WHAT’S IN A NAME: THE DREADED 500-YEAR EVENT
Hurricane season is a wonderfully apt time to discuss the classification of intense storms. It seems like every year we hear over and over how this storm was a 100- or even a 500-year storm.
What does that even mean?
It’s statistics! Yay—everyone’s favorite, right? Unfortunately, scientists don’t have enough data to look back in time and figure out what the worst storm ever in a given region was and then assign categories accordingly. So, as a next-best-thing, they look at the data we do have and then use statistics to determine the probability of a given storm event occurring. The time increments assigned are based on what’s called return period, or the probable time it will take before the given event occurs again. So, for a 500-year storm, the return period would be greater than or equal to 500 years.
Flooding after Hurricane Harvey in Houston, TX in 2017 (image: Mic.com)
Houston, TX, reportedly has had a 500-year event three years in a row! How is that possible? There are a few reasons. First, probabilities are not hard and fast rules. For example, we all know that when we flip a coin we have a 50% chance of getting heads. So once we get heads, does that mean we can’t get it again next time? No, we still have a 50% chance. We could flip the coin 10 times and get heads each time- the probability would still be 50%. Similarly, a 500-year storm has a 1/500 (0.2%) chance of happening each year—so if we have one this year; next year there will be a 0.2% chance that it will happen again, and so on. But seriously. Three in a row?! Well- have you ever had your playlist on random and had a song by the same artist come on three times in a row? Just saying, weird things happen.
Some have suggested changing the way we refer to these storms. For example, a 100-year storm has a 1 in 100, or 1% chance of happening each year, so it might be less confusing to the public if meteorologists and journalists referred to them as 1% storms.
That being said, it’s not all just statistics and semantics. Climate change is real and we are seeing the effects now. As surface waters warm, we really are seeing more frequent, more intense storms. Many regions are using probabilities based on outdated data to classify these storm events. Meaning, what would have been a 500-year storm in the 1980s is now much more probable because the climate is changing. As recent events are included into the data sets, classification schemes will be updated. The mid-western states updated their precipitation frequencies in 2012 for the first time since the 1970s in order to reflect more realistic distributions, and some say it’s time to update them again.
These high-intensity storm events are related to another event we’re hearing about a lot more recently—the statically significant flood. Storms and floods do not track 1 for 1 (i.e., a 100 year storm does not always cause a 100-year flood; they are each calculated based on their own data sets). Flooding is becoming more frequent as well—not only due to increasing rain volume, but also due to strengthening storm surges and rising base sea level. Flooding probabilities and nomenclature are being updated in the same vein as storm event frequencies.
Unfortunately, the only problem that updating these data sets solves is our classification system—our infrastructure is not as easy to update. Cities were built and infrastructure was constructed under far different climactic conditions. Even our more modern systems were typically built to handle the 100-year storm definition as of or prior to the time they were constructed. This is because, historically speaking, it did not make economic sense to plan for an event that had a 0.2% chance of happening. But now, the 500-year event is morphing into the 100-year event, and possibly even more frequent in some regions. As we have discussed previously, environmental planners are not waiting around. Combinations of system upgrades, green infrastructure, and even organized retreat are already being employed in light of our new reality.
At StormSensor, we aim to provide accurate, spatial information for regional technicians and planners to rely on as they update system plans and evaluate upgrades. We have developed machine-learning algorithms and cost-effective, networked sensors that track precipitation, storm flow, sea level, and temperature. Our packages are priced with the goal of making region-wide deployment an actionable possibility.