Defining and Predicting Heat Waves in Bangladesh

New research shows that in Bangladesh, heat wave predictability exists from a few days to several weeks in advance, which could save thousands of lives.

In the United States, extreme heat events have killed more people in the last 30 years than has any other weather-related phenomenon. In Europe, at least 136,835 people died due to heat-related health complications between 2000-2016, which represents more than 87 % of all disaster-related deaths in that area.

Because of heat’s danger, the U.S. and many other countries in the mid-latitudes have invested in warning systems, awareness campaigns, cooling centers and other measures to lessen the human health toll.

However, the thresholds that trigger such actions can vary by country, state or even city. A threshold incorporates temperature and sometimes other climate variables, such as humidity. In order to avoid the threshold being triggered solely by high temperatures, it’s also often based on how conditions have affected people in the past by incorporating mortality data. This ensures that warning systems — and the resources spent to enact them — target the exact conditions that put people’s lives at risk.

Dhaka, Bangladesh. Flickr: mariusz kluzniak.

In tropical countries such warning systems are rare, in part due to the assumption that people living in the tropics are more acclimated to hot weather and thus less likely to be affected by extreme heat events. However, heatwaves are also likely taking a high toll in developing countries and not getting the attention they deserve, according to international disaster reporting. This is partly because quality health data are harder to come by in countries with fewer resources.

The health data constraint also affects the warning systems themselves: most warning systems that do exist in developing countries are based solely on weather data.

Hannah Nissan, a postdoctoral research fellow at the International Research Institute for Climate and Society, is trying to fill in this information gap. Nissan is the lead author on a new paper that proposes a heat wave definition Bangladesh – a first step towards the creation of a heat warning system in a country that currently doesn’t have heat wave forecasts.

Before defining a heat wave, however, Nissan and her colleague Katrin Burkart at Columbia University’s Mailman School of Public Health first examined the health impacts of heat in Bangladesh. They found that heat waves do indeed lead to more deaths in the country. They estimated that during an eight-day heat wave in 2008, for example, at least 3,800 people died from the excess heat. Two-thirds of those killed were 65 or older, indicating that the elderly are especially vulnerable to heat waves.

Heat-related illnesses are also a significant issue, but they are much harder to investigate because of a lack of health data. Evidence from around the world points to a suppression of economic productivity during heat waves, and to occupational heat health problems in sectors working outdoors as well as in industry. Similar research has so far largely been lacking in Bangladesh, but nearly half of Bangladesh’s workforce is employed in the agriculture sector and many work in factories, with the garment industry employing 4 million people — over half of whom are women.

Once the relationship between heat and human health is established for an area, as Nissan and Burkhart have recently done in Bangladesh, climate scientists can look for opportunities to predict extreme heat events in a way that can be useful to decision makers.

In their new paper, Nissan, Burkart and other colleagues from IRI and the Red Cross Red Crescent Climate Centre propose a heat wave definition for the country and assess the potential for heat wave forecasts across various timescales. Using their definition, the authors calculate that death rates rise by about 20% on the days that meet or surpass the threshold. Their threshold definition is also based on the ability of the heat wave conditions to be predicted. In this case, the paper’s authors found potential for predicting heat waves a few days to a few months ahead of time by looking at wind conditions, precipitation and soil moisture.

If an early warning system is in place, decision makers can take actions with timing that makes sense given the forecast period, as shown in the graphic below.

Examples of preparedness activities to reduce the health risks of extreme heat in response to short range (up to 1 week), mid-range S2S (1 week to 1 month) and seasonal (3 months or longer) forecasts.

Recognition of the enormous human cost exacted by heat waves underpins a growing momentum in the international community to support the development of heat action plans and early warning systems worldwide. IRI is partnering with the Global Heat Health Information Network (GHHIN), which will launch later this year with South Asia as a key focal region. The GHHIN will provide coordination and support to governments, professional partners and researchers to build such policies and to develop the evidence needed to enable them.

Ahmedabad, a city in Gujarat province in western India, has been developing a heat early warning system for the last five years, with actions including keeping gardens and parks open all day to increase shade access, public messaging using a variety of traditional and creative formats, identifying areas to install water dispensers and implementing longer-term initiatives such as those to install cool roofs.

While no such early warning system exists yet in Bangladesh, the work of Nissan and colleagues could help pave the way for one, particularly given their finding that heat wave predictability exists from a few days to several weeks in advance. Weather forecasts can be used to forecast if a heat wave is imminent in the next few days or week, and climate timescales could give increased lead time for actions that are hard to do with just three days notice, such as retraining medical professionals and public awareness campaigns. Forecasts for heat on the sub-seasonal scale are still mostly in the experimental realm, but some scientists have assessed heat wave predictability at this timescale in Europe, said Nissan. They found that high mortality could be foreseen in some areas up to three months ahead of time.

Nissan and her colleagues can also now use the climate-timescale predictors they identified for heat wave frequency to investigate how heat wave risks could be changing under climate change, and to develop seasonal forecasts.

“In a changing climate, heatwaves are among the most rapidly rising risks, especially affecting the most vulnerable,” said Maarten van Aalst, director of the Climate Centre, an adjunct researcher at IRI and a co-author of the study. “There are many simple actions we can take to reduce the impact heatwaves have on people, but we rely on good warning systems, and public awareness. Studies like these are critical to fill these gaps, especially in highly vulnerable countries such as Bangladesh, where the need is highest.”

“In lots of places, health data are unavailable to do the type of analysis we’ve been able to do in Bangladesh,” said Nissan. “In other areas, we might know that heat is dangerous but not be able to predict heat waves with any skill. But we have the chance in Bangladesh to develop useful forecasts to enable early action and reduce the number of deaths. There is an opportunity here for Bangladesh to become a pioneer in the region for evidence-based climate services that improve public health.”

Support for this research provided by:

NOAA – International Research and Applications Project 

The Earth Institute

UK Fulbright Commission

German Research Foundation