📄 earth_science

Impact attribution of climate and population change on compound flood impacts from tropical cyclone Idai in Mozambique

This study employs an extended storyline attribution framework to demonstrate that while both climate and population changes since 1975 have increased flood exposure from Tropical Cyclone Idai in Mozambique, population growth is the dominant driver, with climate change disproportionately affecting vulnerable socio-economic groups in ways that spatial aggregation can obscure.

Original authors: Doris M. Vertegaal, Bart J. J. M. van den Hurk, Anaïs Couasnon, Dominik Paprotny, Sanne Muis

Published 2026-06-29

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Doris M. Vertegaal, Bart J. J. M. van den Hurk, Anaïs Couasnon, Dominik Paprotny, Sanne Muis

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a massive storm, Tropical Cyclone Idai, hitting Mozambique in 2019. It brought wind, rain, and a storm surge that turned the land into a giant bathtub overflowing with water. This paper asks a simple but tricky question: Who is to blame for how many people got wet?

Was it the changing climate (making the storm worse)? Or was it the growing population (putting more people in the path of the water)?

To answer this, the researchers used a "time machine" approach. They built a computer model of the storm and ran it four different times to see what would have happened under different conditions:

The Real World (Factual): The storm as it actually happened in 2019 with the 2020 population.
The "Old Climate" World: The storm as it would have happened in 1975 (before modern climate change), but with today's people.
The "Old People" World: The storm as it happened in 2019, but with only the number of people who lived there in 1975.
The "Old Everything" World: The storm in 1975 with the 1975 population.

Here is what they found, explained with some everyday analogies:

The Main Result: It's Mostly About the Crowd

If you look at the total number of people who got flooded, 80% of the increase since 1975 is because there are simply more people living in the area now. Only 9% of the increase is because the climate changed and made the water rise higher or spread further.

The Analogy: Imagine a small party in a living room. In 1975, there were 10 people. In 2020, there are 100 people. If a bucket of water spills, 100 people get wet, not because the bucket got bigger (climate), but because the room got crowded (population). The study found that the "crowding" was the main reason more people got soaked.

The Twist: It Depends on How Deep the Water Is

However, the story changes if you look at how deep the water was. The researchers didn't just count "wet" vs. "dry"; they looked at shallow water (ankle-deep), deep water (knee-to-waist), and very deep water (overhead).

Shallow Water: Most of the people here got wet because of the population growth.
Very Deep Water: Here, climate change plays a much bigger role. The study found that climate change is responsible for turning "moderate" floods into "disaster-level" floods, especially in rural areas.
The Analogy: Think of the flood as a rising tide.
- Population Change is like adding more people to the beach; they just get wetter because there are more of them standing there.
- Climate Change is like turning up the faucet; it makes the water rise higher.
- The study found that while the "crowd" (population) is the main reason more people are in the water, the "faucet" (climate) is the main reason the water is dangerously deep in certain spots.

The Hidden Inequality: Who Gets the Worst of It?

The paper also looked at who these people are. They found that climate change hits different groups unequally, but this is easy to miss if you just look at the total numbers.

The Urban vs. Rural Split: Wealthier, educated people tend to live in cities (like Beira), where the flood water is usually shallow. Poorer, less educated people often live in rural areas.
The Climate Effect: In those rural areas, climate change pushes the water from "knee-deep" to "overhead." This means the people who are already poorer and more vulnerable are the ones facing the most dangerous, life-threatening depths.
The Analogy: Imagine a storm hits a neighborhood. If you just count "how many houses got wet," you might think everyone is equally affected. But if you look closer, you see that the rich houses on the hill only got a wet driveway (shallow water), while the poor houses in the valley got a submerged living room (deep water). The climate change "faucet" turned up the water just enough to drown the valley, but not the hill. If you only looked at the total number of wet houses, you'd miss that the poor got hit much harder.

The Bottom Line

The paper concludes that:

Population growth is the biggest reason why more people were exposed to the flood.
Climate change is the reason the flood became deeper and more dangerous in specific areas.
Looking at the whole picture matters: If you just say "climate change caused X% of the damage," you miss the fact that it disproportionately hurts the most vulnerable people in rural areas by making the water deeper, while population growth is the main driver for the sheer number of people affected.

In short: We have more people in the flood zone because the population grew, but the water is more deadly in some places because the climate changed. And the people suffering the most from that extra depth are often the ones with the least resources to cope.

Technical Summary: Impact Attribution of Climate and Population Change on Compound Flood Impacts from Tropical Cyclone Idai in Mozambique

Problem Statement
Extreme weather event impacts result from a complex interplay of climatic and socio-economic drivers. While traditional impact attribution studies have increasingly focused on the role of climate change, there remains a gap in frameworks that simultaneously attribute impacts to both climate and non-climatic drivers (such as population growth) and disaggregate these impacts across hazard severity and socio-economic vulnerability. Existing studies often fail to propagate socio-economic drivers to actual impacts or account for changes in exposure and vulnerability, particularly in data-scarce regions. Furthermore, aggregated attribution statements often mask the disproportionate effects of climate change on specific vulnerable socio-economic groups. This study addresses these gaps by analyzing Tropical Cyclone (TC) Idai, which made landfall in Mozambique in 2019, causing significant fatalities and affecting 1.85 million people.

Methodology
The authors employ an extended "storyline attribution framework" utilizing a globally applicable model chain to simulate compound flooding under factual and counterfactual conditions.

Scenarios: Four scenarios were constructed to disentangle the contributions of climate and population change since 1975:
1. Factual: Observed 2019 climate conditions and 2020 population data.
2. Counterfactual Climate: 2019 population data with climate trends (rainfall, wind, sea-level rise) removed to represent 1975 baseline conditions (6% reduction in rainfall, 4% reduction in wind speed, 7 cm SLR removal).
3. Counterfactual Population: 2019 climate conditions with 1975 population data.
4. Counterfactual Climate + Population: Both climate trends and population data reverted to 1975 baselines.
Hazard Modeling: Compound flooding was simulated using a reduced-complexity model (SFINCS) driven by boundary conditions from D-Flow FM (tide, surge, waves) and hydrological discharge from wflow. Meteorological forcing combined ERA5 reanalysis and IBTrACS best-track data. Flood maps were generated at 25 m resolution, considering depths $\ge$ 0.15 m.
Exposure Modeling: Population exposure was calculated by overlaying flood maps with Global Human Settlement Layer (GHSL) population data (100 m resolution, downscaled to 25 m). Socio-economic characteristics (wealth, settlement type, age, sex, education, household size) were derived from a global 2015 dataset (Ton et al.) downscaled to the flood resolution.
Attribution Metrics: The attributable change ( $A$ ) was calculated as the relative change in exposed population between factual ( $F$ ) and counterfactual ($CF$) scenarios: $A(\%) = \frac{F - CF}{F} \times 100$ .
Disaggregation: Results were disaggregated by:
- Flood Depth: Low (0.15–0.5 m), Moderate-to-High (0.5–1.5 m), and Very High (>1.5 m).
- Socio-Economic Groups: Defined by vulnerability (e.g., poor, rural, low education, elderly/infants, female, living alone vs. richer, urban, educated, working-age, male, multi-person households).

Key Contributions

Dual-Driver Attribution: The study expands the storyline framework to attribute impacts simultaneously to climate change and population change, demonstrating that population change is the dominant driver for TC Idai.
Disaggregated Impact Assessment: Unlike binary "exposed/not exposed" classifications, the study disaggregates impacts by flood depth categories, revealing that the attribution of drivers varies significantly depending on flood severity.
Socio-Economic Inequality Analysis: The paper attributes climate change impacts across different socio-economic groups, highlighting how aggregated metrics can obscure the disproportionate burden placed on vulnerable populations.
Methodological Expansion: The framework integrates counterfactual exposure conditions (population of 1975) alongside counterfactual hazard conditions, allowing for a more nuanced understanding of how drivers interact.

Results

Overall Attribution: In the factual scenario, approximately 250,000 people were exposed to compound flooding.
- Population Change: Accounted for 80% of the exposed population.
- Climate Change: Accounted for 9% of the exposed population.
- Combined Effect: The joint effect of both drivers accounted for 84% of the exposed population. The combined effect was less than the sum of individual effects due to population growth occurring in areas that only flood under climate change conditions.
Flood Depth Disaggregation:
- The fraction of exposed population attributable to climate change increased with flood depth (3% for low depths, 13% for moderate-to-high, and 24% for very high depths).
- Population change was the dominant driver across all categories but contributed most in absolute numbers to the moderate-to-high flood depth category (89% of exposed population in this category).
- Climate change primarily increased flood depths from moderate-to-high to very high in rural areas, and from low to moderate-to-high in urban areas like Beira.
Socio-Economic Disaggregation:
- When using a binary classification (exposed vs. not), climate change appeared to affect less vulnerable groups (richer, urban) more, largely due to high population density in urban areas where small flood extent changes affect many people.
- However, when disaggregated by flood depth, climate change disproportionately affected more vulnerable groups (poorer, rural). In rural areas, climate change shifted flood depths from moderate-to-high to very high, directly impacting vulnerable populations living there.
- This disparity is masked in aggregated assessments, which fail to capture the shift in hazard severity affecting specific groups.

Significance and Claims
The paper claims that its primary significance lies in demonstrating that spatial aggregation of impact attribution can mask the disproportionate effects of climate change on vulnerable socio-economic groups. By moving beyond binary impact classifications to discrete flood depth categories and socio-economic disaggregation, the study reveals that while population growth is the primary driver of exposure volume, climate change disproportionately increases the severity of flooding for the most vulnerable populations in rural areas.

The authors conclude that future risk management and attribution statements must be policy-informed regarding which drivers and metrics to include. They argue that understanding the interaction between climate and population change is essential for Mozambique, where population is projected to double by 2050, and that ignoring the distribution of impacts across socio-economic groups leads to an incomplete understanding of climate risk. The study also notes limitations regarding the coarse resolution of socio-economic data for Mozambique and suggests that mixed-method approaches combining quantitative attribution with qualitative causal analysis could further elucidate indirect drivers like deforestation and vulnerability dynamics.

The Main Result: It's Mostly About the Crowd

The Twist: It Depends on How Deep the Water Is

The Hidden Inequality: Who Gets the Worst of It?

The Bottom Line

More like this