In the past decade, North Carolina has been smashed by massive rainfalls associated with tropical systems passing over and near our state. During this time, I have fielded many questions about flooding and urbanization. Here are some answers to FAQ’s.

What is a 100-year storm?

This is a rain event that has a 1% chance of occurring in a given year. Thus, over a long-term average, it should occur once every 100 years.

And a 10-year storm?

This rain event has a 10% chance of occurring in a given year. Thus, once (on average) every ten years. This relationship holds for 2-year (50% chance in a given year), 5-year (20% chance in a given year), 50-year (2% chance in a given year), and so forth.

Over what time duration do these storms fall? One day?

The most common “large” storm duration is considered to be one day, but — for example — some cities were impacted by Hurricane Florence for more than three days. The rain associated with Hurricane Helene also fell in WNC for two to three days as the storm moved northward from Florida. There are predictive statistics for durations as short as 5 minutes through events lasting up to a week. One day is used because it was convenient (due to daily data availability) when this decision was made decades ago.

How large a storm was associated with Hurricane Helene in 2024?

The answer depends on where in the mountains you were located. The largest airport in the region (Asheville Regional, AVL) recorded nearly 14 inches of rain in three days. To put this in perspective, the 1000-year storm event was predicted to be 11.3 inches. This suggests that AVL airport may have received a 10,000-year storm! However, the maximum recorded depth of rain that fell at a state-verified station occurred at Mount Mitchell, which received 24 inches of rain! Truly, off the charts!

How large a storm was Hurricane Florence in 2018?

The “other end” of the state has witnessed multiple extreme events during the past decade, highlighted by Hurricane Florence. For example, in Wilmington, it rained approximately 26 inches over a 3-day span. That is roughly a 2,000-year storm. So, a chance of occurring in a given year of 0.05%). In Elizabethtown, nearly three feet of rain fell. That likelihood might be once in 500,000 years. At Raleigh-Durham airport, rainfall totals were less (8.94 inches), but still impressive (approximately a 1 in 50 year event).

How large a storm was Hurricane Matthew in 2016?

As above, it depended upon the location, but Elizabethtown received between 15 and 19 inches within a 2-day period. In case you are curious, that’s between a 500-year storm and a 2,000-year storm.

Wait… Elizabethtown had up to a 2,000-year storm and then an approximately 500,000-year storm in a two-year period?

Yes.

And what are the chances of that happening?

Not high. (A rough estimate is 0.00001%.)

Do you think these extreme events are getting bigger/ worse?

While I can’t speak for all of recorded history, we do have some daily rainfall records since the 1950s that show a trend of increasing “big storms.” See the figure below for Raleigh-Durham airport. You’ll observe that this record stops in 2023 and therefore doesn’t include another 6-inch event (associated with Hurricane Debby).

What is further interesting is that RDU experienced a rather quiet period (from a max daily rainfall perspective) from the 1970’s through the end of the “aughts”. The 1950’s into the 60’s were, however, a period of more extreme rain as well. Still by any measure, maximum rainfall depths certainly appear to be increasing in the past 10-15 years. The dotted line shows the upward trend of maximum daily precipitation at RDU.

Who manages the rainfall statistics and from where can local precipitation data be obtained?

Rainfall statistics are the purview of the National Climactic Data Center, a part of the National Oceanic and Atmospheric Administration (NOAA). Rainfall data are available in North Carolina from the NC Climate Office.

Should I even trust these statistics?

Well, they are the best we have, and they are currently being revised. When most of these predictive statistics were developed, the heaviest rains from the past decade had not fallen. Moreover, most rainfall records are (at best) 70 years old. So, while we have pretty good predictability for smaller (e.g., 2-year or 5-year) events, the predictability for the largest, most extreme storms is simply not as good because don’t have enough data. I mean, it’s hard to be sure what a 500-year storm is when we’re basing that decision on a 50- to 70-year precipitation record!

Seeing pictures of the devastation in both western and southeastern NC makes me wonder, has flooding gotten worse?

Let’s answer that by “breaking down” flooding into various factors: the extent of flooding is dependent on the depth of precipitation, the land cover upon which the rain falls, backwater caused by creek obstructions, and to some extent, slope of the surrounding land. The human impacts felt due to flooding are dependent on the number of people living or working in flooded areas and also our infrastructure. So, while the historic 1916 flood at Marshall (a city in WNC) may have resulted from similar amounts of precipitation as that associated with Hurricane Helene, the number of people impacted by the recent event is orders of magnitude higher than in 1916.

Does urbanization make flooding worse?

Yes. The extent that urbanization exacerbates flooding, however, is dependent on the size of the storm. Urban land use impacts are most felt during smaller flooding events, such as those caused by 3 to 5 inches of rainfall. This is because before land was urbanized (e.g., it was a pasture or a forest), typically 1 to 2 inches of rain would need to fall before soils got sufficiently saturated to produce runoff. That is, it needed to rain up to 2 inches before much runoff would occur. The impervious surfaces (roads, roofs, driveways, compacted lawns) of urbanization dramatically reduce the depth of rain needed before runoff starts. Perhaps to as low as 0.10 inch. Thus, during a 3-inch rain, a typical commercial site might produce 2.8 inches depth of runoff, while a forest might only yield 1-inch of runoff. That’s an increase of nearly 200%!

And for larger storms?

Once a soil is saturated, it’s saturated. So, if it rains 10 inches, the same commercial site is going to yield 9.5-9.8 inches of runoff, while the forest’s yield is 8 inches of runoff. That’s roughly a 20% increase. While not insignificant, urbanization’s impact is more marginal during these extreme events.

Are there other human impacts with regard to flooding?

Yes! Our transportation infrastructure can have an effect. Obviously, we need roads, and sometimes these roads need to cross creeks and rivers. So, we build bridges and culverts. When large storms strike, not only is water and mud/soil transported, but so are limbs, branches, trees, and other large debris. Large debris is able to collect on the upstream face of the bridge or culvert and then slowly build up. Think: first, a displaced tree is “pinned” by the flow on the face of the bridge, and then it snags smaller branches and trash. Before long, the bridge (or culvert) is partly acting like a dam! This then means that water will build up higher along the bridge, thereby backing up downstream until the bridge or culvert/road is possibly breached. In this way, bridges and culverts can exacerbate local flooding.

What’s a culvert?

A culvert is a pipe that goes under a road through which water flows. The pipe (usually concrete or metal) is placed along the creek bed and then soil is placed and compacted around and above the pipe. The roadway will cross on top of the culvert. Culverts can be constructed in a wide range of sizes, from as small as 1 foot to exceeding 10 feet in diameter, and are cheaper than bridges and are used when crossing smaller streams and creeks.

Why did so many towns in WNC flood, whereas flooding in ENC tended to impact town peripheries?

In a word: topography. In WNC there is limited flat land. Flat land is typically the easiest to develop and has been so throughout history. In the mountains flat land is almost always near creek or river valleys. This flat land is where communities historically formed in WNC. In ENC, most of the land is flat! So early settlers tended to establish towns on relatively high land, that was still often close to a major water source.

Why is flat land associated with river valleys in the mountains?

Over the eons, before people populated the planet, it has rained, and rivers have flowed. And there were floods. Floods carry not just water, but eroded soils from the watershed above. You can still see this today by looking at pictures of brown flood water! When rivers overtop their banks and spill out onto the flood plain (the relatively flat land near the stream or river), floodwater starts moving much more slowly and thus doesn’t have the energy to still carry all those soils. Thus, the soil particles fall from the water column, onto the land. Imagine this happening repeatedly for hundreds of thousands of years. That is how flood plains formed and continue to form today. Essentially, the soil of the flat land (flood plain) is that which used to reside somewhere on one of those rocky slopes.

Is there a difference between flood plains and floodways? I’ve heard both terms used.

Yes, there is. Floodways are part of flood plains, but not vice versa. Floodways are where water actively flows downhill during large storms. This includes the creek channel and some part of the land that is adjacent to the stream. Floodplains include the floodway and then also land that is further uphill but still subject to being flooded. Development is restricted in the floodplain and even more so in floodway.

Are flood plains in ENC and WNC different? Because I observe more infrastructure damage in WNC flood plains.

Well, they are both areas that flood, so that’s the same. However, floodplains in eastern North Carolina are often much wider than those of WNC. Flood plains for creeks and streams in the mountains are much narrower than those of ENC because they are surrounded by steep hills and mountains. This poses unique challenges in WNC streams. Because the floodplains are narrower, the water flow (for the same size watershed and same rainfall intensity) will necessarily be deeper than that observed in ENC. Moreover, because water has less room to spread out and because the topography is steeper in WNC, runoff down the slopes is faster, and the velocity of water flow within the streams and rivers is likewise much faster. This faster water flow creates much more destructive power. As a result, roads and bridges are more likely to be blown out in WNC than ENC     .

I’ve heard of a 100-year floodplain. What is that?

This is the area that has been predicted to flood during the 100-year storm. Similarly, the 500-year floodplain is what is expected to flood during the 500-year storm.

In Helene, if people had been prevented from living in the 500-year flood plain would that have helped?

Yes, fewer people would have had their homes and businesses flooded out and likely fewer people would have died. However, Helene far exceeded even the 500-year storm in some places, meaning that many people who thought they or their property were (relatively) safe from flood waters would have been caught in this flood. The same is true for parts of ENC during Hurricanes Florence, Debby, and Matthew.

Could the construction of Stormwater Control Measures (SCMs) have helped?

Stormwater Control Measures (SCMs) are constructed to improve water quality and limit flooding. The most common large-scale SCM is the wet detention pond. Whether they can help limit flooding… well it depends on the storm size. SCMs are specifically designed to improve water quality from 1 to 1.5 inch rain events; compare that to the rainfall depths of Helene, Florence, and Matthew… During these hurricanes, quite simply, very little to no positive impact would have been gained by employing SCMs. If, however, the potential cause of a flood is a 3- or 4-inch storm, then a group of SCMs can certainly help if a sufficient amount of watershed is captured and treated by them.

Could we design SCMs to be more impactful during more extreme events?

In theory, yes, but only in some cases. For example, several SCMs were installed near the banks of the French Broad River in Asheville, NC. Those practices were underwater during Helene. There was nothing a designer could have done to make those “work better.” When SCMs are located on properties above a floodplain, they could be designed to hold water from events larger than the 10-year storm, but doing so necessarily increases their footprint and, therefore, their cost.  And even then, the likelihood that any SCM could improve the situation experienced in Florence and Helene is simply unrealistic.

So, you’re saying that there is very little we can do to engineer a solution to fix the problems caused by storms the magnitude of Helene and Florence?

Unfortunately, yes. Experts are working to develop recommendations to plan for future storm events, but ultimately, if there are people and infrastructure located within floodplains the likelihood of future destructive flooding events will remain high.

Helpful Websites

Weather Data

NOAA Atlas 14 Storm Frequency Predictions

NC Climate Office (NC weather data repository)

Flood Plains

FEMA Flood Maps

Stormwater Control Measures

NC DEQ Stormwater Design Manual

NC State BAE Stormwater Engineering Group