The New Forces in Migration and Displacement

A data explainer project by Evan Baldonado, Jierui Fang, and Joxemari Gallastegi for CS 448B (Winter 2023).

While conflict-related displacement usually is top of mind when the topic of migration comes up, the movement of people caused by climate-related factors has become an increasing focus of discussion in recent years. Whether it is a drought, a flood, or a wildfire, natural disasters and hazards can disrupt communities, devastate livelihoods, and cause widespread relocation. According to the U.N. High Commissioner for Refugees, “natural disasters force an average of 21.5 million people from their homes around the world” (Watson, 2022). Many of the migratory movements are within a country, or internal displacements, as people are forced to abandon their homes in a crisis and seek more stable shelter locally. In this article, we examine patterns and insights in data from the Internal Displacement Monitoring Centre on disaster-related data from the beginning of 2008 to the end of 2021

A Global Phenomena

Displacements caused by natural disasters are not just confined to particular countries or the Global South. Explore the interactive map below to examine the planetary scale of the issue.

2008

Figure 1. Zoom and pan on the map to examine the data more closely. Use the slider to choose which year to display and hover over a country point to see more information. Note that data has been aggregated by country location and are not precise to the location of specific disaster events. The color of each point is shown as the hazard type that caused the most displacement in that country for that particular year.

Where are displacement-causing disasters? The map above shows the global impact of disaster events from 2008–2021, aggregated by country over the course of the year, color-coded by hazard type, and sized to the amount of people internally displaced. With the differential effects of climate change around the world, that brings to mind the question: Will climate change alter or has it already altered the prevalence, severity, or distribution of these disaster displacement events?

Top Trending

Over the past decade, we have seen an increase in frequency of the hazard or disaster types that cause the largest amount of internal displacement. Explore with the dropdown below to see density patterns for the top 4 most common ones in terms of incident count.

Figure 2. The relative diameter of each circle indicates the amount of people impacted for that particular incident. The vertical position of the circle also corresponds to that number of internal displacement. The horizontal position plots the incident along a timeline from the beginning of 2008 to the end of 2021.

Separated out by type, we can see that floods are chief among the culprits of internal displacement followed closely by storms, which are closely related in terms of cause and effect. Wet mass movements, like mudslides, landslides etc., are also among the top number of incidents as a byproduct of the above two hazard types. On the other end of the spectrum, wildfires can also be seen as increasing in frequency though with proportionally less impacted people than the water-related climate events. Less common are droughts, which have a longer timescale and may not be as easy to measure the immediate displacement effects of.

From the visualization, we can also see another potential pattern in terms of data collection. Most incidents in the earlier years tend to have a start date of January 1st, causing an interesting line pattern in the circles or larger circles as individual disasters were aggregated into one. This could owe to more attention being paid, and therefore more rigorous and accurate data collection being done over the years. Nevertheless, there is a clear increase in the density of incidents over the course of more than a decade.

Disproportionate Impacts

Which are the countries that suffer more from natural disasters? Are some that might be prone to constant affection?

Based on the data, we can draw several conclusions. We will rely on small multiples to give evidence. Firstly, contrary to our expectation, countries in the sub-Saharan range of Africa that are affected by global warming and droughts do not appear among the top 5 ranks, showing that these countries are largely unaffected by nature-driven internal displacement, compared to other countries. For instance, Mozambique, South Sudan, and Chad have experienced less displacement than some other countries.

Figure 3. Proportion of population affected by disaster events for representative nations typically perceived as being affected by drought or global warming related displacement

Secondly, small and highly populated islands tend to show a disproportionate amount of displacement relative to their population. For example, Dominica, Tuvalu, and Vanuatu are micro-countries that have experienced significant hikes in displaced populations due to floods and storms.

Figure 4. Proportion of population affected by disaster events for representative island nations that faced significant spikes in impact.

Islands face unique challenges that make them more vulnerable to natural disasters and displacement.

a) They are often more vulnerable to storm and flood-related natural disasters such as hurricanes, cyclones, earthquakes, and tsunamis, which can cause displacement and disrupt the lives of the affected population in a disproportionate way.

b) Islands also tend to have limited land and resources, which can lead to overcrowding and competition for resources, such as housing and employment in a highly dense location within the island. This can exacerbate displacement in the event of natural disasters or other crises, with exceptional challenges in terms of preparedness, response, and recovery, which can exacerbate the impacts on their populations.

c) Islands often have unique social, economic, and cultural characteristics that can affect the frequency and severity of displacement. For example, many island communities rely heavily on tourism, fishing, or agriculture, which can be negatively impacted by natural disasters and other crises, leading to displacement.

Overall, a combination of natural, social, and economic factors makes islands more vulnerable to displacement, which may explain why they have a higher proportion of internally displaced persons relative to their population compared to other countries. However, when it comes to impact, the micro islands are not meaningful in absolute displacement numbers.

This leads to our third conclusion. There are highly populated islands that show constant and relatively high displacement numbers, which leads to an enormous amount of affected populations, such as the following. Despite having >10 million inhabitants, some islands show brutal impact numbers.

Figure 5. Proportion of population affected by disaster events for representative island nations that see consistent impact from displacements.

Thirdly, highly populated islands are also prone to constant and significant displacement. Haiti, Cuba, and the Philippines are countries that are situated in regions that are frequently affected by severe weather conditions such as tropical storms, hurricanes, earthquakes, and typhoons. These countries share a common vulnerability to the impacts of natural disasters due to their limited resources and infrastructure.

Stories Under Water

Based on the visualizations above, it is apparent that floods and storms represent the vast majority of disaster events and disproportionately impact smaller nations. According to a March 2023 article by Isabella O'Malley (Associated Press), “global floods and droughts [are] worsened by climate change.” As human-caused climate change continues, people around the world will be at increasing risk of displacement due to the after effects of floods and storms.

Flood and Storm Related Displacements in 2008

Figure 6. Zoom and pan on the map to examine the data more closely. Note that data has been aggregated by country location and are not precise to the location of specific disaster events. The map automatically cycles through time and is viewed through the timescale of years to account for potential differences in data collection methods as noted in Figure 2.

Who is going under water? The map above shows flood and storm-related displacement events aggregated by country and sized by the amount of internal displacement caused. As the map cycles through the years, highlighted circles call out specific events and elaborate on the stories surrounding them.

Ripples and Reflections

These conclusions suggest that while both natural disasters and displacement have complex and diverse sources and ripple effects, there are trends emerging that show chief causes and populations that are disproportionately impacted. Specifically, island nations and other countries that face water-related disaster events like floods and storms encounter unique challenges in addressing them. By understanding the factors that contribute to displacement, we can better prepare for and mitigate the impacts of climate change and natural disasters on affected populations.

Data and methods

This section briefly describes where and how we obtained our data, as well as how we processed the data:

Our primary data is from the Internal Displacement Monitoring Centre’s dataset tracking displacement-causing disasters from 2008–2021 here. The data was processed with several different methods for the various visualizations including, sorting by type and date, reformatting of capitalization of hazard types to be consistent, and cleaning up of internal displacement numbers when there was a null value. Through exploratory data analysis, we discovered that the accuracy of the start date for many events could be in question as many events in the earlier years had a start date of January 1st and there were several events that were quite large in that pattern. This suggested earlier data collection methods might have aggregated disaster events of the same type to begin on the first day of the year or lacked specificity to a specific date within a year. We solved for that problem in our visualization by specifically abstracting to year when timing was a central focus of the visualization.

For map-based visualizations, the original dataset did not include specific latitude or longitude coordinates for each disaster event, meaning there was no way to drill down an event to the specific location that it happened in. We addressed this by aggregating the events by country and then plotting points at the average latitude and longitude coordinates of the given country. This latitude and longitude data came from a GitHub Gist by tadast.

For the small multiple visualizations that required knowledge of each country's respective population, we supplemented our dataset with population data numbers from OECD and World Development Indicators.

Text from the descriptions of flooding and storm displacement events was taken from various online sources including, Wikipedia, LA Times, NASA Earth Observatory, ReliefWeb, and other online sources (1, 2, 3, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13).