Jörg Kachelmann is Germany’s most recognized weatherman. His hobbies are simple: weather, sailing, and more weather. Day after day, he and his team at meteomedia present forecasts for every region of the republic. They broadcast during the prime slot before the Tagesschau news or in half-hour intervals during the morning show. To achieve this, they have deployed a dense network of weather stations across the country, aiming for accuracy never seen before. Yet, even with all this data, the perfect forecast does not exist. Kachelmann and his colleagues still make mistakes. In this interview, he explains how forecasts are actually built and why you should be skeptical of long-range predictions.
The Myth of the Two-Month Forecast
Recently, scientists in the journal Science reported that events in the stratosphere can “drill down” into the troposphere—our lower atmospheric layer. This connection correlates with specific weather patterns. Headlines immediately jumped to the conclusion that weather would soon be predictable two months in advance. Is this realistic?
Kachelmann is blunt. While there is a kernel of truth to the research, the media spin is nonsense. The behavior of the stratosphere does provide clues for atmospheric developments. It helps meteorologists calculate the probability of warm or cold winters better. But it does not mean you can predict the weather on a specific day next month.
Why Daily Forecasts Are Still Necessary
Does this mean the daily weather report is obsolete? No. Automated tools can produce good results, but humans are still needed. Even if we understand the stratospheric links better, we are not close to hearing a forecast for Berlin on January 1st for the weather on February 26th. That is pure science fiction for now.
The daily forecast remains essential because short-term chaos matters. People need to know whether to carry an umbrella tomorrow, not in two months.
The Problem with the Official Weather Service
Kachelmann’s direct competitor, the German Weather Service (DWD), has lost some of its prestige. He suggests his colleagues may have slept at the wheel. Meteomedia regularly wins forecast competitions, while the DWD struggles with its reputation.
What is going wrong? Kachelmann argues that the DWD wastes taxpayer money on wrong goals. They cut staff at weather stations and sell their services at dumping prices. In return, they give “fun” radio interviews while neglecting their sovereign duties. He prefers a strong, professional partner over a wobbling giant supported by millions in annual subsidies. The DWD also makes the mistake of relying on a single computer model to predict weather for every district of a city. That is insufficient.
The Difficulty of German Terrain
Is Germany a difficult terrain for forecasting? Not particularly. It is diverse, yes. Mountains and seas are part of the equation, but they are manageable.
If Kachelmann has a favorite area, it is the one that challenges him the most. The islands and the Alps are his “favorite forecasting regions.” That is why he prefers to live there. Difficult weather means more interesting work.
How a Forecast Is Actually Made
People often wonder how long it takes to prepare the evening forecast. It is not a quick job. The team works in weekly shifts to build a sense of continuity. A typical day lasts nine to thirteen hours. This does not even include the time spent presenting the weather on air. Since January 2002, they have produced ten daily broadcasts. The morning show alone requires seven segments at half-hour intervals.
Do they just plug numbers into a program? No. They use five different computer models. They also rely on their award-winning MOS (Model Output Statistics). This method compares statistical data from past weather with current model data and measurements. The rest happens in the head. Data collection and interpretation are manual processes.
Gut Feeling vs. Computer Data
Do they trust the numbers or their intuition? Looking at the sky helps, but only in summer. If you see Altocumulus castellanus clouds, those are perfect precursors to thunderstorms. Otherwise, intuition plays a role of about ten to twenty percent. Routine is key. Kachelmann has been working since 1983. That experience matters.
He is still involved in the actual calculations. He does not just present the weather. The team refuses to use a teleprompter. Reading from a screen would exclude input from others. Their announcements are memorized. They create the forecasts they present.
Climate Change and Future Forecasts
Kachelmann follows the debate on climate change with interest and concern. He sees a problem. The question now is whether climate change will make daily forecasts harder in the future. The stakes are rising.
The daily work of predicting tomorrow’s weather does not become quite philosophical in these discussions. We know the odds change. The consensus is that we will experience extreme events like Hurricane Lothar more often in the long term. But when it comes to predicting these particular chaotic disasters, we have generally kept a grip on them.
So if you keep winning weather forecasting competitions, can a chaotic system like the atmosphere ever have a perfect forecast? Or to put it bluntly: would you always bet on your own forecast?
You would bet on it than on the results of other forecasters. But let’s be clear. We are not perfect. We never will be.
Have I ever made a completely wrong prediction? You know the one. I didn’t call for snow in the Swiss Alps, so I drove straight into a whiteout and missed an important meeting?
That didn’t happen. But I got close. In 1984, the Föhn wind collapsed twelve hours earlier than I had predicted. Instead of the promised 20 degrees of sunshine, it rained and it was 8-degree chills. It was at the beginning of my career. Since then I have not made such mistakes.
What do I love most about the weather? Clouds? Rain? Snow? Or just the colors of the sky?
I want my daily snowstorm.
When was the last time I stared at the clouds, lost in a daydream? Or has it stopped?
I don’t look at the sky anymore. I just look at it meteorologically.
The interview was conducted by Marcus Anhäuser.