Recent tropical cyclone changes inferred from ocean surface temperature cold wakes

Ralf Toumi,Shuai Wang

doi:10.1038/s41598-021-01612-9

Abstract

It has been challenging to detect trends of tropical cyclone (TC) properties due to temporal heterogeneities and short duration of the direct observations. TCs impact the ocean surface temperature by creating cold wakes as a “fingerprint”. Here we infer changes of the lifetime maximum intensity (LMI), size and integrated kinetic energy from the cold wakes for the period 1982–2019. We find a globally enhanced local cold wake amplitude 3 days after the LMI of − 0.12 ± 0.04 °C per decade whereas the cold wake size does not show any significant change. Multivariate regression models based on the observed ocean cooling, the TC translation speed and the ocean mixed layer depth are applied to infer LMI and TC size. The inferred annual mean global LMI has increased by 1.0 ± 0.7 m s−1 per decade. This inferred trend is between that found for two directly observed data sets. However, the TC size and the TC destructive potential measured by the integrated kinetic energy, have not altered significantly. This analysis provides new independent and indirect evidence of recent TC LMI increases, but a stable size and integrated kinetic energy.

Highlights

It has been challenging to detect trends of tropical cyclone (TC) properties due to temporal heterogeneities and short duration of the direct observations
We find stronger ocean cooling along the TC moving direction with a slight shift to the right of the motion
We show that the ocean surface cooling can be used as an independent measure of long-term global changes of TC intensity, size and destructive potential

Summary

Introduction

It has been challenging to detect trends of tropical cyclone (TC) properties due to temporal heterogeneities and short duration of the direct observations. Best tracks have been collected for decades with the best observational techniques and analysis protocols of the time This naturally creates temporal heterogeneities and may lead to unphysical detections, for example, the underestimation of TC frequency in the Atlantic prior to the satellite era[11], and the intensity discontinuity in the western North Pacific due to the termination of aircraft r econnaissance[12]. The ADT-HURSAT sacrifices some valuable flight observations for homogeneity but arguably allows more robust trend detection of TC intensity It is found with the ADT-HURSAT that the positive trend of annual mean LMI of global TCs is not statistically significant for 1982–20098, which is contradictory to the significant trend found with the best track. More evidence for the long-term change of TC intensity is needed

Methods

Results

Conclusion