Cloud Seeders Claim Documented
Results In Mexico Experiments

BOULDER, Colo. — Scientists at the National Center for Atmospheric Research here believe they have succeeded in increasing rainfall in existing storm clouds and quantifying the results.

A three-year randomized experiment in the northern Mexican state of Coahuila indicates that rainfall from seeded clouds last longer, covers a larger area, and generates more total precipitation. In some instances, total precipitation doubled over output from similar nonseeded clouds.

In many cases the results of the seeding were statistically significant 20 minutes to an hour after seeding.

The Mexico project was designed to repeat the success of a five-year effort conducted in South Africa in the early 1990s. The new study, which followed several years of drouth in northern Mexico, was funded by the Mexican state of Coahuila with financial support from Altos Hornos de Mexico, a private steel company.

NCAR's primary sponsor is the National Science Foundation.

Researchers flew into the Mexican rain clouds in a Piper Cheyenne twin-engine turboprop airplane, equipped with wing-mounted racks carrying 24 hygroscopic flares and an instrument package to measure basic cloud physics indicators. The flares spewed salted smoke into the moisture-rich updrafts entering the clouds from below. The tiny particles, a mixture of sodium, magnesium and calcium chlorides, attract and absorb surrounding water vapor to create large drops heavy enough to fall as rain.

"We are very encouraged by the results," says lead scientist Roelof Bruintjes of NCAR.

However, he says, the number of cases is marginal for any statistical analysis.

Funding for a planned fourth year of data gathering was cut when the Mexican drouth ended.

This left the total number of cases at 94, compared to 127 in the South Africa study. Researchers say they think more seasons in the field would extend the results and help establish statistical significance.

Scientists say they cannot fully explain how the seeding process works inside the cloud.

"We must be able to explain both microphysical and dynamical responses of the cloud to the seeding procedure," says Bruintjes, "before we can claim full success."

The main tool for "nowcasting" the weather and for evaluating the seeding experiment is a five-centimeter wavelength weather radar. Special software developed at NCAR displayed the radar data and the aircraft's position in real time for directing operations. It also identified storms for evaluating the results.

NCAR is managed by the University Corporation for Atmospheric Research, a consortium of more than 65 universities offering Ph.D.s in atmospheric and related sciences.