Stellar Startup: Doing something about the weather

Hi-tech isn't just about esoteric databases and developing trailblazer technologies. At the bottom line, it's really all about the bottom line: Making money.

By DAVID SHAMAH
weather image 88 298 (photo credit: Courtesy Photo)
weather image 88 298
(photo credit: Courtesy Photo)
Hi-tech isn't just about esoteric databases and developing trailblazer technologies. At the bottom line, it's really all about the bottom line: Making money. Case in point: The TV weather person. While we all turn on the TV news once a day or so because it's what informed people do, we all know in advance what to expect: War, tragedy, greed, scandal, etc. Same old, same old. We could be watching last year's edition, for the depth of variety and innovation. That's why they put the "expanded forecast" with the maps and the "personality" on last; the TV executives know why you're watching, and they want you to keep it on for as long as the broadcast lasts in order to boost the ratings. The weather, unlike the "hard" news items, contains an element of surprise, and appeals to our sense of trying to control the future; the weather person is like a fortune teller, predicting what will happen in the next three to five days. The better the prediction, the more "control" the prophet has over upcoming events. Or something like that. Bottom line: The more accurate the forecast, the more people will be drawn to watch the news, and the more bragging rights the station/forecaster will have, which can be used to attract even more viewers, thereby generating higher ratings and more cash per advertising minute on the broadcast. So: Far from being a stodgy science with a funny name - "meteorology" - weather forecasting is about dollars, cents and shekels, and it shouldn't at all be surprising that hi-tech companies would throw their computerized heart and soul into developing ways to use programming and processing to better predict the weather. Successful weather prediction is a function of two things, says Dr. Barry Lynn, of Weather-It-Is (http://www.weather-it-is-israel.com) - processing power, and an accurate forecasting model. Dr. Lynn (he's a doctor of meteorology, in fact) has got a forecasting model that he prefers to work with - the Weather Research and Forecasting Model (http://www.wrf-model.org) - the processing power to break down weather forecasting for areas as small as 10 kilometer square (i.e., specific forecasts for small areas), and his own programmed algorithms for predicting accurately - more accurately than most - exactly what the weather is going "to do," as most of us put it. And it's not just about providing TV newscasts with higher ratings; improved and more accurate weather prediction could be a boon for many industries that need to wrestle the environmental elements in order to get work done. Take an electric company crew that needs to do major line work, for example. These guys get paid a huge hunk of change for field work, and if the company sends them out on a job, while they sit in the truck instead of working because a surprise electrical storm has made it too dangerous to work, the company - and, of course, its customers - end up footing that bill. An accurate weather prediction for the specific area in question is valuable information for the utility, says Dr. Lynn, and they'll pay - as will oil and gas drillers, farmers, airports and a host of other industries and services. But surely major weather prediction services have already seen this market and taken advantage of advanced technology to come up with more accurate models. Not necessarily, says Dr. Lynn, and certainly not everywhere. Among the problems is that many of the weather services (at least outside the US) are tied to government or educational institutions, and we all know how they can be about changing their methods and/or approaches to solving problems. Much of the prediction algorithms in use are built on less accurate (i.e. more primitive) models, and rewriting them would of course entail a lot of time and money. And the better commercial ventures that do exist serve only the "top" markets in the US and Europe - and charge a hefty fee for their services. Wide swaths of the world, including much of Asia and the Middle East, and all of Africa, are not covered by any advanced service. And often, it's these developing countries, especially in Africa, that could truly benefit from enhance weather prediction. Closer to home, some experts charge that the Israel Meteorological Service bases its forecasts on an inaccurate model and last-generation software. "It's not really their fault," Dr. Lynn says, "because their budget is so small, they can't afford new equipment." Nevertheless, he said, the IMS charges a lot of money for their so-so service - while his predictions have more often than not been much more accurate. Case in point: The threatened snowstorm that eventually did hit Jerusalem in January. Having a vested interest in knowing what to expect, since I was scheduled to drive into and out of the city on storm day, I followed the detailed forecast on Israel Radio closely. As Dr. Lynn reminded me, and as indeed I noticed myself, the weather service changed its forecast at least three times in the days preceding the storm, first predicting a sizable (for Jerusalem) layer of snow on the ground, then backing off altogether, saying the storm would bring no more than freezing rain, and then, in the immediate hours before the storm, a prediction that snow would indeed fall and that there would be a modest accumulation - the prediction, Dr. Lynn said, that he had already made on his Web site days before the storm hit. So, there's plenty of opportunity close to home - as well as in neighboring countries, including Egypt, Lebanon and Jordan, which have almost no weather forecasting services to speak of. Dr. Lynn's site has city by city and region by region breakdowns for all these countries, as well as, of course, for Israel, and the forecast for each area analyzes in depth the available information and produces a unique prediction for what residents of each specific city can expect - unlike the IMS, which produces forecasts by wide regions. But even though they are both in the North, Safed and Haifa have radically different weather - even Haifa and Carmiel can often expect different conditions. Dr. Lynn's site takes into account these differences and produces a what he says is a more accurate forecast for these and dozens of other localities in Israel and the rest of the Middle East. The fledgling company is expanding, and hopes to be able to implement 10 kilometer area predictions in the very near future (after they acquire some more processing power). Dr. Lynn said that his company had worked out a deal to supply information to a forecasting company called Nooly (http://www.nooly.com), and that Channel 10's weather "personality," Danny Rupp, will use the company's data in upcoming forecasts. Meanwhile, you can check out what Dr. Lynn predicts at his site, http://www.weather-it-is-israel.com, for your town - and maybe soon, for your street! startup@newzgeek.com