Useful. Beautiful. Data.

The Met Office Informatics Lab combines scientists, technologists, and designers to make environmental science and data useful. We build prototypes to try out new technologies and techniques to push the boundaries of science, technology and design.

See our work

This week

The things we are looking at right now.


The different things we've done.

The Bell House

Working with local artists to re-imagine climate data

Facebook Chatbot

Developing and deploying a public-facing weather chatbot on Facebook.

Global Temperature Visualisation

Visualising global temperatures through time in 3D


Why scientific and technological research are now intertwined

Labby the Slack Bot

We got bored doing tech odd-jobs, so we made a chat-bot called Labby instead

From Models to Intelligence

Weather models contain a wealth of information. How much can be extracted and put to use?


We all think we understand weather symbols, but does everyone agree?

A weather forecast you can talk to

"Alexa, when should I have a barbeque?" "Thursday should be great for a barbequeue!"


Our illuminating work on the Paignton Zoo Great Big Rhino Project.

Augmented reality sandpit

We've seen some great sandpit-worlds. We're going to build one with weather!

Science Capabilities Pipeline

Using open source tools for effective internal communication

Sky View 360

Training meteorologists with virtual reality

Fly through model fields

Models create rich pictures of the atmosphere, let's look at them in detail.

Using traffic cams for ML

Scraping, building, analysing

Weather and Climate Bot

Working with Microsoft to create a big data weather and climate bot

Our Blog

Keep up-to-date with what's happening in the Lab.

7 February 2018

Instant access to auto-scaling personal Python clusters

The prototype library Daskernetes brings together Dask and Kuberenetes, allowing you to automatically create your own custom Dask cluster

19 December 2017

ChatOps - Automation via chat

Chatting to your IT infrastructure can be useful.

14 December 2017

Iris, dask, distributed and SPICE

Massively parallel processing of large datasets using Iris.

Contact Us

Leave us a message and we'll get back to you as soon as possible.