The Weather Bug

Researcher is itching to develop a more reliable long-term forecasting system

If a butterfly flaps its wings in Iowa, will it storm in Colorado?


But Bill Cotton can tell you that the one has little to do with the other.

Cotton, an atmospheric science researcher and professor at Colorado State University, dedicates his time to more reliable methods of predicting rain or shine in Colorado.

Using the combined computing power of 25 Pentium computers strung together, Cotton is developing a comprehensive, detailed forecasting system that will predict Colorado weather as much as six months in advance. Granted, six-month forecasts won't be pin-point accurate, but they will give agricultural producers and reservoir managers a good general idea of how hot and how wet a summer may be.

That's important. Agriculturists rely on the weather to nourish grassland for grazing livestock or to help irrigate crops. They not only rely on annual rainfall in the spring and summer, they also rely on reservoirs to hold winter snow runoff to be used as irrigation water in dry, hot summer months. For those who gamble with Mother Nature when managing water resources, Cotton's forecasts can mean the difference between a productive year and a long, hot, dry one.

"These forecasts will be long term enough to give such people as reservoir managers and producers enough time to gauge how much water to preserve or how much water they'll have to irrigate crops," says Cotton.

Establishing a forecast is critical to reservoir managers. The winter of 1983 was a good example, says Cotton. There was little snow pack, leading producers and reservoir managers to expect poor run-off. Then, much to their surprise, late March brought heavy snow and rain. Unfortunately, the reservoirs were kept at a high level in anticipation of a dry year and, as the late moisture came in, flooding and erosion caused millions of dollars of damage in many areas.

Cotton's forecasting system is part of the Regional Atmospheric Modeling System (RAMS) developed by the Colorado State University Agricultural Experiment Station. The system allows Cotton to make more comprehensive 24- to 48-hour forecasts than those made by the National Weather Service because it predicts weather variables in a tighter geographical region, showing more detail by pin-pointing conditions every seven miles or so. The National Weather Service system only looks at regional conditions and makes generalizations.

The RAMS' tighter grid system allows Cotton to capture conditions from small storms in specific areas and forecast wind, temperature, evaporation, precipitation, humidity, and other variables that often slip through the National Weather Service's larger grid. These specific variables can help producers gauge how to create an optimum atmosphere for their crops to grow. For example, with this information a producer can find the balance between how much of his irrigation water is evaporating into the atmosphere and how much is nourishing his crop.

Using those variables, historical data, and such complicated factors as sea surface temperatures, snow cover, and weather patterns, Cotton can also make general forecasts several months ahead, especially when gauging snow fall, snow pack, and snow runoff for reservoir storage.

"We try to establish soil moisture, vegetation, and many other details in order to make long-term forecasts," said Cotton. The short-term, high-resolution forecasts can even look at the effects a city has on the weather, with lawns being watered and the numerous other variables created by humans – or the flap of a butterfly's wings.