Controls on Wind Asymmetry in North Atlantic Tropical Cyclones

• Main Author: Dougherty, Erin
• Format: Thesis
• Language: unknown
• Published: University of Virginia 2015
• Subjects: Tropical cyclones; North Atlantic Ocean; coastal storms; North Atlantic
• Online Access: https://dx.doi.org/10.18130/v3-zjrm-rk20; https://libraetd.lib.virginia.edu/public_view/gb19f645b

Each year, coastal areas along the Western Atlantic Ocean incur significant damage from the impacts of tropical cyclones (TCs). Although current warning systems emphasize each TC’s intensity, information on the TC’s structure would prove beneficial. One important structural variable is wind asymmetry––the differential and radial wind extent of maximum winds. Recent work has suggested a number of factors contribute to TC asymmetry, such as factors associated the TC itself and those linked to the surrounding environment. However, it is unknown to what extent these TC-induced and environmental variables account for variability in TC wind asymmetry. The Extended Best Track Dataset, Statistical Intensity Prediction Scheme, and optimum interpolation sea-surface temperature analysis were used to obtain information on TC properties and associated environmental variables. Multiple linear regressions (MLRs) were then performed to determine the factors that control wind asymmetry in North Atlantic TCs from 1988–2012 among all locations and separately for three prominent TC tracks (Gulf, Baroclinic, and Azores). Among all four MLR models, increased asymmetry was associated with an increase in minimum pressure, shorter storm duration, and closer proximity to land. Additionally, asymmetry increased with storm size for all MLR models except the Azores track and with an increase in vertical wind shear in all models except for the Gulf track. In all MLR models, the relatively low explained variances (0.12–0.15) suggest sampling error inherent in the dataset and/or other important variables not included in the model. Despite the low R 2 values, the MLR models exhibit some predictive capability, especially the Gulf MLR with a low root mean-square error of 0.28. Results from this study provide an understanding of the predictability of North Atlantic TC wind asymmetry and may eventually be useful for predicting changes in TC wind structure.