Accurately predicting when, where and how much rain seems like magic. If you’re foretelling the future, and you’re accurate, then it must verge on magic right? Or is it science?
Ask Scott Kehler and he’ll tell you it’s all a matter of science, but it’s got to be good science. Kehler is the chief researcher and co-founder of Weatherlogics, a two-year-old Winnipeg company that specializes in predicting weather for Manitoba farmers.
The lead on their website reflects their passion.
“We can’t help it. There’s nothing we’d rather be doing than studying, predicting, and reporting on weather.”
Kehler says they have a meteorologist on staff who provides clients with a daily forecast for their specific farm every morning, beginning at seeding time. For agricultural clients, they focus solely on Manitoba. That’s because these forecasts are prepared by a person, not spit out by a computer. This human factor keeps the meteorologist and researchers in close contact with their farmer clients. For other industries that don’t require such a close relationship, they provide services coast to coast and into the United States.
“The forecasts you see on-line from big weather agencies and phone apps are computer generated. They can turn out a lot of forecasts very quickly, but they struggle to figure out complicated high-impact events like localized thunderstorms,” says Kehler, adding that Weatherlogics has scientists analyzing weather patterns affecting Manitoba clients.
“Every weather agency has access to the same data, because that data all comes from governments around the world and it’s shared with researchers and meteorologists. Data from satellites and radar and weather stations is available to all of us. Accuracy depends on how you use that data to make a forecast. In our case, real people crunch the numbers.”
Kehler, who was born and raised near Steinbach, Man., says his scientific background is researching nighttime thunderstorms on the Great Plains states, especially Kansas. His team determined that prairie thunderstorms are actually big contributors to rainfall during the growing season, but they are difficult to predict.
“These storms will pop up in a random fashion. We’ll typically track the movement of these possible storms throughout the day as they begin to develop. And we’ll send our subscribers updates of where the storm is tracking and what you can expect from it, heavy rain, hail, tornado or whatever.”
Kehler concedes there’s not much actionable information in the daily morning reports. You may hurry to combine some swaths in one particular field, cancel a spray operation on another or hurry to top-dress nitrogen. He says the five-day forecast has a lot more value.
As farms become larger and more widespread, weather at one end of the farm might be quite different from weather at the other end. The types of crop, growth stages and locations of the crops all play into a farmer’s management plan, as does a farmer’s use of the five-day forecast. An accurate forecast allows farmers to better allocate equipment and manpower. For example, do you send the straight-cut combines to the north end or put on the pickup reels and go to fields in the south end?
“If the five-day forecast during harvest says rain in two days, a subscriber will use that information to decide which fields to concentrate on and which fields can wait. During spraying, the long-term forecast helps in the same way. Guys use the report to plan what work can get finished before rain arrives. Good weather forecasts let you do a better job of setting priorities.”
Weatherlogics currently serves about 100 Manitoba farms. Kehler says his company can take on more Manitoba farms, but there are no plans to expand the ag business outside the province. The intimate nature of providing accurate localized forecasts demands a close watch on specific weather patterns. He feels that expansion to a larger geographic area would dilute the quality of service.
“There are a number of companies providing agricultural weather forecast aimed at specific areas, but most of them are computer generated. There are only a very small number of companies like ours who have genuine meteorologists analyzing data.
“A lot of companies install a weather station on your farm. But the thing about a weather station is it doesn’t predict weather. It only tells you what already happened. It might have a computer that gives you some kind of forecast, but that forecast is generated by a computer on the east coast spitting out millions of forecasts a day. It’s not a science-based forecast.
“That’s the key to what we’re doing. We live in Manitoba, so we know the weather extremely well and we have a direct link to our clients. We produce the reports ourselves and we do our own research. My role in the company is researching new and better ways to predict weather.”
There is one level of service. Every client receives the same close attention. In 2018, the cost was $250 per farm. The rate hasn’t been set yet for 2019.