Climate & COVID-19 Outcome Modelling | Global | Public Health & Climate Intelligence

Lead Research Advisor

Prof. Roshini Sooriyarachchi,

Department of Statistics, University of Colombo ( Published June 2020 )

Problem Statement

During the early and most critical phase of the global COVID-19 pandemic, health systems worldwide faced extreme uncertainty due to the lack of statistically validated evidence linking weather conditions to COVID-19 transmission, mortality, and recovery patterns. Governments and donor agencies needed data-driven tools to guide healthcare planning and resource allocation.

Intervention Design

A global-scale statistical modelling study was conducted to quantify the effect of weather and air quality variables on COVID-19 outcomes, using advanced:

• Generalized Linear Mixed Models (GLMMs)

• Negative Binomial Modelling

• Post-cluster sampling methods

• Regional clustering analysis

The study evaluated how climate variability influenced mortality, recovery, and infection risk at a global level.

Total Program Value

Global public health analytics research (Data sourced from international COVID-19 reporting and climate datasets)

Target Population / Coverage

Global dataset (multi-region)

Key response variables:

• Number of deaths

• Number of recoveries

• Number of active cases (“at risk”)

Study period: Early pandemic phase – 2020

Climatological parameters analysed:

• Temperature

• Humidity

• Wind speed

• Air quality

Measured Outputs

Identification of strong regional clustering effects (Risk factor > 3, highly significant)

Statistical validation of:

• Air quality × temperature interaction affecting mortality (p = 0.0298)

• Air quality × humidity interaction affecting mortality (p = 0.0027)

• Temperature × humidity × air quality interaction affecting recovery (p < 0.0001)

• Temperature × wind speed × air quality interaction influencing infection risk (p = 0.0005)

Validation of robust, well-fitting predictive models via studentized residual diagnostics

Measured Outcomes

Proven climate–COVID-19 outcome linkages at global scale

Strong statistical basis for:

• Pandemic risk forecasting

• Climate-sensitive health planning

• Regional healthcare capacity allocation

Demonstrated that all four weather parameters significantly influenced one or more COVID-19 responses

Independent Verification

• Peer-reviewed academic publication

• International COVID-19 response datasets

• University-based statistical validation

• Independent residual diagnostics confirming model reliability

Sustainability & Long-Term Impact

• Enables climate-linked early warning systems for dengue outbreaks

• Informs donor-funded epidemic preparedness 

• Supports health system resilience and climate adaptation strategies

Donor Relevance

This case study demonstrates how climate intelligence and advanced statistical modelling can:

• Improve pandemic preparedness

• Optimise healthcare resource allocation

• Reduce mortality and system overload risk

• Strengthen climate-resilient public health infrastructure

Institutional Assurance Statement
“This case study is supported by peer-reviewed global research, advanced statistical modelling, and independent academic verification.”