Strengthening disaster preparedness through impact-based forecasting.

Tropical cyclones (TCs) displace several millions every year. Detailed information on the potential impacts triggered by TCs can render the preparation more efficient, targeted and potent. This study introduces a global, open-source impact-based forecast system for TC-related human displacement.
Published in Earth & Environment
Strengthening disaster preparedness through impact-based forecasting.
Like

Share this post

Choose a social network to share with, or copy the URL to share elsewhere

This is a representation of how your post may appear on social media. The actual post will vary between social networks

At the time of writing this blogpost, Category 3 Hurricane Milton has just made landfall in Florida, USA, causing widespread destruction and severely impacting the livelihoods of countless residents. The storm has already claimed at least 14 lives [1], and millions were affected by evacuation orders [2]. While the full extent of damage from Hurricane Milton is still being assessed, this catastrophic event underscores the importance of effective anticipatory action such as early warning systems, evacuation planning, and emergency protection in reducing damages.


Hurricane Milton seen from the International Space Station on 8 October. Source: NASA/Michael Barratt.
undefined

Currently, early warning systems mainly provide information about the physical characteristics of hazards, such as the maximum wind speed of tropical cyclones (TCs, also known as hurricanes in the US and typhoons in Asia) in our case. This approach requires decision-makers to rely on expert knowledge and past experiences to assess potential impacts and plan accordingly, which can be challenging and may lead to inconsistent or incomplete evaluations, and sub-optimal allocation of resources.

Our study addresses this gap by presenting a global implementation of impact-based forecasting of TC-related displacement. In collaboration with the Internal Displacement Monitoring Centre (IDMC) in designing the system, we use open-data and open-source code for this implementation while keeping the computational cost low. This approach makes the implementation accessible to institutions and individuals worldwide, without being limited to (data-)rich regions.

We systematically translate weather information into displacement risk by combining the likelihood and potential severity of TC events, the number of people or settlements in exposed areas, and the vulnerability of these communities. We demonstrate this TC displacement forecast system through a case study of Cyclone Yasa, which displaced over 20,000 people in Fiji in December 2020 [3]. We chose this particular case study because the Pacific Islands are often under-studied, and the island characteristics can show the "hit-or-miss" scenarios. Hence, it highlights the importance of providing probabilistic impact forecasts, as the likelihood of whether impact occurs is crucial information for decision-makers.

Forecasted displacement in Fiji by tropical cyclone Yasa.

figure 2

Furthermore, the uncertainties of the impact forecast stem not only from the probabilistic weather forecast but also from other input data and their interplay with the weather forecast. These include uncertainties in population estimation and vulnerability quantification. We conduct a global uncertainty analysis which considers all the input uncertainties for all recorded TC displacement events from 2017-2020. We argue that for impact forecasts, it is more important to consider all plausible outcomes instead of only the most probable ones, as this provides a more comprehensive understanding of potential risks and supports better-informed decision-making for anticipatory actions. With comprehensive uncertainty information, decision-makers can more clearly plan according to their risk profile: the level of certainty required for protecting a nuclear power plant is different than for implementing mass-evacuation measures.

Additionally, we conduct sensitivity analyses to further understand which uncertain inputs contribute most to the overall uncertainty distribution of the forecast outcomes. We find a general trend that at long forecast lead times, weather forecast uncertainty dominates, but the uncertainty in vulnerability quantification becomes more important towards the landfalls of TCs. This information can guide decision-makers on what uncertainties to consider when devising anticipatory action plans.

The implementation of this impact forecast system and the findings in this paper are a collaborative effort between the academic research group at ETH Zurich and stakeholders in IDMC. As all the codes and findings are openly accessible to the public, we hope to encourage more open science initiatives by expanding collaborations between academia and relevant stakeholders. Such partnerships are becoming increasingly crucial in addressing the consequences of heightened probabilities of extreme events due to climate change.

References:

[1] Fortin J. and Mayorquín O. Hurricane Milton’s Impact: What We Know So Far. New York Times (2024). https://www.nytimes.com/2024/10/10/weather/hurricane-milton-damage-florida.html

[2] Mazzei P. and Taft I. Fears of Hurricane Milton Drive Millions From Their Homes in Florida. New York Times (2024). https://www.nytimes.com/2024/10/08/weather/hurricane-milton-florida-evacuations.html

[3] IDMC. Global Internal Displacement Database. IDMC. https://www.internal-displacement.org/database.

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in

Follow the Topic

Environmental Sciences
Physical Sciences > Earth and Environmental Sciences > Environmental Sciences

Related Collections

With collections, you can get published faster and increase your visibility.

Biology of rare genetic disorders

This cross-journal Collection between Nature Communications, Communications Biology, npj Genomic Medicine and Scientific Reports brings together research articles that provide new insights into the biology of rare genetic disorders, also known as Mendelian or monogenic disorders.

Publishing Model: Open Access

Deadline: Jan 31, 2025

Clinical trials

In this call for papers, Nature Communications invites submissions of papers reporting interventional trials, including those yielding negative results for which primary end-points were not met.

Publishing Model: Open Access

Deadline: Dec 31, 2024