Skip to content

Technical Case Study

From: Kenyan Ecosystem on Air Pollution and on Health and Climate Change and the Netherlands Organisation for Applied Scientific Research (TNO), in affiliation with the Dutch Ministry of Health, Welfare and Sport

Approach

To inform climate vulnerability assessments and national adaptation plans, climate risks to health need to be considered in an integrated way, as vulnerable populations are often exposed to multiple climate-related hazards simultaneously, with compound effects on health.

This case study in Kenya outlines an integrated approach to assess population health risk from two common climate-related health hazards: heat and air pollution. Weather (heat) and air pollution data can be modelled, integrated and interpreted by:

  • Converting operational meteorological data into a heat stress relevant parameter for workers and other vulnerable groups.
  • Using globally available meteorology and emission inventory data from the EDGAR database into a dispersion model to calculate exposures to pollutants.
  • Mapping co-exposure to heat and (climate change induced) air pollution across the area of interest (in this case, counties in Kenya) to support quantitative climate vulnerability assessments, including to identify vulnerable population subgroups and preventative measures.

By working with relevant stakeholders, sustainable ways to generate and use information on climate risks to health can be identified and adopted, in order to inform national vulnerability assessments and adaptation planning. 

Key concepts

  • Heat stress: net heat load from metabolic heat, environmental factors, and clothing which results in an increase in heat storage in the body
  • Wet bulb globe temperature (WBGT): a heat stress index representing the thermal environment to which an individual is exposed. Can be used to establish the presence or absence of heat stress.

Why take action?

Climate change affects human health through various interrelated pathways such as extreme weather events, heat stress, air quality, water quality and quantity, food security and safety and vector distribution1. Low- and middle-income countries (LMICs) face more severe impacts because of current vulnerabilities, including underlying levels of disease, limited infrastructure and services, weak economies, and poorly informed governance and decision-making processes. Moreover, the overall health burden of climate change is almost certainly underestimated because of the complexity of the causal pathways between climate-related hazards and health outcomes and a lack of robust surveillance and monitoring systems in many countries2 .

Under the UN Framework Convention on Climate Change, countries are developing National Adaptation Plans (NAPs), which need regular updating and refinement in the coming decades. To support these, guidelines on ‘’Climate change and health: vulnerability and adaptation assessment’’ (V&A) were published by the WHO2 . The aim is to provide basic and flexible guidance on conducting a national or subnational V&A assessment of current and future vulnerability to the health risks of climate variability and change, and the policies, programmes and capacities of health systems that could increase resilience, taking into account the multiple determinants of climate-sensitive diseases and health outcomes.

In the WHO guidelines, several health outcomes were selected that need to be addressed to protect population health: mortality due to heat, air pollution, infectious disease and undernutrition.  Effects on worker productivity and ability to provide livelihoods are specifically mentioned. LMICs especially struggle to provide quantitative risk assessments, which are essential to prioritize their adaptation actions. Across countries that provided initial NAPs, many still lack quantitative assessments of health risks.

In Kenya, plans for the vulnerability assessments to inform the input for national adaptation plans are being made led by the Kenya Ministry of Health, Division of Environmental Health.

TNO, in collaboration with the high-level visiting delegation in June 2024, developed a seed money project that was selected and funded by the Dutch Ministry of Health, Welfare and Sports (VWS). The project prototyped assessing the risk of heat and air pollution taking into account population vulnerabilities, by using monitoring data in combination with models over a great part of Kenya. We decided to look at heat stress and outdoor air pollution together because these have shared underlying drivers but may also synergistically increase health & performance impacts beyond the sum of individual effects.3

Figure 1: Average hourly heat value (WBGT) per county in Kenya, compared to World Athletics Organization Guideline values

Key messages

The combined air pollution – heat health risk model developed in Kenya demonstrated that:

  • Heat stress risk maps showing hourly risk of heat stress at 7 by 7 km resolution can be developed by combining European Centre for Medium-Range Weather Forecasts data with a heat risk assessment framework for wet bulb globe temperature (WBGT, as outlined in ISO7243 and ISO9920). WBGT takes into account all relevant environmental factors relevant for human health: temperature, humidity, wind speed and radiation. Figure 1 above shows how hourly heat stress risk profiles can be compared to existing World Athletic Guideline Values.
  • Specific assessments for vulnerable groups and potential of preventive strategies can be made by taking into account the contribution of heavy physical (work) activity,  clothing, and timing of activities. Figure 2 below shows the contribution of performing heavy work such as farm work to the risk of heat stress (B versus A: % hours above the WBGT reference value) and how avoiding work during the hottest hours of the day can reduce the risk (C: % decrease in the hours above the WBGT reference value by working between 7-11h and 15-19h instead of 9-17h).
  • Air quality models including source origins can be implemented at national level (in our case Kenya) as shown in Figure 3. The validation of the implementation against sparse observations is promising and provides strong guidance towards improvements. The model provides concentrations together with the source origins; who is exposed where to what source of pollution. Understanding the source contributions as actionable information supports making informed emission reduction measures.
  • Results of combined monitoring and modelling on heat and air pollution can be used to inform a quantitative climate vulnerability assessment and climate adaptation planning in several ways. For more details see “in practice” section below. What is needed to implement such an intervention was discussed during a workshop with panel discussion as part of the 3rd Health & Climate Conference in Machakos, Kenya. Together with Ministries, county representatives, and relevant (research) institutes, the ecosystem and route towards vulnerability assessment was explored.
A
B
C

Figure 2: Example of heat stress value (WBGT) exceeding safety reference in counties of Kenya, according to behavior and time.

 

How to get started

Figure 3: Modelled air pollutant concentrations in Kenya

In Kenya, plans for the vulnerability assessments to inform the input for national adaptation plans are being made led by the Kenya Ministry of Health, Division of Environmental Health, with support from WHO, the Dutch Ministry of Sports, Welfare and Health and the United Kingdom Foreign, Commonwealth & Development Office (FCDO). A group of healthcare professionals, policymakers, public environmental health specialists and scientists have been trained by the WHO and Ministry of Health to be able to educate various stakeholders on assessing climate risks, using vulnerability and adaptation tools and developing the Kenyan Health National Adaptation Plan.

Ongoing national initiatives on greening the health system and improving indoor air quality can likely serve to inform a vulnerability assessment of qualitative nature, through expert qualitative inputs. Our concern was that quantitative considerations, namely, exposure pathways described in the WHO guidelines, will not be taken into account in the first phase of the vulnerability assessment. Moreover, most analyses remain qualitative. With this project, we showcased that with existing data and models, we can support a quantitative vulnerability assessment for two climate-sensitive interrelated health hazards.

Several key learnings can be drawn from our experience for the development and implementation of a quantitative climate-sensitive health risk assessment.

Key learning: Timely access to meteorological data

For an accurate assessment of environmental heat and air pollution variables, it is important to have access to meteorological data at any time. This concerns both historical data and forecasts. Apart from a few observation stations, data accessibility in Kenya is not considered to be sufficient by stakeholders. Moreover, for evaluating the effects of heat and for developing policy instruments, it is necessary to have a data stream that is continuous in space and not only at observation stations. For this, models can be used in combination with the most high-resolution weather monitoring data. For the distribution of air quality, the same applies, since to make good quality predictions, output from a weather model is needed. In this project, the global weather model of the European union was chosen. TNO showed that these data are unlockable for Kenya. However, it is known that for heavily hilly terrain with mountains, the resolving power (resolution) of this model is not always sufficient. Input from local meteorologist, e.g. Kenya Meteorological Department (KMD), is desired.

Key learning: Combining global with local data for high-quality inputs for air pollution

For air pollution modelling, high quality emission input is crucial. For this first implementation project a globally accessible database of anthropogenic emissions of greenhouse gases and air pollution on Earth (EDGAR) was used, along with a leading chemical transport model for air pollutants (LOTOS-EUROS). During a workshop with national stakeholders, relevant Kenyan parties that collect or have access to local emission inventories or are interested to set up mid-range air quality networks were identified. It is of utmost importance that these stakeholders are coordinated and onboard and provide their data to further train the models and when operationalizing air quality models over Kenya.

Key learning: Training modelers and data interpreters

Modelling was done by the TNO team based on the operating LOTUS EUROS model and WBGT modelling framework, and experience in Europe. For a sustainable and routine operational system, a local team of modelers and data interpreters need to be trained focusing on Kenya and East Africa. These modelers should have academic skills but work in institutions that are tasked with the responsibility of air quality modelling and work in permanent positions.

Key learning: Establishing a sustainable operational system with key stakeholders

A sustainable and routine operational system benefits the entire Kenyan ecosystem and requires collaboration between different stakeholders to optimize output and outcomes from the actionable information on adaptation, mitigation and policy formulation. Relevant stakeholders include:

  • Ministry of Heath, Division of Environmental Health Kenya, including the Kenya Air Pollution center of Excellence/ Clean Air Africa KEMRI
  • Ministry of Environment, including the National Environment Management Authority (NEMA) and the Kenya Meteorological Department (KMD)
  • Ministry of Energy and Petroleum
  • County Governments
  • Kenyan National Public Health Institute
  • AMREF Healthcare Africa
  • KEMRI Wellcome Trust
  • Kenya Industrial Research and Development Institute (KIRDI)
  • Moi University
  • Kenyatta University
  • Private Sector, notably, Air Quality Systems East Africa (AQS)

Key learning: Striving for local coordination

Coordination of efforts between the different Ministries as well as the capacity within the Ministry that is tasked with coordination and has the orchestrator role needs to be strengthened. The Kenyan stakeholders are still very much depending on outside support and therefore not acting in a holistic and coordinated manner. There is a lot of competition between the different stakeholders for the same funding. This is also influencing prioritization of the different topics.

Key learning: Scaling multilateral funding

This demonstrator project was funded by the Dutch Ministry of Health, Wellbeing and Sport (VWS). The initial funding was 72K Euro to prototype and demonstrate that this is possible and actionable input for different key stakeholders and informs the national adaptation plans. To ensure this is up and running on a quarterly basis to provide input and monitor over time needs collaborative efforts from our bilateral support for the Netherlands and the UK as well aligned multilateral support to GFATM and the Climate Fund. As TNO, we can support a sustainable and routine system with another investment of 400K Euro to:

  1. Providing assistance in harmonizing emissions inventories in East African countries along the lines of a project that focused on South America.
  2. Setting up the modelling system including relevant input data for several east African regions.
  3. Giving in-depth training of LOTOS-EUROS and its input (land-use, meteorology, and emissions) to become independent users. The incentive is a long term cooperation, hence the training concerns young professionals and academics. We have positive experience with the model of knowledge transfer and capacity building initiatives in Colombia, Uruguay, Portugal, and Croatia.
  4. “Train the trainer”, in our vision, Kenya will act as a hub in the region for air quality.

Giving online training sessions to cover basic operation (simulations) of the open source LOTOS-EUROS for satellite countries.

Tracking progress

Progress is measured on a quarterly basis as real time information might be a level up. The aim of setting up this system should be to have quarterly modelling with actionable data provided in a dashboard with combined air quality and heat risk assessments done that can allow for monitoring overtime. Stakeholders will meet to evaluate the progress during in-person meetings in February 2025 aligned with the 15th KEMRI Annual Scientific and Health Conference (KASH).

In practice

Routine practice should be that these dashboards are discussed by all Ministries in : 1) risk assessment (who is exposed, who is vulnerable, where?); 2) impact and health assessment studies linking the risks to health outcomes taking into account the complexity of air pollution and health; 3) surveillance and monitoring systems tracking the risks and related health effects; 4) early warning systems (dashboards or apps)

More information

For more information, please contact Yvette Fleming (yvette.fleming@tno.nl) and Bas Henzing (bas.henzing@tno.nl) from the Netherlands Organisation for Applied Scientific Research (TNO): 

Key resources

References

  1. World Health Organization. (‎2021)‎. Quality criteria for health national adaptation plans. World Health Organization. https://iris.who.int/handle/10665/339454
  2. World Health Organization. (‎2021)‎. Climate change and health: vulnerability and adaptation assessment. World Health Organization. https://iris.who.int/handle/10665/345968

  3. Rahman et al., The Effects of Coexposure to Extremes of Heat and Particulate Air Pollution on Mortality in California: Implications for Climate Change. Am J Respir Crit Care Med. 2022 Jun 21;206(9):1117–1127. doi: 10.1164/rccm.202204-0657OC

 

Scroll to top