“NASA Goddard Debuts New Climate Simulation Center “ video file interview transcript:

Mike Bosilovich:

Clip 1

A model starts from a very large computer program, but it is all gauged by what we know about the Earth and how the Earth’s atmosphere works. So that we have temperatures we have precipitation we have the sun radiating on to the earth. We have the earth radiated back to space. We take all this information and the model produces the fields that we can predict in to the future and get model predictions.

Clip 2

We are looking at the individual weather events and we’re building a long time series of those weather events and collecting them over a long period of time. So the idea is then that we can build a record of what the climate looks like based on weather events. In fact data assimilation is really mostly about the weather. But we can start doing reanalysis and start understanding climate better the longer time period we use.

Phil Webster

The exciting thing about these particular machines is that they are based on the halem processors by Intel. And those processors are twice as fast as the previous generation machines. So we can get something twice as fast and we can provide the scientists computing power to double their capability to support the really important stuff that’s being done with weather and climate research.