A week ago we argued that a quiet hurricane forecast is not a safe one, because seasonal outlooks predict basin activity and not whether a storm reaches your address. We did not expect the point to be demonstrated quite so promptly.
Where the season stands
As of today, the 2026 Atlantic season has not produced a single hurricane. Five tropical storms have formed; none has reached the 74-mph threshold.
That is genuinely unusual. In the roughly sixty-year satellite era, the average date of a season's first hurricane is July 27. The latest on record is September 11, a distinction shared by 2002 and 2013. We are three days from that mark. If the Atlantic stays quiet through the end of this week, 2026 sets a record for the longest wait.
The reason is the El Niño that has been strengthening through the season. El Niño tends to increase wind shear over the tropical Atlantic and encourage dry, sinking air — conditions that pull developing storms apart before they can organize. NOAA's below-normal seasonal outlook called for 7 to 13 named storms with 2 to 6 hurricanes, and so far the storm count is tracking while the hurricane count is not.
And yet
On September 1, Tropical Storm Edouard came ashore near Johnson Bayou, Louisiana, at 2:20 p.m. local time with maximum sustained winds around 60 mph — fourteen miles an hour short of being a hurricane at all.
What it did anyway:
- Rainfall of 3 to 6 inches across the upper Texas coast and east-central Texas, with localized totals approaching 9 inches north of Houston
- Flood warnings across southwestern Louisiana, including Lake Charles, and much of the upper Texas coast
- More than 150,000 customers without power
- Storm surge of 1 to 2 feet above expected tide levels along the southwest Louisiana coast
Nobody will remember 2026 as a bad hurricane season. The people who spent that week pumping out a house will remember it fine.
Why the category system misleads
The Saffir-Simpson scale — the familiar Category 1 through 5 — measures wind speed only. It says nothing about rainfall, storm surge, or how slowly a system moves. That is a real limitation, because in recent decades the majority of US tropical-cyclone fatalities have come from water rather than wind.
This produces a predictable failure. "Only a tropical storm" gets heard as "not a threat," and people make decisions accordingly: staying put, driving through water, skipping preparations they would have made for a named hurricane. Edouard was, on the wind scale, unremarkable. Its rainfall was not.
The forecast language has been shifting toward water impacts for exactly this reason, and it is worth internalizing before the next storm: ask what it will drop, not just what it will blow.
The wider frame
Three things are true at once here, and the honest version keeps all three:
- The seasonal forecast is performing well. A quiet hurricane count in a strengthening El Niño is what the science predicted. This is a forecasting success, not a failure.
- Individual risk was never captured by that number. A single landfall is what matters at one address, and it arrived regardless.
- Warming shifts the risk toward water. A warmer atmosphere holds more moisture, so storms that do form are favored to be wetter — which makes wind-based categories a progressively worse proxy for danger. That connects to the record ocean heat we reported this summer, and to the drought and record US heat we wrote about last week, where dried-out ground sheds heavy rain rather than absorbing it. You can follow the indicators on our data dashboard.
For the rest of the season
Peak activity runs from mid-September into October, and El Niño's suppression is a tendency rather than a guarantee. The practical guidance has not changed since last week, and Edouard is the argument for it: know your evacuation zone, plan for water rather than wind, and treat "only a tropical storm" as a description of wind speed rather than a forecast of your week.
