Integrated air-quality decision platform

The problem

Air-quality agencies usually have a monitoring network, access to satellite products and, sometimes, a forecast model, but each lives in its own system. Staff spend time stitching them together by hand, gaps in station data go unnoticed, and when a ministry asks what a policy change would do, there is no quick way to show it. The result is a dashboard that reports the past rather than a tool that supports decisions.

What you get

• Automated ingestion pipelines for station observations, selected satellite products and model output, with QA/QC and data-freshness monitoring
• A spatial database (PostgreSQL/PostGIS) and gridded data services (THREDDS/GeoServer) as the single source for all layers
• Fused views: observations overlaid on satellite and model fields, with simple model-versus-observation comparison
• Threshold-based alerts for exceedances, configured with the client's own standards
• A scenario module that presents pre-computed model runs (for example baseline versus emission-reduction cases) side by side
• A web dashboard, deployment on the client's servers, documentation and staff training

How it works

1. 1

   Scope

   Agree the decisions the system must support, the area and resolution, and which observation networks, satellite products and model runs are available or realistic to obtain.

2. 2

   Data backbone

   Build the ingestion pipelines, database and data services first, with QA/QC and monitoring, so every later feature sits on reliable, current data.

3. 3

   Fusion and alerts

   Align observations, satellite and model data in space and time, add comparison views and threshold alerts, and review them with the agency's scientists.

4. 4

   Scenarios

   Integrate scenario runs from an existing or newly configured chemical transport model workflow. Scenario scope is set by the modelling capacity actually available.

5. 5

   Deploy and hand over

   Deploy on the client's infrastructure, document operations, and train staff to run and extend the system.

Tools

• Python (xarray, Pandas)
• PostgreSQL/PostGIS
• THREDDS Data Server
• GeoServer
• Google Earth Engine
• WRF-Chem / HYSPLIT output
• Django REST Framework
• React / Next.js