In Search of the Hundred-Year
Forecast with PV-WAVE^®

Climate prediction	Large amounts of data can be manipulated and presented as visual images for analysis

Joint funding by the Department of the Environment and the UK Meteorological Office has allowed the creation of the Hadley Centre for Climate Prediction and Research in Bracknell. Its mission is to predict climate changes over the next 100 years. With terabytes of data to be analyzed, the centre has turned to visual data analysis to speed up the process.

Output from climate simulations runs on supercomputers combined with data from satellites, and data records going back many decades presents the Climate Prediction Centre with terabytes of data (that's data measured in trillions. The problem is not finding such data -- it's finding methods to analyze it.)

When scientists want to analyze data as voluminous as this, there isn't a simple formula that can be used to find a simple answer. Even the most powerful supercomputer in the world will not pop out a simple yes/no answer to the most critical question ever asked -- Are we heading for a climate change or not?

The Hadley Centre would, of course, be the first to know if it were that straightforward. At their disposal is one of the most powerful supercomputers in the world - a Cray^® Y-MP.

Instead, the climatologists must use advanced visual data analysis (VDA) software to analyze their data and make predictions. Using VDA, large amounts of data can be manipulated and presented as visual images for analysis by the most powerful interpretation machine in the world -- the human brain.

Scientific staff use powerful HP^® Series 700 workstations to analyze their data. According to Alastair Sangster, the workstations must cope with some demanding tasks. "Our network configuration is such that each workstation also supports a single X-terminal, i.e., normally two users share a single CPU. This can put a heavy load on the host machine, particularly when both users are using graphics packages."

Within the Hadley Centre are around 100 scientific and support staff. They are divided into groups, with each group looking at different aspects of climate and climate modeling, working with different sets of data.

Mark Webb and Andy Jones, for example, are cloud specialists. It is well known that clouds can have both a warming effect and a cooling effect. They act as a blanket, keeping in warmth, but also as a barrier, stopping sun from reaching the earth's surface. To be able to provide an accurate prediction of average global temperature, the action of the clouds must be understood and built into the model. Webb and Jones use PV-WAVE software from Visual Numerics to visualize the effects. Model data can be mapped onto satellite observations, giving insight into the relationships between clouds and radiation.

"It may sound straightforward," said Webb, "but there are many, many variables. Clouds can be bright or dark, and both long- and short-wave radiation must be considered. Before we had PV-WAVE, it was difficult to pick out relevant details from the data. Flexible data manipulation allows you to get the 'big picture,' then to zoom in on the interesting features. The interactive environment is also a great improvement, allowing the development of new analysis techniques in a fraction of the time previously required. With PV-WAVE it's not a case of can I do it, but simply how do I do it," concluded Webb.

Another area of research is the interaction between ocean and atmosphere. Simon Tett is looking at the reasons why in some years the temperature of the sea surface in the Eastern Pacific is significantly warmer than average and in other years the temperature is cooler than normal. By using PV-WAVE, Tett and others aim to build more accurate models of the oceans and understand the effects of the oceans on the climate of the world. "El Niño is a good example of this effect," commented Tett. "It is an oscillation in the temperature of the East Pacific. Depending on its status, Australia will have a wet or dry year."

Jonathan Gregory is involved in prediction of sea levels. It is a complicated problem, as several effects have to be taken into account. As the world warms, the expansion of the sea will make sea level rise. Glaciers will eventually melt, and the permanent ice sheets at the poles may eventually melt too. However, as the atmosphere warms, greater amounts of moisture will be held in the air. It is also possible that in the short term, snowfall may increase in Antarctica, producing more ice there.

The results of the Hadley Centre's work are published via technical papers and journals. Included in the work are constantly updated predictions of climate change to the end of the 21st century.