The final word on Irma

A tough hit from a seemingly distant tropical system

Jared Smith
Sep 18, 2017 · 12 min read
The corner of Bee Street and Lockwood Drive on the afternoon of September 11, 2017. Photo credit: Erin Spencer

For a storm with its center 300+ miles away, Tropical Storm Irma dealt quite a blow to the Charleston area with heavy rain, waves, tides, and even a tornado.

A week after Irma, life is mostly back to normal. Schools have restarted; kids and parents alike are grumbling about makeup days. Floodwaters have been cleaned up and the beaches are already being renourished after taking a hard hit.

Data continues to be compiled by various government agencies, including the National Weather Service, on Irma’s impact to Charleston. NWS Charleston, SC put together a nice summary of Irma’s impacts on its county warning area (which includes a good bit of Southeast Georgia in addition to the Charleston metro area). I’ll be looking more at local impacts here.

Tides, Flooding, and Erosion

Tides You’ve Never Seen Before…Again

Every year for the last three years, I’ve gotten to use this line:

“Unless you were here for Hugo, you haven’t seen tides like this.”

The first time was with the 2015 flooding event, spurred on by moisture from Hurricane Joaquin well to our east and a sharp upper-level trough digging into the area from the west. During that event, we saw tides in Charleston Harbor top out around 8.16' mean lower low water (October 4, 2015). Later in October, tide levels surged again, this time to 8.65' (October 27, 2015).

With Hurricane Matthew, the tide topped out at 9.28'. (Had the 6' of storm surge Matthew generated coincided with high tide, it would likely have challenged Hugo’s all-time record of 12.52'.)

We weren’t going to get off so easy with Irma’s surge this time. While it was lower than Matthew’s at 4.87', it coincided with a high tide that was already influenced by a recent full moon. Plus, stiff northeast winds had helped to further pile on water, with waves from Irma battering the area almost for a whole week prior.

There were many concerns in the few days leading up to Monday that Irma would be a top-three tidal flooding event. Surge guidance, at one point, had alluded to a 13' tide in the harbor midday Monday, with two major flooding events during the previous high tides. Over time as the model initialized better (it is driven by the GFS), these numbers trended downward.

Thankfully, the two major flooding events during previous high tides did not verify. There was still salt water flooding, though, and there was still the matter of the upcoming high tide and storm surge.

Monday’s High Tide

Monday’s high tide was extremely impressive. Only Hugo and the 1940 hurricane outdid Irma, whose tide (pending verification) stands as the third-highest all-time at the harbor. Much of downtown Charleston went underwater for hours.

A kayaker south of Broad during Irma, right outside a front door. Photo credit: Kaley Briesmaster

The biggest local story from Irma was, by far, the beloved Folly Boat floating away from where it was deposited by Hurricane Hugo, crashing onto a dock. Not even Matthew moved the boat!

The Beaches

The beaches took an absolute pounding from Irma. Waves crashed over the rocks at Folly, and Edisto Beach was pushed back over Palmetto Boulevard again. Isle of Palms has already resumed work to restore the sand dunes to protect the island from further inundation.

Inland Impacts

These impacts were felt well inland, too, especially in West Ashley. Snap Map was lit up with lots of flooding footage from the Castlewood area, and for the third year in a row, water entered Crosstowne Church.

Flooding reports even came in from as far away as Huger, where rising waters forced evacuations.


Rainfall totals from Tropical Storm Irma from September 10–12 for the National Weather Service, Charleston, SC county warning area. Source: NWS Charleston, SC

Overall, Irma was not quite the rain-maker that we saw during the 2015 event or Matthew. However, there was plenty of heavy rainfall — that, combined with the exceptionally high tide, helped to exacerbate the flooding situation across much of the Charleston area. This situation necessitated the rare “Flash Flood Emergency” for downtown, verbiage that hasn’t been employed since the 2015 event.

Storm Total Rainfall

  • 8.97" of rain at a weather station off Bees Ferry Road in West Ashley
  • 7.28" of rain at a cooperative observer’s gauge in Summerville
  • 6.44" of rain at Mt. Pleasant
  • 6.30" of rain at the NWS office in North Charleston


Winds were not the major story of Irma, but they certainly were a factor, with gusts near hurricane force at the beaches and numerous reports of downed trees.

Peak Gusts

Peak gusts from Irma at airport stations:

  • 60 MPH at the airport
  • 59 MPH at Charleston Executive Airport on Johns Island
  • 43 MPH at Mt. Pleasant Regional Airport

Beach-based stations set up by WeatherFlow recorded even higher gusts:

  • 72 MPH at Folly Beach
  • 68 MPH at Isle of Palms
  • 68 MPH at Sullivan’s Island


Supercells with possible waterspouts line up off the Carolina coast, taking aim on the coastline. Image: GR2Analyst

Tornadoes were a major complicating factor throughout Irma’s approach on the US, especially over Florida. Meteorologists watched nervously as Irma’s primary feeder band, embedded with numerous rotating storms, began to move northward into the Charleston area. Tornado watches were issued early in the day, and by noon, tornado warnings began being issued for land areas.

For information on all tornadoes, please see the Public Information Statements issued by the National Weather Service in Charleston.

The first tornado report was a waterspout on Isle of Palms, captured by ABC News 4 as it approached the coast. No word on whether it made landfall; NWS surveys seem to show that was not the case.

Overall, four tornadoes were confirmed in the Charleston area during Tropical Storm Irma. Only one of these was a result of the strongest initial feeder band; three others touched down later in the evening. Most of the tornadoes were of the EF-0 variety, causing minor damage, but one on Johns Island was rated at EF-1 with winds over 100 MPH. None of the tornadoes were particularly long-lived — a hallmark of tropical cyclone-spawned twisters. This made for rather difficult detection — in fact, only the Joint Base Charleston tornado received a formal tornado warning, further underscoring the importance of being sheltered during a tropical system as tornadoes during these events are extremely difficult to warn on.

Path map for the four confirmed tornadoes from Tropical Storm Irma in the Charleston, SC metro area. Source: National Weather Service, Charleston, SC

Joint Base Charleston: 2:48 PM

KCLX base reflectivity (left) and storm-relative velocity (right) depicting the Joint Base Charleston tornado at 2:48 PM.

On radar, the Joint Base Charleston tornado was the most “slam-dunk” for storms requiring warnings. There were reports that as the storm came ashore, a wind shift was noted at the harbor. GR2Analyst “normalized rotation” values were well over 1, lending to high confidence that a tornado was likely. NWS warned the storm around 2:37pm with the estimated touchdown around 2:48 PM.

This storm went on to produce an EF-0 tornado with damage to the control tower and tree damage on a nearby golf course. Nobody was hurt. Here’s NWS’s assessment:

The weak, short-lived tornado, associated with the outer rainbands of Tropical Storm Irma, touched down near the Joint Base Charleston flight line and traveled along a discontinuous path toward the north-northwest. The tornado pushed back the edge of the roof on the control tower, damaged facia of a nearby building near the flight line, and damaged the roof of another metal building on the base. Farther northwest on the Wrenwoods Golf Course, the tornado blew down two large oak trees and snapped large branches off other trees. The tornado then dissipated just beyond the north side of the golf course, about 2 minutes after touchdown.

Johns Island: 5:45 PM

KCLX base reflectivity (left) and storm-relative velocity (right) depicting rotation associated with the Johns Island tornado at 5:45 PM.

The most damaging tornado of Irma was also extremely difficult to detect and very short-lived — only lasting about two minutes. In those two minutes, though, some significant damage was done to Legare Farms on Johns Island, including a home pushed off pilings, as seen in a tweet from Dave Williams, chief meteorologist at WCIV.

The tornado was rated as EF-1 with top winds of 107 MPH. It was on the ground for around two minutes and just shy of half a mile. From NWS:

The tornado touched down along the west bank of the Stono River near the junction of Abbapoola Creek. The tornado initially traveled north northwest near the Stono River, destroying a shed, snapping a large live oak tree at the trunk, and damaging several other trees. In this area, the tornado also pulled from the ground a well pump and cement anchors for a chain link fence. The tornado then knocked a home off pilings; the displaced residence incurred significant roof damage and crushed a pickup truck which was parked under the damaged house. The tornado then turned toward the north and northeast, damaging multiple trees before lifting along the west bank of the Stono River. The tornado was on the ground for less than 2 minutes.

James Island: 6:19 PM

KCLX base reflectivity (left) and storm-relative velocity (right) of the suspect area of rotation associated with the James Island tornado.

The tornado on James Island was very short-lived. Despite its weak nature (an EF-0 with max estimated winds of 83 MPH), it still managed to uproot trees, do roof damage, and loft a little debris for the short time it was on the ground. From NWS:

The weak, short-lived tornado, associated with an outer rainband of Tropical Storm Irma, touched down in a marsh north of Schooner Creek and tracked north-northwest. The tornado uprooted a live oak tree and damaged another tree at the edge of the marsh. The tornado then damaged a residence on the southeast side of Seaward Drive, blowing off part of the southeast and northwest facing roof. The tornado then crossed Seaward Drive, damaging another tree and blowing 4 shutters off the southeast facing side of a residence on Lynne Avenue. Debris carried from the Seaward Drive residence also damaged a soffit on the Lynne Avenue home. The tornado then lifted south of Parrot Point Creek, less than 2 minutes after touchdown.

Oakhaven: 7:13 PM

A little shy of 7PM, there was a subtle rotating feature noted on KCLX WSR-88D velocity data heading NNW toward Isle of Palms and Sullivan’s Island. This rotating feature lasted for a few scans before becoming fairly ill-defined in the data. The storm that produced this rotating feature moved ashore with very heavy rain and gusty winds. No tornado warning was issued.

An hour later, the reports started trickling in.

I’m naturally skeptical of tornado reports on social media without photos, convincing radar evidence, and in situations where storms would be very rain-wrapped. This isn’t a referendum on anybody sending in the reports, but rather a recognition that in the heat of the moment, some of these features can be mis-identified. (I didn’t realize how often this happens until I took a Skywarn training course several years ago.) At first, there were some questions over whether there was a microburst.

But then the tornado reports started to gain a little more steam. Photos floated around private Facebook groups, and one of the thresholds I look for — multiple, independent reports — began to be met.

I referred these reports to the National Weather Service, who ended up surveying the Oakhaven area as part of their post-Irma assessment. On Thursday, NWS confirmed an EF-0 tornado had occurred in Oakhaven.

A weak, short lived tornado, associated with an outer rainband of Tropical Storm Irma, formed over the marsh area between Mount Pleasant and Sullivan's Island. The weak tornado first moved into the south end of Pine Island View Road, then moved quickly north-northwest over the eastern ends of Oak Landing Road and Green Path Lane, before dissipating back on Pine Island View Road. The time of the tornado was estimated using eye witness reports and radar data. Although no significant structural damage was observed, there were many large tree limbs down, some uprooted trees, and one wooden fence completely blown down. The tornado was estimated to have lasted only about two minutes before dissipating.

On Friday, Twitter user Steve Suggs tweeted a video of the genesis of the EF-0 tornado:

The Oakhaven tornado was a prime example of a short-lived, highly difficult-to-detect tornado from a tropical system. Looking back at the radar data, one would note a couple locations of broad, weak rotation in the general area.

KCLX storm-relative velocity data at 7:11 PM, September 11, 2017.

The area of rotation over Isle of Palms is broad and not terribly much to write home about, but as the video above shows, perhaps it may have been enough.

KCLX storm-relative velocity at 7:13 PM, September 11, 2017. A tornado was reported in the Oakhaven subdivision at this time.

Two minutes later at 7:13 PM, the tornado was estimated to have begun its path through Oakhaven. It wasn’t around for long — only two minutes — and it was rather small, with a width of only around 90 yards. Weak rotation continued to be indicated around the Isle of Palms connector, two miles east of the tornado report. Beyond this time, any hint of rotation had weakened on radar.

One thing to keep in mind is that the further from the radar site, the higher up the radar beam sees into the storm. Over Isle of Palms and parts of Mt. Pleasant, the radar is seeing around 5,000 to 5,500 feet up into the storm. With a highly marginal signature on radar and no real-time ground truth, this is a nightmare scenario for any warning forecaster. It’s a big reason why the weather enterprise emphasizes the importance of being ready for a tornado at any time within a tornado watch area, especially with tropical systems that spin up and spin down just as quickly. Thankfully, nobody was injured and there was no significant property damage.

As of this writing, there have been no additional tornadoes confirmed in the Charleston metro from Irma.


  • Watch the cone of uncertainty, but know that it’s not the end-all, be-all. There was a lot of emphasis on Irma’s center track. By Friday, when it became clear the storm’s center would stay west and the wind field would expand, there was still a lot of public emphasis on the cone. Many felt as if the #chswx enterprise was hyping up the storm — in fact, one Post & Courier commenter claimed I was milking my “15 minutes of fame.” It seems clear that was not the case. There is a big reason the Hurricane Center put the wind field on the cone maps this year, and if I may extend to them another suggestion: Project the forecast wind field onto the projected path. Decouple the emphasis on the center line and the probability cone and really make it about impacts away from the center.
  • Deterministic model runs are not forecasts. I got a lot of questions about what the Euro was going to do during this event. It did a good job handling a lot of the track, but that won’t always be the case. NHC forecasters work hard to put together a forecast using a blend of models and, most importantly, their observations and expertise. We should all use that data to make our decisions and not trust one outlier model run.
  • Surge has impacts on inland communities, too. Irma was a good reminder that anyone in the tidal zone is vulnerable to surge. There were a lot of reports of people feeling surge well up the Ashley River. Now imagine if this was a direct hit.
  • Charleston’s problems with flooding are only getting worse, and how we are building is not helping matters. We’ve got to stop creating our own disasters by filling in wetlands for new construction. There needs to be serious discussion about this at all levels of government, and not a moment too soon.

Thank You

Finally, thanks to all of you for being such an engaged community during the event. Your photos and reports were invaluable and provided the #chswx enterprise, from broadcasters to private meteorologists to the National Weather Service, invaluable information as we worked to keep the public informed and warned. It’s safe to say that @chswx just doesn’t work without your input, period.

A special thanks goes to those of you who signed up on Patreon or sent in donations. Every little bit helps continue to further my mission of providing hype-free, straight-to-the-point weather information to all of you. Your patronage has already helped expand the amount of data I have access to, which only helps me provide more informed coverage.

Now, let’s hope future storms stay out to sea. I only really have one of these in me every year. :)

Follow Jared Smith’s Charleston Weather updates on Twitter and Facebook. An updated forecast and current conditions are always available at, too. Bookmark it on your home screen! To support independent, hype-averse weather journalism, consider becoming a patron on the Charleston Weather Patreon page.

Charleston Weather

Weather forecasts and discussions for the Charleston, SC Tri-County area (Berkeley, Charleston, and Dorchester).

Jared Smith

Written by

Software engineer and weather geek. Dev team manager @BoomTownROI, principal nerd @chswx.

Charleston Weather

Weather forecasts and discussions for the Charleston, SC Tri-County area (Berkeley, Charleston, and Dorchester).