Back on Aug. 29, the rain came in buckets and an east-northeasterly wind rattled windows for hours. So the high tide that afternoon did what anyone who'd stuck a finger in the air would have expected: It rose and rose.
In the midst of that unnamed storm dashing up the coast, the gauge at Sewells Point in Norfolk measured a peak water level deep into the minor flooding stage. Lots of low-lying streets in Hampton Roads were swamped.
It was easy to explain.
But three weeks later, five straight flooding high tides caused seasoned weather watchers to scratch their heads. There were gusts in that Sept. 18-20 period, yet not like on Aug. 29, and barely a quarter of an inch fell over those three days at Norfolk International Airport, compared with nearly 4 inches on that one blustery day in August. At times during those September days, the weather was downright pleasant even as many streets were impassible.
So what riled the tides? Several things, some of which are as basic as where in the heavens the moon was then.
More important than the explanation, some scientists say, are what those flooding tides last month may foretell. As sea levels keep rising, they warn, Hampton Roads is likely to see more such prolonged periods of high water when there's no immediately obvious reason.
The region started out more prone to flooding tides that week in September than in late August, simply because of astronomical conditions. A new moon was to arrive Sept. 20, and in that phase its gravitational pull is stronger, driving tides higher. Because of that and other factors, the long-term tide charts predicted water levels would begin more than half a foot higher on average Sept. 18-20 than they would have on Aug. 29, before weather effects were taken into account.
Then came a hurricane named Jose.
If it had taken a different path, Jose might now be cursed in the same breath as Harvey, Irma and Maria among this year’s storms. It peaked as a Category 4 hurricane with winds of 155 mph, but skirted the Caribbean and finally died, after an odd, twisting loop, well off New England as a post-tropical cyclone. It never made landfall anywhere.
Jose did, however, make an impact – the only significant impact so far among 2017's big storms – in Virginia’s coastal region. Though it was just barely a hurricane when it crept hundreds of miles from shore in mid-September, it stirred long, powerful waves that pushed extra water into the Chesapeake Bay over several days.
Jose’s long-distance punch was a classic example of what makes autumn the peak period for flooding tides in Hampton Roads. In the hurricane season that officially ends Nov. 30, even dwindling storms far off the coast can drive the water higher.
And because of sea level rise, the effects of storms like Jose are being exacerbated. It doesn't take a hurricane like Matthew, which pummeled Hampton Roads a year ago, to cause problems. Disturbances way out in the Atlantic that once might have been waved off as inconsequential locally can now cause tidal flooding.
That appears to be the case with Jose. The peak tide it induced at Sewells Point, roughly 5.23 feet above a reading called “mean lower low water,” was about 0.73 feet above what the National Oceanic and Atmospheric Administration considers minor flood stage there. Subtract the more than 0.75 feet of sea level rise over the past half-century, however, and that tide wouldn’t have crossed the flooding threshold, even with the lunar cycle approaching a phase that nudges tides higher.
There’s no doubt that the region is experiencing “much more flooding and much more damage from flooding just because of sea level rise,” said Tal Ezer, a professor of ocean, earth and atmospheric sciences at Old Dominion University.
Ezer has studied data from tide gauges across coastal Virginia and has done some of his most extensive research using data gathered at Sewells Point. That gauge is often used as a proxy for the effects of high tides, including those with the potential for flooding, across the entire region.
For Sewells Point, Ezer has calculated the number of hours per year that the water level is above various thresholds. Even for the most extreme scenario he considered – 2 feet above a taller measure called “mean higher high water” – he has found an overall increase in recent decades.
There was only one year each decade from the 1950s through the 1980s in which the water hovered for 10 hours or more above that level, his research showed. There were five such years in the 1990s, five in the period from 2000 to 2009, and six more in the seven years from 2010 through 2016.
For 2016, Ezer calculated more than 35 hours at 2 feet or more above mean higher high water at Sewells Point. During those hours, the water would have been above the minor flooding threshold.
Ezer and an ODU oceanography colleague, Larry Atkinson, have theorized that slowdowns in the Gulf Stream are a factor in some flooding tides across the mid-Atlantic, with the effect particularly noticeable at times when storms disrupt the current's flow. The Gulf Stream, because it's so powerful, is key to local sea levels, Ezer said, because it pulls water away as it heads farther into the ocean around Cape Hatteras.
Larry, a senior forecaster in the Wakefield office of the National Weather Service, has been documenting tidal flooding in Hampton Roads through a different measure. He counts each high tide that hits at least minor flood stage at Sewells Point, and has found that they averaged 2.8 per year in the 1980s, 6.7 per year in the 1990s, 6.6 per year in the 2000s and 11 per year from 2010 through 2016.
Last year, the Sewells Point gauge crossed the flooding threshold 18 times, Brown found. That tied it for the second-highest tally, behind only 2009, a particularly busy year for storms.
Brown also has broken down the flooding tides by month. The data show – likely not surprising to experienced local weather watchers – that they tend to peak in the fall.
Since 2000, October has averaged 2.6 flooding tides at Sewells Point, the most of any month, he calculated. September came in second at an average of 1.9, followed by November at 1.2.
Hurricane season largely explains the autumn peak.
Brown said storms that might seem initially to present no threat to Hampton Roads have to be watched closer than just a few decades ago because of sea level rise.
"A tropical system a thousand miles away, if we start getting swells from it for a couple of days in a row, water will start to pile up in the bay," he said. "That can cause issues even if the weather doesn't seem like it's bad here."
He said some storms that caused only minor flooding in Hampton Roads a few generations back would now bring what's considered moderate flooding.
"It's not really that these storms are getting stronger," Brown said. "It's just the water level is starting out a half-foot higher than it was 30 or 40 years ago." That increase in sea level rise, exacerbated because the coastal land is sinking, is one of the fastest in the country.
Amid the growing threat, the weather service in recent years has expanded the number of tide gauges in Virginia and elsewhere in the mid-Atlantic that have 72-hour flood forecasts. Jeff Orrock, the meteorologist in charge of the Wakefield office, said forecasters start with the astronomical tide predictions, then add or subtract from that an "anomaly" that's based on water levels over the past five to 30 days.
At this time of the year in the bay region, they tend to add more often than not when factoring in that anomaly, and recently have tacked on as much as a foot of water to the baseline astronomical high-tide forecasts.
The final ingredient in how high a tide is ultimately predicted to crest is the expected weather around that time – particularly the speed, direction and duration of winds.
The bay's numerous tributaries are tricky to forecast, partly because because their openings face in every direction, in some cases opposite each other within the same city or county. Their varying length, width and depth also come into play.
Shallow, narrow waterways can more quickly be overrun by tides driven up by the wind. Then again, a strong enough wind from the opposite direction can contribute in that same waterway to a blowout tide, when the water drops dramatically low. They've occurred locally in some places in recent weeks, but are much rarer than flooding tides.
Hurricane Maria, the most recent big storm to land on the local tide forecasters' radar, was officially predicted to have contributed to a single flooding tide at Sewells Point, at 4.8 feet above mean lower low water. At one point, some forecast models had the tide pushing 6 feet there on Sept. 27, which would have been the highest level of the year for that station.
It ended up peaking at just below the minor flooding threshold.
"Every event's different," Orrock said.
This year's highest astronomical tide – again, the tide that's predicted without any weather effects factored in – is due in Hampton Roads on the morning of Nov. 5. At Sewells Point, the forecast is for 3.41 feet above mean lower low water.
John Boon, an emeritus professor at the Virginia Institute of Marine Science, said the convergence of two cycles around that time – a full moon and a lunar perigee, when the orbiting moon is due to make one of the year's closest approaches – will set the table for higher-than-average tides.
That still may not translate into much flooding Nov. 5. Other tidal "building blocks," like sustained, strong winds from the ocean, would have to come into play. If flooding does occur, scientists like Boon say, sea level rise also will have played a part.
Boon helped create VIMS' Tidewatch system, which gives predictions and recent tidal records for 10 stations in Virginia and Maryland. He said he's found himself being asked more often in recent years why "the tides seem higher than normal."
It's pretty basic, he said: "I would generally say it has to do with what our current sea levels are." And that they've been steadily rising.
Mayfield writes about the environment for The Virginian-Pilot.