From Cloud Classifying to Climate Change Clustering

Cloud Classes Last year, while browsing in the The Quaker Bookshop at Friends House on Euston Road in London, I happened across a copy of The Invention of Clouds : How an Amateur Meteorologist Forged the Language of the Skies¹ (Hardcover) by Richard Hamblyn. It's a biography of Luke Howard, a Quaker pharmacist and amateur meteorologist, who created the system of cloud classification² that we still use today³.

In southeast England, we've had lower-than-average rainfall for the past two years. This may have been part of the motivation for the BBC to focus on climate change. A few weeks ago, I joined the 100,000+ others who have joined the BBC-sponsored climate change simulation experiment. As the climate model (specifically the Met Office's Hadley Centre Coupled Model, version 3) is running in the background, soaking up those unused clock cycles on my computer, I can have the model display one of four climate variables, or ‘facets’, it's calculating: temperature, pressure, precipitation, or cloud cover. BBC Climate Change model screenshot The plan is to get as many computers as possible to run this climate model.

The simulation starts with weather conditions as they were at the beginnig of 1920. It calculates, in half-hour steps, for eighty simulated years, what the climate state would be until the simulated year 2000. At the simulated millenium, the program compares the calculated climate state with the actual climate state. If the two states are close to each other, the assumption is that the simulation was fairly accurate and carries on calculating until simulated 2080. The result of my simulation will be plotted, with the results of the others, on a scattergram4. The rationale is that the experimenters will be able to assess the quality of the simulations based on where the results cluster.


1. I've been a sucker for this kind of book ever since I read Dava Sobel's Longitude : The True Story of a Lone Genius Who Solved the Greatest Scientific Problem of His Time (Hardcover).

2. See also cloud types.

3. For "How-Tos", see How to Identify Cloud Formations, and How to Identify Cirrus Clouds, and so on.

4. See also dendograms and cladograms