Date of Award

Summer 2026

Language

English

Embargo Period

8-11-2026

Document Type

Master's Thesis

Degree Name

Master of Science (MS)

College/School/Department

Department of Atmospheric and Environmental Sciences

Program

Atmospheric Science

First Advisor

Oliver Timm

Committee Members

Mathias Vuille

Keywords

Trade Wind Inversion, Extreme Rainfall, Kauaʻi

Abstract

Extreme rainfall is a persistent hazard in the Hawaiian Islands, where steep terrain and short catchments convert heavy rain into damaging floods with little warning. The trade wind inversion, a subsidence-driven stable layer that caps the marine boundary layer over the subtropical North Pacific, limits the depth of convection and is therefore expected to govern whether available moisture can be converted into extreme rainfall. Previous work has characterized the inversion climatologically and has related the El Niño-Southern Oscillation (ENSO) and the Pacific-North American (PNA) pattern to Hawaiian rainfall totals, but the behavior of the inversion on the specific days when extreme rainfall occurs, and the question of whether the climate modes act on extreme rainfall by way of the inversion, have not been established. Both modes can influence the inversion through large-scale subsidence, with the PNA pattern modulating the North Pacific subtropical high and ENSO altering the trade wind flow, yet whether they reach extreme rainfall by way of the inversion rather than through moisture transport alone remains untested.

This thesis addresses these questions using radiosonde observations from Līhuʻe, Kauaʻi, in three parts: first characterizing the seasonal and 00Z-12Z climatology of lower-tropospheric thermodynamic structure, then testing whether extreme rainfall days depart from that climatology and whether a coherent signal develops before and after each event, and finally evaluating whether ENSO and PNA phase modulate trade wind inversion absence and extreme rainfall occurrence, and whether the same drivers govern both. First, a seasonal climatology of convective available potential energy (CAPE), lifted condensation level (LCL) pressure and precipitable water is constructed from IGRA2 soundings for 1993-2025, alongside a climatology of trade wind inversion strength and occurrence for 1981-2013, the period covered by the available inversion identification record. Second, thermodynamic conditions on extreme rainfall days, identified from the NOAA Storm Events Database for Kauaʻi, were compared against this climatology, and superposed epoch analysis was used to trace the evolution of the atmospheric column from three days before to three days after each event. Third, odds ratios, a measure of how much more or less likely inversion absence is under one condition than another, and contingency tables with bootstrap and Monte Carlo permutation were used to test whether ENSO and the PNA modulate the frequency of inversion absence and the occurrence of extreme events.

Extreme rainfall days are thermodynamically distinct, showing elevated CAPE, elevated precipitable water, a lower cloud base, and a weakened or absent inversion consistently across every month and both sounding times. The strongest result is the association between inversion absence and extreme rainfall: the odds of inversion absence are 5.14 times higher on wet season event days (95% CI 4.14-6.43) and 6.94 times higher on dry season event days (95% CI 4.82-9.90) than on ordinary days, both significant at p < 0.001. A coherent signal precedes events by two to three days, with wet season CAPE rising from 199 J/kg three days before the event to 709 J/kg on the event day against a climatological median of 94 J/kg, while inversion strength falls to 0.40 °C against a climatological 1.80 °C. Both climate modes modulate these conditions in the region. In the wet season, El Niño reduced the odds of inversion absence across all sounding days (OR 0.72) and coincided with 52% fewer extreme event days than expected, whereas negative PNA doubled those odds (OR 2.01) and coincided with 134% more extreme event days than expected. The PNA effect exceeded the ENSO effect and retained its sign in both seasons.

In summary, the results indicate that the trade wind inversion governs extreme rainfall and provides a thermodynamic pathway linking modes of climate variability to extreme events over Hawaiʻi. However, the evidence is correlational and cannot conclusively establish the pathway. The absence of inversion is strongly associated with extreme rainfall but is not sufficient for it. The principal limitation is that the inversion record ended in 2013, which bounds every inversion-dependent result reported here.

License

This work is licensed under the University at Albany Standard Author Agreement.

Additional Files
Kauai_extreme_events_catalog.xlsx (81 kB)
Complete catalog of extreme rainfall events for Kauaʻi County, Hawaiʻi, 1996–2025, compiled as supplementary data
Download

Share

COinS