Irving Krick's 'Weathercasts'

Weather: That was then, this is now

Compared to forecasting methods used in the days of Irving Krick’s Weathercasts, today’s weather forecasts, observations, and means of dissemination are infinitely more accurate, and quick as lightning. The biggest difference between forecasting then and now is computer power. Krick relied on a single artificial construct of the atmosphere—or model—generated with limited information and updated at what must have been an extremely slow rate. Now there are dozens of much more highly advanced computer models, constantly running on high-speed supercomputers, with most of them updated twice daily. Some, such as the RUC (rapid update cycle) are updated every three hours, and thus give rolling forecast updates for the following 12 hours. Other models forecast out to five days or more. Still others are tailor-made for studying hurricanes and tropical storms.

Modern models use complex algorithms to come up with forecast conditions. Most of the information plugged into these algorithms wasn’t available in Krick’s time (with the exception of weather balloons, which are still time-tested reporters of temperatures and pressures aloft). Among the information not available to Krick is multi-channel satellite imagery, which lets us see storms form well in advance, follow their motion, and determine their characteristics. A vast network of weather buoys also transmits vital information on pressure, winds, waves, and temperatures. Automated weather observations from ASOS and AWOS installations help us keep track of weather in near real time. Ditto airliner broadcasts of winds and temperatures aloft. All of this information—and more—is fed into supercomputers, which then crunch the numbers and come up with graphic representations of anticipated atmospheric conditions. Most of the models depict four aspects of the atmosphere.

One is vorticity and pressure levels in the upper levels, at 18,000 or 30,000 feet msl. This tells forecasters where atmospheric “spin” is located. Counterclockwise rotational movement can give rise to surface low pressure, and so predict fronts and severe storms. Another depiction is the atmosphere’s thickness between the surface and 18,000 feet; this helps locate where fronts, lows and highs will be, and the nature of any precipitation they may contain. Another representation shows moisture concentrations in the mid-levels of the atmosphere. And the final image shows anticipated rainfall amounts and temperatures. But while all this information is computer-derived, the job of interpreting still falls to a meteorologist sitting at his display console.

And let’s not forget the revolutionary impact the Internet has had on weather dissemination. Anyone with a personal computer—something that no one envisioned in the late 1950s—can access huge amounts of forecast information. This ranges from the raw information from the models themselves, to the very specific observations and forecasts provided by such means as AOPA Online, DUATS providers, and several private weather services.

It doesn’t end there. In the ultimate manifestation of real-time weather, nothing beats today’s datalink weather and on-board weather sensors. Airborne weather radar was in its infancy in Krick’s day, but is now standard on larger aircraft, continually painting the location of precipitation—and even turbulence. Lightning detectors, available since the 1970s, also provide real-time weather information in the cockpit. Most recently pilots have clamored in droves to datalink weather, which provides in-cockpit, near-real time views of high-resolution Nexrad radar imagery, fronts, satellite imagery, surface observations, TAFs, winds aloft, airmets, sigmets, TFRs, and lots more, giving pilots of even the smallest piston singles the ability to monitor a huge amount of weather information. It’s equipment that many instrument pilots consider essential for instrument flying, and that most manufacturers provide as standard equipment in new airplanes. Datalink has revolutionized the way we obtain in-flight weather and make in-flight weather decisions, and has almost certainly helped reduce weather-related accidents. Apart from a few science fiction writers, this too was way, way outside the scope of conventional thinking in the Weathercast days. Now we have a way of identifying and avoiding the kinds of small-scale, intense weather phenomena that Krick could never hope to predict.

He painted in broad strokes with limited data, using vague techniques to make vague assumptions far into the future. With models, online, onboard, and datalink weather, we use scads of data to pinpoint adverse weather in great detail, and in time frames relevant to our flights. That’s weather information of true value, because it’s the small-scale weather events that pose the most danger. Sweeping forecasts like Krick’s are nice to know, but have little value once your wheels leave the ground.— TAH

For the first 21 years of The AOPA Pilot, the magazine carried a monthly column that published long-range forecasts of VFR and IFR weather. The columns were the product of Irving P. Krick Associates of Denver. Krick, a professor of meteorology at the California Institute of Technology in the 1930s, was a controversial figure in the meteorological community, as we’ll soon see. But Max Karant, The AOPA Pilot’s editor in its early days, was convinced of Krick’s accuracy.

The AOPA Pilot contracted with Krick to provide his “AOPA Weathercasts” in an attempt to give members reliable weather outlooks. These one-month projections were quite audacious—and extremely specific—in their proclamations. From 1958 to 1972, they predicted the hours per week that various areas of the nation could expect instrument meteorological conditions, and the dates when certain popular routes would experience either VFR or IFR weather—all a month in advance! Actually, the predictions would have to have been made some six to seven weeks in advance of the magazine’s publishing date, taking into account the amount of time needed to prepare the magazine for publication.

How could such exact predictions have been made in those days? How could even the most senior meteorologists of that time point to a map and say, “Six weeks from now, from April 9 through 14, there will be 55 hours of instrument weather in central Kansas?” It makes you wonder.

Today, entire rooms filled with state-of-the-art Cray supercomputers run 24/7 to make forecasts. They crunch data from a wide range of observations using lofty equations dealing with atmospheric temperatures, pressures, and motions, to name but a few variables. From these and other rapidly updated inputs, the supercomputers generate models—their ideas of what the atmosphere will do in the future. Most of the observations used to make today’s models—such as satellite imagery, remote sensing of ocean waves and temperatures, cloud-top temperature readings, Doppler weather radar, and ASOS/AWOS and, of course, the Cray’s themselves—didn’t even exist when Krick made his “Weathercasts.” (The first weather satellite, TIROS I, was launch-ed in April 1960, and could only transmit black-and-white imagery in the infrared spectrum).

But even with Cray-powered modelling, modern forecast accuracy of mesoscale meteorology (weather covering five to 500 nm, and lasting one to three days) has its limitations. Five-day forecast accuracy is fairly good these days, three-day forecasts are often quite accurate, and accuracy is improving all the time. But a one-month forecast of instrument weather in a small portion of the United States? We’re still not there.

And yet, Krick said he could make these kind of forecasts with 80 percent accuracy in 1958. How?

The story goes back to World War II, when Krick was involved in forecasting for the U.S. Army Air Forces. He claimed he assembled all the weather maps and data for the northern hemisphere from 1899 on, then observed three basic forecast elements. One was that there were 11 types of sequential weather patterns. Another was that the positions of several semi-permanent high- pressure systems—located at about 30 degrees north latitude—heavily influenced these weather patterns. And the other was that these systems “moved with the sun,” meaning they travelled north in the summer, and south in the winter. For the North American continent, Krick said that the position of the North Pacific High was the key controller of weather.

These are not shocking revelations. We’ve known about the behavior of the North Pacific High, and its influence on adverse weather entering the Pacific Northwest (and from there, on to the east) since the 1800s. So the real question becomes: How did Krick translate his theoretical concepts into such concrete, long-range forecasts?

The answer is: We don’t know. Some believe he simply used his primitive computer—a Remington Rand Univac 120—to try to duplicate the weather-making, large planetary-scale, high-altitude pressure waves that routinely circle the mid-latitudes. Others think he relied on the influence of the moon or planets. No one really knows for sure. But he seems to have somehow banked heavily on repetitive atmospheric patterns.

The professional community shunned Krick’s work, and some even called him a charlatan. True, he did become involved in the first successful cloud-seeding experiments, and yes, he rubbed shoulders with many great names in meteorology, but so much else is dubious about his work that the American Meteorological Society (AMS) censured him in the 1950s, citing the improbability of accurate forecasting for months ahead.

There were other claims that aroused suspicion. Krick claimed to have made the critical forecast that dictated the timing of the D-Day invasion of Europe in June 1944. Wrong, according to Group Captain James M. Stagg of the British meteorological team that headed up the forecasting effort. Krick, he said, was just one American on the Allied forecasting team.

Krick also claimed to have made a 1969 forecast of the drought that hit the American Midwest in the mid-1970s, plus numerous accurate forecasts for various Olympic games, presidential inaugurations, and other notable events. But, because of the sketchiness and obscurity of his methods, and his penchant for blatant self-promotion, both the AMS and the U.S. Weather Bureau (the forerunner of the National Weather Service) looked down on Krick. Most of his contemporaries considered him a maverick or a snake-oil salesman.

Even so, Krick’s “Weathercasts” ran until 1972, when they were changed to show frontal systems, lows, highs, and areas of precipitation. These ran until December 1979, when his contract ended and AOPA Pilot ceased the forecasts’ publication.

Karant, in spite of what he may or may not have known about Krick’s background, was a staunch believer. Although half of AOPA Pilot readers wrote in to say they found Krick’s forecasts useless, Karant was quick to point out that the other half felt them uncannily accurate. This division of opinion was also evident among AOPA staffers. It seemed like pilots either loved the forecasts or hated them.

Krick died in 1996, taking his forecasting secrets with him. We’ll never know what tricks he and his Univac knew about the atmosphere, in spite of what must be the most definitive book about the man, Storm: Irving Krick vs. the U.S. Weather Bureaucracy, by Victor Boesen. Want to read it? It’s online. Chapter 10 includes passages about Krick’s involvement with AOPA.

Quack or genius? Either way, Krick’s forecasts were a colorful and amusing chapter in AOPA Pilot’s rich history.

E-mail the author at [email protected].