Project Description

Air pollution remains one of the leading environmental health risks across Africa, yet many cities lack the monitoring infrastructure needed to provide timely information for public health interventions and environmental management. To address this challenge, AirQo has developed an AI-powered forecasting platform that predicts PM2.5 concentrations at both hourly and daily timescales.

The AirQo forecasting system combines data from AirQo's growing network of calibrated low-cost sensors with numerical weather forecasts, historical pollution measurements, temporal patterns, and machine learning algorithms to generate reliable air quality forecasts. The platform produces:

Hourly PM2.5 forecasts that predict pollution levels for the next 24 hours, enabling citizens and decision-makers to understand how air quality is expected to change throughout the day.
Daily PM2.5 forecasts that provide a seven-day outlook, supporting planning for public health advisories, outdoor activities, environmental management, and policy decisions.

The forecasting models are specifically designed for African environments and continuously improve as additional monitoring data become available. Each prediction incorporates meteorological variables such as temperature, humidity, wind speed, wind direction, atmospheric pressure, cloud cover, and precipitation, together with historical PM2.5 observations and temporal features. This allows the system to capture both local pollution sources and changing weather conditions that influence air quality.

To improve transparency and decision-making, the platform provides much more than predicted PM2.5 concentrations. Users receive forecast confidence estimates, expected pollution ranges, Air Quality Index (AQI) categories, health guidance, and interactive visualizations that make complex environmental information easy to understand.

The platform serves a wide range of users, including:

National and local governments for environmental management and policy.
Health agencies for pollution early warning and public health communication.
Researchers and universities for scientific analysis.
Businesses integrating air quality intelligence into digital products.
Communities and individuals seeking to reduce personal exposure to air pollution.

Built on a scalable cloud infrastructure, the forecasting system supports multiple cities and countries while delivering forecasts through APIs and interactive web applications. The platform automatically updates forecasts as new observations and meteorological predictions become available, ensuring users always have access to the latest information.

The forecasting capability is available through the AirQo AI Platform and interactive forecast map, where users can explore current conditions alongside hourly and daily PM2.5 predictions for monitored locations across Africa.

The project demonstrates how artificial intelligence, environmental sensing, and operational data science can be combined to deliver practical, real-time air quality intelligence that supports healthier communities and evidence-based environmental decision-making.

daily and hourly PM2.5 forecast

Overview

Details

daily and hourly PM2.5 forecast

Overview

Details