Natural Hazards

Cost of natural disasters

The CRED presents the bill: the socio-economic cost of natural disasters.

The CRED presents the bill: the socio-economic cost of natural disasters.

Which type of natural disaster is the most frequent? And which one causes the largest economic losses? Which populations are mainly affected? What are the necessary steps to reduce natural disasters’ impact? If you have ever wondered about any of these questions, you’d be interested to know that there is an institute answering all of them with a series of reports and ad hoc publications.

We are talking about the Centre for Research on the Epidemiology of Disasters (CRED). The CRED is based, since 1973, at the Université Catholique de Louvain in Belgium and since 1980 it’s a collaborator of the World Health Organization (WHO). Their main goal? Study public health during a mass emergency as well as the structural and socio-economic impact of natural and technological disasters and human conflicts. They maintain the world’s most comprehensive database (EM-DAT) on occurrence and effects of technological and natural disasters from 1900 to the present day: more than 22,000 events and counting. [Read More]

Mapping population dynamics to advance Disaster Risk Management

Mapping population dynamics to advance Disaster Risk Management


Today we have the honour to introduce Sérgio Freire as our guest. Sérgio Freire is a Geographer, currently working as Scientific/Technical Project Manager at the European Commission’s Joint Research Centre (JRC), Directorate E. Space, Security and Migration, Disaster Risk Management Unit, based in Ispra, Italy. His main activities focus on developing applications of the JRC’s Global Human Settlement Layer (GHSL) in the context of disaster exposure, risk, and vulnerability analysis, including modelling population distribution at a range of spatial and temporal resolutions. Current activities also include global mapping and characterisation of human settlements, and developing satellite-based indicators to support monitoring of Sustainable Development Goals.



  1. When we think about disasters, we firstly mean natural hazards characteristics. However, potential harm comes even from vulnerability and exposure. Can you please explain to us what these elements are and which role they play in the risk equation?


In fact, natural hazards are ‘normal’ acts of nature that are part of the living planet that is Earth.

These only make the news and become disasters when they affect people (or property, systems) that display vulnerabilities to those specific phenomena. A strong earthquake in the middle of the Sahara desert may have little or no impacts due to scarce population and settlements, i.e., the absence of exposure. On the other hand, an earthquake of comparable magnitude occurring in cities of dissimilar countries may cause very different impacts and casualties due to the divergent structural vulnerabilities of built-up structures. However, for extreme events or hazards above a certain magnitude, exposure is a major driver of impacts.

Figure 1. Evolution of global population exposed to the highest seismic hazard, by decade. Bars refer to the total population in Modified Mercalli Intensity levels VIII to XII (right axis) and lines refer to percent population change relative to the previous period (left axis) (Source: Freire S., D. Ehrlich, S. Ferri, 2015. Population Exposure and Impacts from Earthquakes: Assessing Spatio-temporal Changes in the XX Century. Computer Modeling in Engineering & Sciences (CMES), SI: ‘Modeling of dangerous phenomena for risk mitigation’. Vol.109(2): 159-182)


[Read More]

Our first Interview is ready!

Our first Interview is ready!

Today we are happy to post our first interview and to thank our first interviewee, Paola Crippa for her contribution. The topic focuses on mortality from high concentration of particulate matter generated from widespread wildfires. This topic wants to be just the starting point to address another and broader theme: dealing with lack-of-data for research purposes in developing countries.

This will be inspired by one of the most recent researches published by Paola: “Population exposure to hazardous air quality due to the 2015 fires in Equatorial Asia” http://www.nature.com/articles/srep37074


1. Which problem did you address in your research?

Vegetation and peatland fires occur frequently across Equatorial Asia, as they are used to manage the land, clear vegetation and to prepare and maintain land for agriculture. Wildfires emit pollutants that can cause poor regional air quality and are extremely harmful to human health. As a result, each year thousands of premature deaths occur across Equatorial Asia. In fall 2015, these fires burned out of control in Indonesia as a result of the extremely dry landscape caused by strong El Nino conditions. In our study, we use a state-of-the-art air quality model (the Weather Research and Forecasting model with Chemistry, WRF-Chem) at high spatial-temporal resolution to quantify the impact of these fires on air quality and human health. We found that 69 million people were persistently exposed to unhealthy air quality conditions caused by fire emissions and that this pollution may have caused 11,880 (6,153–17,270) excess mortalities. Our results emphasize the need of a coordinated effort between scientists and policymakers to assess the impact of land use changes and human-driven deforestation on fire frequency, to possibly mitigate the impacts of these hazardous events on human lives.

2. Do mortality estimates from simulations actually agree with the corresponding real data?

We evaluated our model simulations relative to both ground- and satellite-based observations of aerosol properties and we are confident that our simulated results provide a realistic representation of the 2015 wildfires, and hence can be used to infer the impact on air quality and human health. We integrated our hourly maps of pollutant concentrations with population density data and estimated the number of people persistently exposed to unhealthy and hazardous air quality conditions during fall 2015 with respect to World Health Organization and Pollutant Standards Index guidelines. While these metrics gave us confidence in our assessment of population exposure, it was unfortunately not possible to validate our mortality estimates since no local hospitalization data were available for the period of interest.

3. No real data were available? This is certainly a strong limitation for the research community but also for those that deal with risk management. What is your position in this regard?

In order to estimate the number of premature deaths occurred as a result of exposure to degraded air quality conditions, epidemiological evidence to link pollutant concentrations and hospitalizations and mortality data are needed. Unfortunately, in Equatorial Asia, as well as in most developing countries, these cohort studies have never been performed, or at least those data are not available to scientists. In our work, we used exposure-response functions developed from studies conducted in Europe and United States where pollutant concentrations are much lower than those registered during the events we studied. Therefore, our mortality estimates are likely conservative. This is indeed a big limitation not only for scientists but also for policymakers when trying to reduce the negative impacts of natural hazards since no robust evidence of the magnitude of those events is available. If local governments would be able to collect, organize and release these data, this would allow scientists to better serve the community by providing better mitigation strategies.

4. Do other countries invest more in data collection allowing for a better coupling between simulations and ground-truth data?

In Europe and United States, epidemiological studies linking exposure to mortality and, most importantly, hospitalization data are easier to access. While still not as easily accessible as most publicly funded satellite and climate model repositories, we hope that Western governments would implement a standardized national or international database that can be used to produce considerably more reliable exposure maps. This would allow a better assessment of mortality in polluted areas such as London, but the level of exposure in less developed countries is on a different level of magnitude, with millions of human lives at risk, including children and elderly citizens. Since any extrapolation of Western data in these areas is problematic, the international community must invest in the development of local studies and data collection.

5. Do you think that simulations like yours can be useful not only in a post-disaster phase but as a risk prevention tool? 

One of the great advantages of using numerical models such as WRF-Chem is that they can be also used in forecasting mode, meaning that they can be used to predict where and how fast pollution would be transported from emission sources and consequently provide information for reducing population exposure. They can be also used to make projections as a function of emission scenarios. This is particularly important in regions subject to rapid land use change and human-driven deforestation, such as Equatorial Asia or South America. An example of successful integration of numerical model forecasts with mitigation strategies can be found in Santiago de Chile, where the government declares alert days based on numerical weather model forecasts of unhealthy pollution. This is the result of a close and constructive collaboration between scientists and policymakers.