Weather apps off by 100% spur grower warning at Big Iron forum

A rainfall app showed 2.5 inches of rain at one North Dakota location when the true figure was 1 inch — an error above 100% that Satshot CEO Lanny Faleide cited as evidence that cellphone weather data remains too unreliable to drive major farm decisions.

Faleide, president and CEO of Satshot, a company using satellite and aerial imagery for production management, made the point during an Agweek-sponsored forum last month in West Fargo, held during the Big Iron Farm Show on the Agweek 360 building stage. He was joined by Nolan Berg, director of agronomy at Heads Up Plant Protectant; Tommy Grafenauer, science and operations officer at the National Weather Service's Grand Forks office; and Lydia Blume, a meteorologist with WDAY television.

"I am a tech nerd for my age. I have all of these weather apps. But what really concerns me is that interpretation of the data that is coming out," Faleide said. "We are interpolating a lot of data, but I am concerned about the accuracy of the data that is coming through and that is prompting decisions that maybe are on the edge of danger."

The panel's consensus: AI and phone apps cannot yet fully inform big on-farm decisions, especially on weather.

What did the panelists say about technology limits?

Berg's company specializes in biological seed treatments that defend soybeans against white mold and sudden death syndrome, and its products perform best when cold, wet conditions threaten. He said growers now track weather daily on their phones and urged them to weigh long-term averages rather than short-term trends.

"I would suggest you don't look at last year and next year. Look at the average — where it is trending," Berg said.

He also flagged the data gap that no software can close. Recalling his earlier work in variable-rate precision agriculture, he described field anomalies that systems misread because farmers had spread extra manure or shut off spray in a given spot. "The systems don't know that stuff, so they can't be perfect," he said. Much of what happens on a farm goes untracked, limiting AI's usefulness for major decisions.

How should growers measure rainfall?

Blume said radar often overestimates rainfall, and she trusts physical gauges over modeled estimates.

"I will trust even an automatic rain gauge that tells them (rainfall amounts) on their app, or just a good old plain rain gauge," she said. "When it's set up correctly, it's clearly the most accurate way to measure rainfall."

She added that traditional, trusted methods will remain useful wherever meteorology and agriculture intersect.

Will AI replace meteorologists and agronomists?

Grafenauer, a 25-year National Weather Service veteran, said fears of AI displacing jobs in meteorology and agriculture are overstated. "In my opinion … there is always going to be a human component. I truly believe technology can only take us so far," he said.

Faleide credited seed genetics and software that assigns fertilizer and chemical types and rates to precise locations and timings with revolutionizing production agriculture. But he called reliable weather information vital, pointing to National Weather Service and WDAY data as anchors.

The forum also covered regional drought conditions and current weather patterns. The panel's bottom line for producers: pair new digital tools with experienced judgment and verified weather sources before committing input dollars.