Modeling 'Science on a Sphere' to save the environment

Creating a 10-day Earth prediction system using NOAA's Science on a Sphere would produce more than 52 exabytes of data, but could help scientists address our greatest environmental challenges.
Written by Mari Silbey, Contributor on

Dr. Alexander MacDonald believes scientists hold a great responsibility when it comes to educating the public about the environment. Research is critical, but "mass demand for action will only flow from mass understanding of the predicament."

At the recent Internet2 member conference, Dr. MacDonald, Chief Science Advisor for the National Oceanic and Atmospheric Administration (NOAA) and Director of the Earth System Research Laboratory, delivered a keynote speech focusing on his invention of and ongoing work with Science on a Sphere. Part of an environmental literacy effort by NOAA, Science on a Sphere consists of installations around the world that display dynamic animated images of the Earth on an interactive globe. MacDonald preaches the theory that if we want to understand our world, "let's look at it in the shape it's actually in."

The simple presentation of Science on a Sphere belies its underlying complexity. First conceived in the 1990s, the installations encompass a tremendous amount of data across three fundamental layers: physical, chemical and biological. From clouds and currents to sulfur dioxide levels to plant and animal growth, Science on a Sphere enables detailed study of the numerous elements that impact -- and are impacted by -- the environment.

The amount of data involved in Science on a Sphere is almost inconceivable. Measurements must be recorded in multiple dimensions, taking into account both time and space. To chart the course of a storm, for example, meteorologists might track the formation of different types of clouds minute by minute. That's just one element necessary for predicting a storm's path. Imagine collecting data for every variable impacting storm development, and then multiplying that out over time.

MacDonald has run the calculations for creating a 10-day Earth prediction system. Taking into account 100 data fields -- with categories like temperature, carbon dioxide levels, etc. -- and measurements at one-minute intervals, such a model would produce more than 52 exabytes of data. To put that in context, it would require well over a dozen years to transport that much data over a one-terabit-per-second Internet connection.

Is that level of data-intensive study worthwhile? MacDonald points out that if we had understood the impact of chlorofluorocarbons earlier, we could have taken action to protect the Earth's ozone layer. If we can better understand trends in ocean pH levels today, we might be able to determine if there's a way to save underwater creatures at risk of losing their calcium shells to increasing ocean acidity.

MacDonald sees it as crucial that scientists not only study the environment in detail, but communicate about this research in a way that leads to problem solving. "A key to helping people is that we have a giant bandwidth receiver in our head," says MacDonald, "everyone does, but how do you get to that in the right way?"

Science on a Sphere is one answer. Thanks to 97 permanent installations and numerous traveling exhibitions, 25 million people were exposed to the project last year. That type of visibility and shared insight about the environment may help stimulate a widespread preservation response.

MacDonald hopes so. To him, the stakes are no small matter. He believes we're at a time in history where technology has given us the power "to create either a heaven or a hell" on Earth.

(Images courtesy of NOAA)

This post was originally published on Smartplanet.com

Editorial standards