A camera on a tripod shoots the same landscape for decades. If you don't touch the lens, the images get hazy: the landscape changes, but the focus stays put. That's roughly how climate norms work. According to World Meteorological Organization rules, they are calculated as the average over the last 30 years. But the climate doesn't stand still, so the norm starts lagging behind reality, like an old lens chasing a moving subject.

The authors propose a different tool: a filter that adjusts focus to motion. Instead of simply averaging 30 years, they fit a local model at each point in time: a constant, a linear trend, and six seasonal waves. Then they regularize, meaning they find a compromise between a filter without trend and one with trend. Parameters are chosen to minimize the error in estimating the norm, accounting for autocorrelation in anomalies. The result is a lens that auto-focuses.

Tests on sea surface temperature in the Niño 3.4 region showed that the standard 30-year lens is almost as good as the new one. The average optimal window width is 29.38 years, and anomalies differ from NOAA estimates by no more than 0.3 °C. For temperature, the old focus still works.

But for winds in the tropical Pacific, the picture is different. Out of 228 series, the standard filter was optimal in only three cases. The old lens systematically underestimates wind norms, so anomalies are overestimated by up to 2 m/s. It's like a speedometer reading 2 meters per second faster than the actual speed: seems minor, but for a sailboat or an atmospheric model, the error is critical.

The new filter is not just a replacement for the average. It's a way to focus the lens on a moving climate. For sea surface temperature, it confirms that the standard is fine. For winds, it shows for the first time that ignoring the trend distorts anomalies. Now that the camera is focused more sharply, we can discern the true El Niño signals without mistaking them for a defect in the optics.