reask global Catastrophe Risk Modelling and forecasting services

Number of Affected People – A Vague Measure

Summary: The lack of a standardised measure of rapid impact reporting is discussed in this blog using the example of tropical cyclones Batsirai and Freddy, which hit Madagascar in 2022 and 2023 respectively. The current reporting standards provide various figures, such as the number of people affected or displaced, without indicating the dimension of severity. This deficit of coherence and clarity can lead to confusion and misinterpretation of the situation. The following blog suggests a comprehensive and comparable measure to report the impact of tropical cyclones in a timely manner and highlights the importance of providing fast liquidity for disaster relief actions.

Arsenal vs Chelsea 7:5

When I buy a sports newspaper, I would be very surprised if it only announced how many corners Arsenal and Chelsea took but did not mention the result. It is obvious that every newspaper always reports on the final result as well, since the number of corners is not necessarily representative of the outcome of the match. Conversely, reporting deficient information seems the accepted standard when it comes to a first estimate of the impact after a tropical cyclone (TC) makes landfall, especially in vulnerable countries. This has always puzzled me, and I would like to discuss some ideas on how to improve this in the future.

Batsirai vs Freddy 116’000 : 299’000

Just a year after the devastating tropical cyclone Batsirai hit Madagascar, probably the strongest storm ever in terms of accumulated cyclone energy, made landfall in almost the same place on 21 February: tropical cyclone Freddy.

I tried to do some research to understand the impact of Freddy in relation to Batsirai (for Madagascar only). The various data I found didn’t really help to inform me, but rather confused me.

One problem with current post-event reporting methods is the lack of a commonly agreed measure that provides a comprehensive and comparable picture of the impact of a tropical cyclone. Instead, different sources report many figures such as:

  • number of people affected;
  • number of people exposed;
  • number of people exposed to hurricane-strength winds;
  • number of people displaced, and
  • number of fatalities

This lack of consistency can lead to confusion and make it difficult to understand the severity of the situation. Furthermore, some sources may use some of these terms even interchangeably.

The second issue, which I find even more problematic, is that the number of people exposed or affected does not indicate at what level of severity they were affected. At most, it gives the number of people who were exposed to hurricane-strength winds, which is exactly one point of reference.

TC Batsirai and TC Freddy are a good example of this issue: Reliefweb declares a figure of 116,000 for Batsirai1 and 299,000 for Freddy2 for the number of affected population in Madagascar. Without further research, these two data points lead to the assumption that Freddy was hitting Madagascar more severely than Batsirai. But is this really the case?

Full distribution thanks to Metryc

The following map compares the two events visually by track and wind field. In addition, the population can be overlaid. Please click on the image to open the interactive map:

Fig 1: Population affected in Madagascar by tropical cyclone Batsirai and Freddy (interactive map).
Fig 1: Population affected in Madagascar by tropical cyclone Batsirai and Freddy (interactive map).

The attentive viewer of the interactive map will notice that this visual comparison of the two wind fields reveals that the maximum winds of TC Batsirai hit the coastal town of Mananjary precisely, while the maximum winds of TC Freddy decreased significantly shortly before hitting the town.

Figure 2 provides the full distribution of Freddy and Batsirai in terms of number of affected people by wind speed increments. Please click on the image to open an interactive graphic:

Fig 2: Population affected in Madagascar by tropical cyclone Batsirai and Freddy (interactive map).
Fig 2: Population affected in Madagascar by tropical cyclone Batsirai and Freddy (interactive chart)

Thanks to our machine learning based wind fields that we generate on a global scale 24 hours after every event by our product called Metryc we can disclose that Freddy indeed affected more people than Batsirai up to wind speeds of 110 km/h, which corresponds to wind speeds of a strong tropical cyclone on the Saffir Simpson Scale.

The sting is in the tail

Having a closer look at the tail of the cumulative distribution reveals that TC Batsirai affected significantly more people compared to TC Freddy for wind speeds in excess of 110 km/h. 200’000 people were affected by wind speeds at the level of a Cat 1 for Freddy and 375’000 for Batsirai. There were no people affected by winds of more than 150 km/h for Freddy, but still 44’000 for Batsirai!

I suppose this is the moment to realize that one simple reported figure of affected people can lead to a wrong perception and even to wrong actions eventually.

African Risk Capacity

By the way, this observation seems to be confirmed, at least in relative terms, by the announcement from ARC providing a lower payout for Freddy ($1.2m) than for Batsirai ($10.7m), under the assumption that the structure is comparable between the two years3.


In addition to providing an improved way of reporting impacts quickly and consistently, this approach can also be applied to parametric insurance, which in turn provides quick liquidity to initiate rapid disaster relief measures after an event.

At Reask, we create wind field footprints for every storm worldwide within 24 hours of an event using our Metryc product. It can be run against any kind of indices showing a full, comparable and consistent distribution, for all historical as well for all future events. It is also possible to segregate by demographics and by region to better target beneficiaries.

If there is a credible political will, the Loss and Damage Fund agreed to at the COP274 could demonstrate this by for example building out a global index based on this approach for vulnerable countries. This would be robust, reliable, easy to understand and most importantly closer to the most vulnerable people.

If you have any questions, thoughts or suggestions, please feel free to contact us at

About the Author

David Schmid is Reask’s Head of Parametric Product, leading the adoption of our products including Metryc, our tropical cyclone parametric insurance pricing and calculation tool.

David has more 14 years’ experience for the world’s leading re/insurance companies and holds a Master of Science from the University of Zurich.



Recent Posts


Close Menu