Weather Forecasting
The Climate Centre is dedicated to understanding and mitigating the impacts of weather events to improve preparedness and response strategies, ultimately helping communities better withstand and recover from these events. We do this through the development and maintainence of weather models and applications, primarily our EarthEye application and our Earth System Weather Model (ESWM).
We originally developed EarthEye as a smartphone application for monitoring weather conditions. However, this project represented an overextension of our capacity for multiple reasons, primarily due to a lack of staffing and funding, and we have instead decided to maintain EarthEye as a web application and integrate it with the development of our weather forecasting model, the Earth System Weather Model (ESWM). You can interact with our EarthEye application below, displaying sample forecast average temperature data from July 2025. You are able to do a text search for a handful of large cities, and zoom and pan the map to look at different locations across the globe. The average temperature, latitude and longitude will be displayed as you move your mouse cursor to specific locations. Please note that EarthEye is best viewed on a desktop with a sufficiently wide screen, otherwise the display will appear too cramped.
The Earth System Weather Model (ESWM) is an AI-driven system for short-term meteorological estimates 5-10 days ahead. It enables quick processing of large datasets, and acts as a primary input to the Global Weather Outlook, an interally produced report, which provides estimates of temperature conditions compared to baseline, by various geographies. We hope to develop and expand the Earth System Weather Model as an AI weather forecasting system to help support understanding of weather events with respect to climate change. The ESWM uses initial conditions from the GDAS dataset which is based on observational records, and the GFS which is a numerical weather prediction model, and uses AI algorithms based on FourCastNet to predict future weather states for temperature, rainfall and wind across the globe.
*Please note that our weather forecasting and model should not be used as a replacement for other weather models or forecasts. All weather forecasts should be considered together when planning for future events.
We have also previously produced the ERA5-derived GHCND-adjusted Temperature Reanlysis (EdGaTR) product as a global temperature dataset. However this is not an actively maintained product and is not accessible for use either internally or externally. We also undertake climate forecasting, which you can read more about here. While data is used across many of our projects, our main data analytics program is the Climate Update series of projects, which you can read more about here.