March 2, 2005

Snowjobs About Weather and Climate

Filed under: Climate Models

Want to raise the blood pressure of an entire region? If you’re within a few hundred miles of Washington DC, just say “snow” to a TV camera. But if you’re more interested in planetary hypertension, simply substitute the words “global warming”.

It turns out that the forecasting methods for both snow and global warming are quite similar. And what happened with the Mid-Atlantic snowstorm of February 28 tells us much about the climate of the next 100 years.

Both snowstorms and global climate change are predicted by computer models. For snow, the computer starts out with a global “snapshot” of the atmosphere, produced simultaneously, around the world, by weather balloons at 7am and 7pm Eastern Standard Time. They sample temperature, humidity, altitude and wind. Once these variables are measured, the computer goes to work, shoving moisture and heat around the planet, creating what we call “weather.” What changes from day-to-day in these models is the “snapshot” that puts the whole shebang in motion.

Climate change models have similar physics. But because they are concerned with 100-year (rather than 100-hour) timeframes, they all start with the same initial conditions and then slowly modulate them with the warming effects of the main human “greenhouse” emission, carbon dioxide, along with the cooling effects of the particles that go in the air at the same time we burn fossil fuel. Together, these produce changing patterns of surface temperature and rainfall as the years progress.

For both weather and climate, there are a number of different computer models. That’s because we have only a partial knowledge of the way the atmosphere truly works, requiring different models to make different assumptions. For example, we know that some clouds cool the surface, while others warm, but we really don’t know how much. So different models use different values, which may change their rate of projected snowfall (in the case of weather) or annual temperature (when dealing with climate).

The February 28 snowstorm was one of those many occasions in which different models predict different outcomes. One of the two main federal models, the Global Forecasting System (GFS) projected modest amounts of moisture, while the other, the North American Model (NAM) had nearly one and three-quarter inches of liquid coming down not far from Washington DC. If that fell as snow, DC could have experienced its biggest storm since the paralyzing whopper of January 1996, which set the all-time record for a single snowstorm at Dulles Airport.

What do we teach forecasters to do in such a situation, where two equally sophisticated computer models say such different things? “Look out the window, stupid.”

Literally. When models differ, forecasters look to see which one has been tracking things more correctly, or which one was right the last time there was a threat of a similar storm.

In this case, NAM was consistently placing the center of the storm too far to the west. So it was thrown out in favor of GFS, and the forecast for much of Northern Virginia and Maryland was for 4-8 inches of snow—enough to cause the usual panic, but hardly a paralyzing blow.

Indeed, that’s largely what fell, with the exception of urban Washington, where the city’s heat melted some of the snow when it hit the surface. The District averaged 2-3 inches rather than hitting the minimum of 4. Frankly, given all the uncertainties in snow forecasting, the National Weather Service forecast indeed was good enough for government work.

For the 100-year global warming window we have dozens of climate models to choose from. So which one (if any) is right? Hey, look out the window again!

These models tend to share a common characteristic: once warming starts from human activities, it tends to take place at a constant rate. So, all one needs to do is demonstrate that warming is taking place, see if it is indeed at a uniform clip, and you know which model to select.

Greenhouse effect theory predicts that warming takes place inordinately in cold, dry air in the winter and much less in the humid summer. Over the last several decades, indeed, the largest warmings in our hemisphere are in Siberian, Canadian, and Alaskan winters, where the air is coldest and driest. Okay. Greenhouse warming.

Now, is it indeed a constant rate? No doubt about it. The trend line couldn’t be straighter. So what is the rate? About 0.8ºC (1.4ºF) per half-century, a rather modest amount.

So there’s your long, long term climate forecast. All we have to do is adjust the models for reality, something that happens most every day with the short-term weather forecast.

And, before getting into a snowstorm panic about this, remember that there was the same amount of warming spread over the entire 20th century. Life span doubled, some crop yields quintupled, and the average U.S. citizen became wealthy beyond the wildest dreams of the historians of a century ago. Rather than reacting hysterically to it, we survived and prospered. So why should we get upset about a similar warming in the next fifty years?

No Comments

No comments yet.

RSS feed for comments on this post. TrackBack URI

Sorry, the comment form is closed at this time.

Powered by WordPress