Natural and anthropogenic contributions to the hurricane drought of the 1970s–1980s

The study addresses why North Atlantic hurricane activity was markedly lower in the 1970s–1980s compared to preceding and following decades, and whether this “hurricane drought” was driven by natural variability or anthropogenic forcing. Hurricane power dissipation depends on storm number, intensity, and duration, all linked to environmental factors and especially local sea surface temperature (SST). SSTs in the tropical North Atlantic were depressed in the 1970s–1980s, coincident with the drought. Prior work implicates anthropogenic sulfate aerosols in cooling Atlantic SSTs, yet climate models accounting for sulfates alone underrepresent the observed multidecadal SST depression and associated hurricane reduction. The authors hypothesize a positive feedback: sulfate aerosols weakened the West African monsoon and Sahel rainfall, which increased Saharan dust emissions and trans-Atlantic transport; this dust further cooled the main development region (MDR) and suppressed hurricane activity.

Empirical and modeling studies have linked anthropogenic sulfate aerosols to North Atlantic SST variability and hurricane activity, with SO2 emissions peaking in the 1970s–1980s before declining due to controls. However, CMIP-class models undercapture multidecadal SST variations and the magnitude of the hurricane drought. Saharan dust has the largest aerosol radiative effect over the tropical North Atlantic and can depress SSTs and reduce hurricane activity, but dust variability is underrepresented in models. Observations show peak dust loading over the Atlantic in the 1970s–1980s and strong correlations between Sahel drought and Atlantic dust. Volcanic aerosols can also modulate Sahel drought when hemispherically asymmetric. These lines of evidence motivate considering dust as an amplifying feedback to sulfate forcing.

- Quantification of hurricane drought: Using the IBTrACS dataset, the power dissipation index (PDI) was computed and decomposed into storm number, intensity, and duration for 1960–2017 versus 1970–1990 to gauge the magnitude and components of the activity change. MDR (6–18N, 60–20W) summer SST anomalies were used as predictors of multidecadal hurricane activity. - Dust reconstruction: MDR dust optical depth (AOD) over the 20th century was reconstructed by combining (1) Barbados boundary-layer dust concentration (rescaled to match AVHRR MDR AOD during 1982–1987), (2) AVHRR satellite AOD over the MDR starting in 1982, and (3) a Sahel precipitation index (SPI; 20–10N, 20W–10E) available since 1901. Time series were June–September averaged (SPI annually averaged), low-pass filtered with a 7-year cutoff, and a regression of Barbados dust onto SPI extended the record back to 1901. Uncertainty in the SPI–dust relationship was quantified via BCa bootstrap and Wald confidence intervals. - Sulfate asymmetry index: An ad hoc hemispheric asymmetry index in sulfate aerosol optical depth was computed from GISS-E2-1-G simulations of tropospheric sulfate and a stratospheric volcanic aerosol dataset, as the difference between 0–60N and 60S–0 means over 35W–55E, representing hemispherically asymmetric forcing relevant to Sahel rainfall. - Single-column model (SCM) simulations: The MIT SCM was run under a weak temperature gradient (WTG) constraint with prescribed dust to quantify the sensitivity of MDR SST and TC-relevant variables to dust AOD. The model used a 25 m slab ocean; fixed cloud profiles; Fouquart–Morcrette radiation; Emanuel–Živković-Rothman convection. Dust optical parameters (single-scattering albedo ~0.89, asymmetry parameter g ~0.68, LW extinction tuned with τ10µm/τ0.55µm = 0.45) were set to match observed surface and TOA forcing efficiencies; sensitivity tests varied g (0.58–0.78) and ω0 (0.84–0.94). SST response δSST/δτ and changes in the thermodynamic component of the genesis potential index (GPIx = 4/3(PI–35)^2, with PI potential intensity and x mid-tropospheric saturation deficit) were diagnosed. - Low-frequency component analysis (LFCA): Applied to HadISST Atlantic SSTs (1870–2017) using 25 leading EOFs and a 7-year low-pass to extract patterns and components maximizing low-frequency variance. The first component corresponded to the global warming mode; the second low-frequency pattern/component was examined for hemispheric asymmetry and correlation with Sahel drought. Parseval’s theorem with multitaper PSD estimated the fraction of MDR SST variance (20–100 years) explained by the dust–sulfate mode.

- Magnitude of drought: PDI was ~55% larger in 1960–2017 than in 1970–1990. Decomposition suggests contributions of +22% hurricane number, +14% intensity (∝ wind speed^3), and +11% duration. MDR summer SST anomaly averaged −0.13 K during the 1970s–1980s. - Dust reconstruction and links to forcing: Low-pass filtered Barbados dust vs Sahel precipitation correlation R = 0.77. Reconstructed MDR dust optical depth during the drought was higher by 0.043 ± 0.010 relative to 1960–2017. Sulfate AOD hemispheric asymmetry correlated with Sahel precipitation at R = −0.76, consistent with sulfate-driven Sahel drought; asymmetric volcanic eruptions (e.g., El Chichón 1982, Novarupta 1912) aligned with Sahel drought spikes. - SCM sensitivity: δSST/δτ ≈ −1.4 K per unit dust AOD (range −1.3 to −1.8 K per τ across optical property perturbations). The observed dust increase (Δτ ≈ 0.043) implies an MDR SST depression of −0.06 K (bounds −0.04 to −0.10 K), explaining about 46% (31–77%) of the total −0.13 K anomaly. Dust-induced reduction in normalized thermodynamic GPI was ~15% between average 1960–2017 dust and 1970s–1980s dust, comparable to the observed 22% decrease in hurricane numbers (not including shear effects). - LFCA modes: The first mode closely tracks tropical-mean SST (global warming component). The second low-frequency pattern exhibits pronounced Northern Hemisphere horseshoe structure and large interhemispheric temperature differences, consistent with hemispherically asymmetric sulfate forcing; its component correlates with Sahel drought (R = 0.69). This dust–sulfate mode explains ~88% of MDR SST variance on 20–100 year timescales and accounts for −0.14 K MDR SST change during the 1970s–1980s, similar to the observed −0.13 K.

Findings support a mechanism where anthropogenic sulfate aerosols weakened the West African monsoon and induced Sahel drought, enhancing Saharan dust emissions and trans-Atlantic transport. The radiative effects of increased dust amplified cooling of the tropical North Atlantic MDR, reducing potential intensity, increasing mid-tropospheric dryness, lowering the thermodynamic genesis potential, and ultimately decreasing hurricane number, intensity, and duration during the 1970s–1980s. The dust–sulfate LFCA mode captures most multidecadal MDR SST variance and aligns with Sahel drought, indicating dust-mediated amplification of sulfate forcing. This helps reconcile why models lacking realistic dust variability underrepresent the SST depression and hurricane drought magnitude. The results argue against a dominant role for an internal multidecadal oscillation (AMO) in explaining the 20th-century drought; instead, external forcing (anthropogenic and volcanic) and dust feedbacks are key. Implications include limited expectation of a forced, multidecadal return to low activity solely from internal variability, and emphasize improving aerosol–dust representation to project regional hurricane risk. While global warming mode changes have limited direct effect on PDI on multidecadal scales, they can increase rainfall rates and enable poleward extension of weaker storms, affecting impacts beyond PDI.

The study proposes and supports a dust-mediated feedback mechanism amplifying anthropogenic (and volcanic) sulfate aerosol impacts on tropical North Atlantic SSTs, explaining much of the 1970s–1980s hurricane drought. Reconstructed dust AOD increases, SCM-derived SST and genesis sensitivity, and LFCA-derived dust–sulfate modes collectively indicate dust radiative forcing can account for roughly half of the observed MDR SST depression and a large share of the hurricane activity reduction. These insights clarify past variability and suggest that current high activity levels are unlikely to decline due to a natural multidecadal oscillation alone. Future work should improve constraints on sulfate and dust radiative properties, better represent dust variability and aerosol–cloud–wind feedbacks in models, and assess contributions of vertical shear and ocean dynamics to TC changes.

- Uncertainty in sulfate aerosol radiative effects and satellite retrievals of dust AOD (e.g., AVHRR underestimation) affects quantitative estimates. - Dust optical properties and vertical profiles are tuned within plausible ranges; real-world variability may broaden radiative forcing efficiency. - SCM uses WTG approximation, fixed cloud profiles, slab ocean without heat flux convergence, and omits environmental responses (e.g., wind changes), shear effects, and cloud feedbacks, potentially biasing sensitivity. - Dust reconstruction relies on proxies (Barbados dust, Sahel precipitation) and rescaling assumptions; uncertainties in the Barbados-based AOD and MDR definition are not fully propagated. - LFCA attribution to sulfate asymmetry is based on correlation and pattern characteristics; causality and separation from other forcings have residual ambiguity. - Natural variability at quasi-decadal or shorter timescales is not fully addressed.