A retrospective on the 2025 Atlantic hurricane season

TL;DR

The paper analyzes the anomalous 2025 Atlantic hurricane season, characterized by intermittent activity with three distinct clusters despite broadly unfavorable baseline conditions. It evaluates environmental factors (SSTs, vertical shear, high SLP, TUTT) and relative SST anomalies, and examines teleconnections (MJO, Kelvin waves, NAO) as drivers of cluster formation. Forecasts suggested above-average activity under neutral ENSO, but the season ended with above-normal ACE ($ACE = 132.5\times 10^4$ kt$^2$) and a below-average number of storms, driven by several high-intensity hurricanes, including three at category five. The analysis links equatorial wave forcing and thermodynamic shifts to clustered tropical cyclone genesis, offering insights for improved seasonal forecasting and highlighting the role of large-scale atmospheric oscillations in Atlantic TC activity. Data and code used for replication are provided to support further research.

Abstract

The 2025 Atlantic hurricane season saw above average activity overall, with extended quiet periods separated by three distinct clusters of activity. The broad-scale conditions were often unfavourable for cyclogenesis and common drivers of activity such as La Nina were not present. However, short-term variability, including periods of weak shear and episodic equatorial wave driving, led to the clusters of activity. When storms were able to overcome the unfavourable conditions, above-average SSTs provided the energy for intensification, leading to the formation of five hurricanes, of which three (Erin, Humberto, and Melissa) reached category five.

Figures (7)

• Figure 1: Summary map of the Atlantic basin and 2025 tropical cyclone tracks. Tracks are coloured by category (tropical depression TD; tropical storm TS; hurricane H; major hurricane MH) with line width proportional to maximum sustained wind speed. The main development region (MDR) used in figure \ref{['fig:SSTs']} and wave analysis region used in figures \ref{['fig:ACE_MJO_KW']} and \ref{['fig:tracks']} are indicated in blue. Track data for 2025 obtained from ATCF realtime sampson2000
• Figure 2: Daily-average SSTs ($^\circ$ C) in the Atlantic main development region (10--20$^\circ$ N, 20--85$^\circ$ W) in 2023--2025, compared with the long-term (1991--2020) average. Data obtained from NOAA OISSTv2.1 huang2021.
• Figure 3: Cumulative Accumulated Cyclone Energy (ACE) in 2025 compared with the long-term (1991--2020) average. Lifecycles of 2025 tropical cyclones inset, coloured as in figure \ref{['fig:summary']}. Data for 2025 obtained from ATCF realtime sampson2000 and for 1991--2020 climatology from HURDAT2 landsea2013.
• Figure 4: Upper-troposphere (200 hPa) geopotential height contours (grey dashed, 50m interval) and flow streamlines (light green) with 850--200 hPa vertical shear (coloured) in June to September 2025. Troughs are indicated by thick blue lines. Data from preliminary ERA5 reanalysis hersbach2020.
• Figure 5: Monthly-average anomaly (compared to 1991--2020 climatology) in June--July (a--d) and August--October 2025 (e--h) of mean sea-level pressure (MSLP), relative sea-surface temperatures (rSST), 850--200 hPa temperature lapse rate, and 500 hPa relative humidity (RH). Data from preliminary ERA5 reanalysis.