several days before the wind intensified that the worst threat from Harvey would be posed by its enormous moisture content. Both the energy for strengthening, and the rate of evaporation from the Gulf were being fed by an exceptionally warm eddy in the northwest Gulf of Mexico, making this strengthening inevitable.
Intensity prediction remains the weakest link in hurricane forecasting, while track prediction has approved substantially in the last decade. The problem with focusing on intensity in the case of a hurricane making landfall is a tendency to breathe sighs of relief when intensity slackens as the storm moves inland. In Harvey’s case, there was general model agreement the weak steering currents aloft would cause the storm to nearly stall once it moved onshore. Obviously, this would mean the incredible amount of water vapor (we refer to that as “precipitable water”) would be able to inundate huge tracts of land for long periods of time. This is what led to forecasts of catastrophic inland flooding well in advance of landfall.
Even by early Thurday evening, National Hurricane Center public advisories were warning of devastating flooding in clear language, later amplified to catastrophic:
“RAINFALL: Harvey is expected to produce total rain accumulations of 15 to 25 inches and isolated maximum amounts of 35 inches over the middle and upper Texas coast through next Wednesday. During the same time period Harvey is expected to produce total rain accumulations of 7 to 15 inches in far south Texas and the Texas Hill Country over through central and southwest Louisiana, with accumulations of up to 7 inches extending into other parts of Texas and the lower Mississippi Valley. Rainfall from Harvey will cause devastating and life-threatening flooding.”
These advisories are issued to the public, as well as emergency managers and other agencies. There are other conference calls that occur between managers and the National Weather Service and, of course, broadcast meteorologists put these statements out on air and in social media, along with their own interpretation and analysis. Still, not all of the public is getting this information, oftentimes by ignoring normal news and weather sources that provide this vital detail.
So, when the storm crawls inland, and the predicted dramatic weakening takes place, psychological guards can fall. There can be a perception of initially riding out the storm successfully, with the dire flood forecasts still in the background in many people’s minds. I don’t know that emergency managers are prone to this type of thought process, but I know many in the public are.
The major problem here is the worst flooding disasters are usually tied to slow-moving, weakened tropical systems. Allison produced a disaster in Houston and other parts of Texas in 2001, causing 41 deaths and leaving 30,000 people homeless. Allison was just such a weakened, slow-moving system. Tying dangers to the Hurricane Category scale misses the mark in cases such as still-unfolding Harvey.
Worse yet, when the storm weakens many in my profession use the phrase: “Hurricane X has been DOWNGRADED to a Tropical Storm.” That term alone creates an impression of diminished hazards presented by the storm. That was never to be in the case in storms like Allison and Harvey. I would propose altering that phrase to “Hurricane X is now a Tropical Storm” with no category/hazard implication in circumstances such as these. If it’s a fast-mover, and doesn’t present such enormous and, in this case, unprecedented flood hazards, I’d be fine with downgrading language. I’m pretty certain I’ve made this judgment error myself, though when I did, it was for another part of the country where my words could not be heard. The one year I worked in Tampa was tropical cyclone-free for west Florida.
As for the rainfall amounts actually forecast, as the storm drew near they were about as spot-on as the science allows. Before landfall, virtually no meteorologist is going to literally predict 50 inches of rain, because that would be unprecedented and would go against climatology. But once we’re talking about 15 to 30 inches of rain and up, catastrophic flooding is inevitable … in ANY city. This reminds me of Gov. Andrew Cuomo’s criticism of the National Weather Service during the "wall of snow storm" of 2014, because the service hadn’t specifically mentioned 7 FEET, which, in fact, only occurred in one or two spots. The Buffalo National Weather Service and the private sector had forecast, days in advance, a paralyzing, nearly stationary band of intense lake snow that would cause major disruption with accumulations of several feet.
In the case of Houston, to be forewarned was not to be forearmed. No city’s infrastructure could begin to handle this unprecedented amount of rain. Houston’s situation was worsened in that it is on mainly wetlands that have been paved over. The swampy land used to hold water. Now, all the paved surfaces do the opposite. They produce runoff and absorb nothing. There is nothing that can be done about this in the short term, but future planning will need to consider this major danger. Part of what is known about the ongoing global warming is the increase in heavy, flooding rain events in middle and southern latitudes in the United States where moisture sources such as the Gulf prevail. More warmth = more evaporation = more precipitable water = heavier rain.
We already have seen this occurring with increasing frequency in the last two decades. That is not to say global warming “caused” Harvey. We had such storms before warming accelerated. But it does suggest the warming aided in producing such exceptionally warm waters in the Gulf so that when a tropical cyclone finally came along, its rainfall potential would be greater than when waters were not quite so warm. Harvey was the first major hurricane to make U.S. landfall since Wilma, in 2005. The hypothesis of warming leading to more hurricanes has not stood up well to peer review. However, there is good evidence warming can make what hurricanes do develop more destructive upon landfall, at the very least in terms of flooding.