Anthropogenic effects on tropical cyclones near Western Europe

The study addresses whether and how human activities have altered tropical cyclone (TC) frequency (TCF) near Western Europe, a region with relatively low baseline TC activity but high exposure and vulnerability. While TCs occasionally affect Western Europe and can cause high impacts, understanding TCF changes is difficult due to limited observational records, model limitations, and lack of an accepted theory for global mean TCF. The authors leverage observations and targeted modeling to ask: How has TCF changed and will it change near Western Europe, and why? The motivation includes historical European TC encounters, potential for high hazard, and the need to disentangle anthropogenic forcing from natural variability in a midlatitude-affected region.

- Prior work shows relatively high confidence that TC peak winds, precipitation, and storm-surge-related flooding may increase with global warming, but TCF remains among the least understood components (e.g., Knutson et al. 2019, 2020; Walsh et al. 2016). - Challenges for TCF include limited paleoclimate proxies, model biases in reproducing observed TCF, and inconsistent projections across models, alongside the absence of an accepted theoretical framework for global mean TCF variability. - Western Europe has documented TC encounters, often involving extratropical transition (ET), and these events can produce significant impacts; extensive literature characterizes ET processes and European impacts. - Previous studies using prescribed SST patterns under warming suggested possibly increased TCF near the European Atlantic coast; however, consensus is lacking and results can depend on modeling approach, coupling, and ensemble size. - Recent advancements in high-resolution global models and large ensembles (e.g., GFDL SPEAR) enable more reliable attribution of anthropogenic influences (aerosols vs GHGs) on TCs and decomposition of genesis vs track contributions.

Observations and metrics: - Best-track observations: North Atlantic HURDAT (subset of IBTrACS) for 1966–2020, from first attainment of tropical-storm strength (≥34 kt) to last time at that strength, regardless of ET status; data restricted to Equator–60°N. TCF defined as number of 6-hourly TC appearances per 5°×5° grid per year. - Environmental datasets: HadISST SST (1° grid) and ERA5 monthly reanalysis for August–October (peak TC season for basin and Western Europe). - Climate indices: ONI (ENSO), IPO (2nd EOF PC of filtered global SST), AMO (detrended North Atlantic SST mean), and AMM. Statistical analyses: - Singular Value Decomposition (SVD): Both global and regional SVDs relate SST fields to TCF. For each mode, paired expansion coefficients (ECs) for SST and TCF and squared covariance fraction (SCF) quantify coupled variability. Regional SVD evaluates correspondence between local TCF and global modes (global warming-like vs IPO-like) grid-by-grid. - Significance testing: Two-sided bootstrap for epochal TCF changes; two-sided t-tests for trend differences between experiments. Modeling framework: - Model: GFDL SPEAR Earth system modeling system with ~50 km atmosphere/land and ~1° ocean/sea-ice (refined near equator). A TC tracking algorithm identifies TCs every 6 h using wind ≥16.5 m s−1 and a warm-core criterion at identification; ET does not terminate tracks, mirroring observational handling. - Large ensembles: • AllForc: 30 members, 1966–2014 historical forcing; 2015–2100 SSP2-4.5 and SSP5-8.5. Includes anthropogenic (GHGs, aerosols, ozone) and natural forcings (volcanic, dust, solar). • AllForc_NoAE: 12 members, 1966–2020, like AllForc but anthropogenic aerosols fixed at 1921 levels. • NatForc: 30 members, 1966–2100, anthropogenic forcings fixed at 1921 levels (natural-only). - Forcing attribution via ensemble-mean trend differences: • Anthropogenic aerosol effect (1966–2020): AllForc − AllForc_NoAE. • Anthropogenic GHG effect (1966–2020): AllForc_NoAE − NatForc. • Anthropogenic GHG effect (2030–2100): AllForc − NatForc (post-2030 aerosol emissions in Europe/US assumed flattened by pollution control). Decomposition of TCF changes: - Total Analysis: Decomposes TCF trend into genesis contribution, track contribution, and their nonlinear interaction by expressing TCF as the integral of genesis frequency times track probability. - Origin Analysis: Attributes changes in TCF within a target region (magenta box near Western Europe) to genesis and track changes in remote genesis locations, identifying source regions (e.g., MDR) that drive local TCF changes. Genesis diagnostics: - Genesis Potential Indices: Emanuel–Nolan GPI (ENGPI) and Dynamic GPI (DGPI). DGPI components include vertical wind shear magnitude (200–850 hPa), meridional shear vorticity (500 hPa), vertical motion (ω500), and low-level absolute vorticity (850 hPa). DGPI linearized via log-transform to quantify relative contributions of environmental factors to DGPI trends over the MDR. Target region definition: - A magenta box near Western Europe identified via observed TCF increases and dominance of the global-mean SST (anthropogenic) mode in regional SVD, used consistently for model–observation comparisons and decomposition analyses.

- Observed historical change (1966–2020): • Significant increase in TCF over the tropical Atlantic and near Western Europe; near Western Europe, increases can reach up to ~200% relative to the local climatological annual mean TCF. • No discernible trend in the fraction of ET or the northernmost TC latitude in best-track data, arguing against artifact-driven trends. • Regional SVD indicates the global warming-like SST mode dominates the TCF increase in the deep tropics, US East Coast, and near Western Europe; IPO-like variability is secondary. - Attribution from large-ensemble simulations: • Historical period (1966–2020): - Anthropogenic aerosols drive an increase in TCF near Western Europe; GHGs and natural-only forcings do not produce this increase. Combined aerosol+GHG still shows increased TCF, matching observations in magnitude and pattern. - Decomposition (Total Analysis) near Western Europe indicates contributions to the aerosol-driven TCF increase: genesis +50%, track +39%, nonlinear +11% (sum 100%). Track changes play a secondary but notable role. - Origin Analysis points to the Main Development Region (MDR; 10°–20°N, 60°–15°W) as the key source: increased genesis in the MDR and some track changes from near-Africa origins feed TCF increases near Western Europe. • Future period (2030–2100): - Under SSP2-4.5 and SSP5-8.5, simulations project a decrease in TCF near Western Europe primarily due to anthropogenic GHG forcing, statistically significant under SSP5-8.5. - Decomposition shows genesis changes dominate the future decrease; track changes slightly oppose the total signal: genesis ~107%, track ~−16%, nonlinear ~9% (summing to 100% of the total magnitude). - Environmental controls on genesis (DGPI): • Aerosol effect (1966–2020): Increased MDR genesis linked mainly to reduced vertical wind shear (DGPI component contributions over MDR: V-SHEAR 44%, OMEGA 25%, VORT 17%, M-SHEAR 14%). Reduced shear arises primarily from weakened zonal shear associated with a weakened subtropical jet at 200 hPa, likely connected to enhanced midlatitude lower-tropospheric warming and thermal wind adjustment. • GHG effect (2030–2100): Decreased MDR genesis tied chiefly to weakened ascent over the MDR (DGPI components: OMEGA 56%, M-SHEAR 39%, VORT 7%, V-SHEAR −2%), consistent with a relative SST pattern featuring weaker warming in the MDR than surrounding midlatitudes/South Atlantic/eastern equatorial Pacific, producing a drier, less convective MDR. - Hazard implications near Western Europe: • TCF rises historically (1966–2020) and declines in future (2030–2100); mean TC intensity near Western Europe shows no clear trend in observations or simulations. • ACE (combining frequency and intensity) increases historically and is projected to decrease, largely mirroring TCF changes rather than intensity changes.

The study resolves the research question by attributing observed historical increases in TCF near Western Europe primarily to anthropogenic aerosol changes that modified large-scale environmental conditions over the Atlantic MDR. Reduced European/US aerosol emissions in recent decades produced midlatitude tropospheric warming and a weakened subtropical jet, lowering vertical wind shear and enhancing TC genesis in the MDR, thereby increasing TCF near Western Europe. Looking forward, under scenarios with flattened aerosol emissions in Europe/US and increasing GHGs, the MDR experiences relatively weaker SST warming compared to surrounding regions, leading to drier conditions and weakened ascent, which suppresses genesis and reduces TCF near Western Europe. Thus, the same region sees opposite trends historically versus projected futures due to different anthropogenic forcing balances. These findings are significant because they separate aerosol and GHG effects using large ensembles and show that regional aerosol controls can strongly modulate basinwide genesis and downstream TCF near Western Europe. They also clarify that frequency, not intensity, dominates changes in ACE-based hazard for this region. The results align with broader projections of decreased North Atlantic TCF under GHG warming, while highlighting that earlier studies prescribing SSTs could yield different regional conclusions. The decomposition and origin analyses underscore the critical role of MDR genesis pathways in shaping European-adjacent TCF outcomes.

- The observed increase in tropical cyclone frequency near Western Europe during 1966–2020 is likely driven by anthropogenic aerosol changes, which reduced vertical wind shear via subtropical jet weakening and boosted MDR genesis. - In contrast, under future GHG-dominated forcing (with controlled regional aerosols), models project a decrease in TCF near Western Europe by 2030–2100 (especially under SSP5-8.5), primarily due to suppressed MDR ascent associated with relative SST patterns, leading to reduced genesis. - Changes in large-scale environmental conditions governing genesis explain both the historical rise and projected future decline in TCF; track effects contribute historically under aerosol forcing but are minor in the GHG-dominated future. - Hazard implications: ACE near Western Europe increases historically and likely decreases in the future, mainly reflecting frequency changes; no robust trend in mean TC intensity near Western Europe is found. - Future research directions: incorporate precipitation-related hazards and compound risks; analyze direct landfall statistics with larger samples and refined regional definitions; improve intensity-resolving capability via higher-resolution ensembles; further disentangle aerosol–GHG–internal variability interactions; assess sensitivity to alternative SSPs and aerosol emission pathways.

- Observational constraints: Reliable records are limited to the satellite era (post-1966); landfall counts in Western Europe are too sparse for robust statistical testing, necessitating a regional-box proxy. - Methodological constraints: SVD is a statistical tool and cannot perfectly separate anthropogenic forcing from internal variability; interactions between the two may persist. Attribution relies on ensemble-mean trend differences and assumes post-2030 aerosol flattening in Europe/US. - Model limitations: SPEAR’s ~50 km atmosphere limits resolution of major TC intensity; uncertainties in parameterizations and aerosol representations remain. Prescribed forcing scenarios and ocean–atmosphere feedbacks could affect regional SST patterns and ascent. - Scope limitations: Precipitation-related hazards and other impact metrics were not analyzed; ET dynamics are included in TCF by design but detailed ET process changes were not a focus.