🧮 The drift you can't see coming
Sensor drift is rarely dramatic. A thermistor doesn't fail — it ages a fraction of a degree per year. A humidity element doesn't break — it slowly loses sensitivity at the high end of its range first, so your station starts looking accurate on dry days and wrong only during humid ones, which is exactly when you're least likely to notice. The pattern that gives drift away isn't a single bad reading; it's a slow, consistent gap between your station and a trusted reference that grows month over month rather than appearing and disappearing randomly.
📖 Why one homeowner's rain total was off by 40%
A station three streets over from mine logged barely half the rainfall the rest of the neighborhood recorded during a fall storm. The owner assumed the gauge was simply in a drier microclimate. Eight months later, cleaning the funnel revealed a tightly packed mat of cottonwood seed fluff that had been silently filtering out most of the water all season, a little more each month as it compacted further.
💡 The five-minute sanity check
Before any deep audit, check whether your reported dew point is ever higher than your reported air temperature. Physically, dew point can equal temperature (at 100% humidity) but can never exceed it. If your log ever shows dew point above temperature, that's not weather — it's a calculation or calibration fault in either the temperature or humidity channel, and it's worth tracing down before trusting anything else that month.
🔍 Reading your own history like a detective
Most citizen-science weather platforms let you overlay your station against the three or four nearest neighboring stations on the same graph. Do this once a season, not just when something looks wrong. A station that tracks its neighbors closely on temperature but consistently runs drier on rain, or windier on gusts but normal on sustained wind, is telling you something specific about which single instrument needs attention rather than a wholesale recalibration.
🔧 What's actually worth keeping a spare of
Anemometer bearings and rain gauge tipping mechanisms are the two components most likely to need mid-season attention because they're the only parts with continuous moving contact. Keeping a spare bearing kit or a complete replacement rain collector on hand means a failure becomes a same-day fix instead of a multi-week data gap waiting on shipping, which matters most during the exact storm season you most want unbroken records.
📝 Treat this log as a second dataset
The real value of repeating this audit monthly isn't any single check — it's the audit log itself becoming a record of your station's mechanical history. When you eventually see a strange jump in your climate data, your first move shouldn't be to question the weather. It should be to flip back through your own audit notes and see what changed in the hardware around that date.
