The 2026 Atlantic hurricane season hasn’t started yet, but one of the country’s longest-running university forecast programs is already sounding an alarm. Researchers at the University of Arizona project 20 named storms, nine hurricanes, and four major hurricanes this year, with an accumulated cyclone energy (ACE) index of 155 units. If those numbers hold, the season would rank well above the 1991-2020 average of roughly 14 named storms, seven hurricanes, and a median ACE near 123, placing 2026 in the company of some of the most punishing years in recent memory.
What the Arizona forecast actually says
The outlook comes from the Department of Hydrology and Atmospheric Sciences at the University of Arizona, which has issued statistical seasonal hurricane forecasts for years. The 2026 prediction table lists 20 named storms, 9 hurricanes, 4 major hurricanes (Category 3 or higher), and an ACE value of 155.
ACE is worth understanding because it captures more than a simple storm count. The metric combines wind speed, storm duration, and frequency into a single number. A season can produce a modest number of storms yet still rack up a high ACE total if those storms are long-lived and intense. Conversely, a season packed with weak, short-lived tropical storms may generate a lower ACE despite a high headline count. At 155, the Arizona team is signaling that 2026 could bring not just many storms but storms that last longer and hit harder than in a typical year.
For perspective, the hyperactive 2020 season produced an ACE of roughly 184, while 2005, the year of Hurricanes Katrina, Rita, and Wilma, generated about 250. The 1991-2020 median sits near 123. A forecast of 155 places 2026 in clearly above-average territory without reaching the extremes of those historic seasons.
The science behind the numbers
The Arizona team’s methodology is grounded in two peer-reviewed papers published in Weather and Forecasting, a journal of the American Meteorological Society. The first paper introduced the statistical model the team uses for seasonal North Atlantic hurricane activity, detailing how predictors are selected, how the model is structured, and how its skill is measured against historical baselines. The second paper extended that framework to focus on predicting ACE and major hurricane counts, arguing that accumulated cyclone energy is a more informative seasonal metric than storm counts alone.
Both papers draw on HURDAT2, the Atlantic best-track dataset maintained by NOAA’s National Hurricane Center. HURDAT2 provides six-hourly position and intensity records for every Atlantic tropical cyclone stretching back decades, giving the Arizona team a deep historical record to train and verify its models.
According to the team’s accompanying materials, warm Atlantic sea-surface temperatures are the primary driver behind the elevated 2026 forecast. Those warm conditions, the team notes, outweigh any residual suppression from El Niño patterns. The ensemble approach integrates multiple predictors to improve accuracy beyond what simple climatological averages would suggest. The broader university site confirms this work is part of an ongoing research program, not a one-off projection.
How it compares to other early outlooks
The Arizona forecast is one of the first academic outlooks to land for 2026, but it won’t be the last. Colorado State University’s Tropical Meteorology Project, led by Dr. Phil Klotzbach, typically releases an initial seasonal forecast in April, with updates in June and August. Tropical Storm Risk (TSR), based at University College London, also publishes early-season projections. NOAA’s Climate Prediction Center usually issues its official outlook in late May, combining statistical and dynamical models that may weight predictors differently than the Arizona approach.
Until those forecasts arrive, it is difficult to know whether the broader expert community will converge on numbers as high as Arizona’s or settle on a somewhat lower range. Comparing the spread of predictions across multiple groups will give a much clearer picture of the consensus heading into June 1.
What we still don’t know
Several important gaps remain. The Arizona department’s summary page reproduces the forecast table and discusses general drivers, but the full set of numerical inputs, including the exact sea-surface temperature anomalies and wind-shear values fed into the 2026 model run, has not been released in a detailed technical appendix. Without those inputs, independent researchers cannot replicate the specific 20/9/4/155 output or test how sensitive it is to small changes in the predictors.
The peer-reviewed papers that underpin the model were published before the 2026 season, meaning the model’s skill scores were calculated against earlier years of HURDAT2 data. Whether the team has recalibrated using the active seasons of the early 2020s is not confirmed in publicly available materials. That matters because if Atlantic temperatures in spring 2026 sit outside the historical training range, the model’s predictive relationships could behave differently than they did during the calibration period.
The Arizona team acknowledges the role of warm sea-surface temperatures as a primary driver but does not publish a formal margin of error around the 155 ACE figure on its public-facing summary. Readers should treat these numbers as central estimates, not precise guarantees. No seasonal forecast, from any group, can predict where individual storms will form or make landfall.
What coastal residents and businesses should do now
A forecast of 20 named storms and 9 hurricanes does not guarantee that any single community will take a direct hit. But it does raise the odds that multiple landfalls will occur somewhere along the Gulf Coast, the Eastern Seaboard, or in the Caribbean. In an active season, small shifts in steering currents can determine whether a long-lived storm curves harmlessly out to sea or drives inland over a densely populated coastline.
That uncertainty is exactly why early preparation matters. Homeowners in hurricane-prone areas should review insurance policies now, particularly flood coverage. Standard homeowners insurance does not cover flood damage, and National Flood Insurance Program policies carry a 30-day waiting period after purchase before they take effect. Waiting until a storm is in the forecast is too late.
Beyond insurance, practical steps include documenting property with photos and video, assembling or refreshing emergency supply kits, and reviewing local evacuation routes. Businesses with supply chains, facilities, or employees in coastal zones should revisit continuity plans and consider how sustained storm activity could disrupt logistics, power infrastructure, or staffing. Emergency managers can use an above-average forecast as a prompt to stress-test shelter capacity, communications systems, and mutual-aid agreements before the season’s statistical peak in August and September.
Why one forecast matters less than the pattern behind it
No single seasonal outlook should drive panic or complacency. The Arizona projection is one statistically grounded estimate among several that will be issued before the 2026 season begins. Its elevated numbers are consistent with a warm Atlantic and a recent stretch of active years, but the absence of detailed input data, formal uncertainty ranges, and independent replication leaves room for the final tally to come in higher or lower.
What the forecast does reinforce is a pattern that has been building for years. Atlantic sea-surface temperatures have been running at or near record levels, and the conditions that fuel intense hurricanes, warm water, moist air, and low wind shear, have been persistently favorable. In that environment, planning around the hope of a quiet year is a gamble that coastal communities can’t afford to take. The University of Arizona’s numbers are an early signal. When NOAA, Colorado State, and other groups weigh in over the coming weeks, the picture will sharpen. The time to prepare is before it does.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
