In this series, the Gallagher Research Centre (GRC) talks to academics and innovators whose work has had a profound impact on the reinsurance industry.

This month the GRC catches up with global Tropical Cyclone (TC) expert Dr Greg Holland. With more than 150 journal articles, thousands of citations and a wind field model that bears his name, Greg’s research career has covered a variety of topics; from studying TC formation, frequency, severity to the impacts of climate change. Greg’s work developing the ‘Holland model’ in 1980 (a wind and pressure model) went on to become instrumental in the (re)insurance industry as it became widely adopted in the development of the first hurricane catastrophe models used for risk management.

You started out with a Maths degree. What made you interested in switching attention to research TCs?

I did start out studying Maths at University and went on to get a First. While doing this I noticed an advert that the BOM (Bureau of Meteorology, Australia) was running a cadetship scheme to fund people through University, as long as you’d go and work for them after you finished. This became a game-changer for me. Overnight I went from being a penniless student to being one of the richest kids on campus and paid to be there! But more importantly, it took me into meteorology.

The real beauty of studying meteorology is that your laboratory is right outside of your window every day. It’s an ever-present opportunity of learning. TCs are one of the most extreme forms of our weather so there was always a natural interest there.

The wind and pressure model (1980) you developed went on to have a profound impact — particularly in the emergence of probabilistic hurricane models. Did you have any idea back then how the research you were doing would be used within the insurance industry?

None at all. It was also a slow burner and didn’t get adopted overnight. After I left forecasting, I was asked to help write a book about forecasting methods for TCs. As part of that research, I’d reviewed a paper by the Army Corps of Engineers on TC observations and modelling. I noticed that the model just didn’t fit the observations — the physics were all wrong! So, I set about fixing it! That’s basically how the Holland model was born.

Why has the ‘Holland model’ become so widely used?

It’s very accessible — something that’s not always the case with scientific methods. The model was so simple to code up, people were really able to understand it and then implement it for their own use.

In terms of the insurance industry, the scale of damage from Hurricane Andrew (1992) was obviously a major factor. I then started to work with some of the modelling companies as they began to develop stochastic event sets of hurricanes.

Interestingly though, I think the biggest impact the model has had is on property building codes. The Holland model began to be adopted into standard practices in building designs as resilience became of critical concern.

What is the ‘Holland Model’?

Typhoons, Tropical Cyclones (TC) and Hurricanes are different terms for the same weather phenomenon — a rotating organised system of clouds and thunderstorms originating in tropical or sub-tropical waters with low-level circulation. A key feature of TCs is their symmetrical structure.

Published in 19801 the wind field model described by Greg Holland defined the radial distribution of surface pressure within a TC. It defined how the relationship between the pressure (p) at the centre of the TC and the pressure at the periphery (the difference being defined as ∆p) is related to the radius (r) to maximum winds (RMW). This allowed risk modellers to recreate the wind field of TCs from a basic understanding of their central pressure.

The Luminaries Interview Series: Greg Holland
Figure 1: a) Holland Wind field Model1 b) Representation of Hurricane Rita wind field

Why was it so important?

Understanding this key relationship allowed catastrophe modellers to create credible simulations of the wind field throughout the passage of a TC. As the model became incorporated into building codes, such as those in Florida2, it led to enhanced resilience in the construction and retrofitting of buildings in high-risk areas to TCs.

What are the most critical questions TC researchers should be focused on right now?

There are many, but I think flood research is critical. It’s fast becoming the headline story for many TCs. The Clausius-Clapeyron relation tells us that for every 1 °C (1.8 °F) rise in temperature the atmosphere can hold an additional 7% moisture content — so in a warming climate, we can expect more extreme precipitation with these events. For example, 70%3 of the claims during Cyclone Debbie (2017) were coming from water related damage.

At the same time, I see real opportunity to further refine our models with the use of machine learning and downscaling models to increasingly higher resolutions. Likewise, the availability of data on building codes and property exposure is infinitely improved. We can capture the risk to properties with increasing precision.

Extreme precipitation from Tropical Cyclones: Debbie 2017

Warmer than normal sea surface temperatures led to Cyclone Debbie unleashing unprecedented rainfall across Queensland, Australia in 2017. The towns of Undercliff and Upper Springbrook reported rainfall totals of 544mm in 24hrs and 890mm in 48hrs respectively. A combination of accumulated flow in rivers, flash flooding and wind-driven water ingress led to widespread flooding.

Research3 suggests that Climate Change will result in increasing accumulated rainfall from these events. This can also have the secondary impact of rainfall able to penetrate further inland, as Cyclone Debbie showed. This could mean we see flooding as being an increasingly important component of TCs.

The Luminaries Interview Series: Greg Holland
Figure 2: Cyclone Track Debbie (IBTrACS)(2017) — Accumulated 7-day rainfall 23rd-30th March 2017 (NASA).

Researchers are increasingly able to evidence the attribution of Climate Change to single events. The latest IPCC reports show an increasing consensus on the world experiencing a higher frequency of the most dangerous Hurricanes (Cat 4-5) in the future. Is the debate on climate change and hurricane activity settled?

We have good confidence on future TC intensity globally. Where there’s less confidence is on the how larger scale patterns (El Nino Southern Oscillation) interact and impact regional variability and frequency. Modeling small scale features of these events should really be an area of focus.

ENSO and correlation to Tropical Cyclones

The El Nino Southern Oscillation (ENSO) in particular, is one of the most important influences on TC activity. The genesis, track, landfall and intensity of TC are largely controlled by large-scale environmental conditions. Given ENSO’s strong modulation on both the atmosphere and ocean, it modulates critical aspects of TC development across all global oceans. La Nina phases of ENSO are associated with more favourable conditions for North Atlantic Hurricane development than EL Nino.

The Luminaries Interview Series: Greg Holland
Figure 3: Ten-year record of the Oceanic Nino Index (ONI) and North Atlantic Hurricane tracks 2020-2022

You’ve had a career in research but interacted closely with the insurance industry. What lessons do you think each side can learn from the other?

It’s a critically important relationship for both parties. I’ve had some of my best ideas in the corridors of insurance companies.

The benefits to academics are that they get to see their ideas implemented into real-world solutions. Bear in mind that academics are just curious kids who want to investigate something. Having industry partners provide feedback and guidance on the application of research is incredibly useful.

Industry partners understandably have a different priority. While this can involve being more demanding on the timescales of research, I think it’s important for industry to recognise the need to be patient; good research can take time to give you answers, but they will come!

How can we help?

The Gallagher Re Model Research & Evaluation team have developed a series of global track, frequency and severity adjustments that allow clients to test the impact of future climate scenarios on their portfolio. The work leverages the latest peer-reviewed scientific publications, and the adjustments are available across multiple climate scenarios, time frames and vendor catastrophe models. For further information please contact your Gallagher Re client representative.

To learn more about TC impacts in 2022 as well as ENSO and Climate Change, please read the Gallagher Re Natural Catastrophe Report 2022.

The Gallagher Research Centre (GRC) is a dedicated fund for collaboration between academics and the reinsurance industry. It provides access to independent, peer-reviewed academic work to support applied innovations across all Gallagher Re’s advisory and transactional products related to both natural and man-made perils.

Sources

1. Holland, G. J. “An analytic model of the wind and pressure profiles in hurricanes,” Monthly Weather Review as cited by American Meteorological Society, 1 Aug 1980. PDF file.

2. “Development of Loss Relativities for Wind Resistive Features of Residential Structures,” version 2.2, INTRARISK Applied Research Associates, 28 Mar 2022. PDF file.

3. Bruyere, C. L. et al. “Physically-based landfalling tropical cyclone scenarios in support of risk assessment,” originally published in Weather and Climate Extremes, 26, 100229. Dec 2019. PDF on Elsevier.

Disclaimer

The opinions and views expressed in the above articles are those of the contributor only and are for guidance purposes only. The authors disclaim any liability for reliance upon those opinions and would encourage readers to rely upon more than one source before making a decision based on the information.