Frustration was in the forecast at the Phoenix office of the National Weather Service. Meteorologists knew remnants of Hurricane Dean might be heading toward Arizona, which meant lives were at stake. The state’s history is dotted with accounts of tropical systems bringing torrential rains and killer flash floods to the desert.
But the weathermen looked in vain for data from the south that could provide clues to Dean’s future path.
Due to budget cutbacks in the Mexican weather service, weather balloons are released over that nation with far less frequency than in years past. That means their American counterparts’ predictions are grounded in guesses of such crucial factors of the wind’s direction and speed, temperature, humidity and air pressure.
As a result, experts say, the accuracy of weather forecasts in Arizona is suffering badly.
The effects could be felt this week, as Tropical Storm Henriette, currently off the Mexican coast in the eastern Pacific Ocean, threatens to bring rain in the Southwest.
Weather stations across the world release balloons shortly before midnight and noon Greenwich Mean Time (5 a.m. and 5 p.m. Mountain Standard Time), providing about 1,500 global observations twice a day.
In Arizona, the Flagstaff and Tucson offices of the weather service do this year-round. But in Phoenix the balloons are released only during the monsoon.
On the morning of Aug. 24, the Phoenix office’s forecast mentioned the difficulties in trying to track Dean. The last measure of the air over Mexico came two days before in Chihuahua, 500 miles to the southeast.
“But big deal,” read the meteorologist’s note, its annoyed tone standing out among the dry discussion of subtropical shortwaves and wind fields.
Since the Aug. 22 measurement, matters had changed so dramatically that the weatherman admitted he was flying blind. “We are going to have to wing it,” he wrote.
The guess that Dean’s moisture would surge into the state was wrong. Instead, it took a more westerly path toward California and drenched the mountains north of San Diego.
That mistake left Art Douglas high and dry. Douglas, former chairman of Creighton University’s atmospheric sciences department, had traveled to southern Arizona in anticipation of Dean’s arrival.
“The computer models, if they have no data from Mexico, what (the forecasters) rely on is not very good and you can be so terribly off,” Douglas said. “And this was a case of being so terribly off.”
Douglas currently works as a consultant to the Mexican weather service, the Servicio Meteorológico Nacional, and he blames a mix of factors for the balloon problem.
First, the agency recently updated the instrument pack attached to the balloons, only to see the cost of that technology nearly triple in price. Those packs, called radiosondes, can cost as much as $200. Multiply that price by two launches a day at more than a dozen sites, and the money can add up.
Then, the budget shortfall got lost in the government shuffle created by a change in presidential administrations. There may be funding in 2008, but for now there’s little solution other than infrequent launches.
“Don’t think the Mexicans aren’t concerned,” Douglas said. “They really consider this to be a travesty.”
Until the Mexican government buys more balloons, forecasters on this side of the border are developing workarounds.
John Skindlov, a meteorologist at Salt River Project, said Global Positioning System technology has been employed in an unorthodox manner. The signal between GPS satellites and receivers is affected by airborne moisture, and forecasters can study the signal to determine relative humidity.
“It’s better than nothing,” Skindlov said.