Climatology is a reference forecast provider (ID 89) that uses historical average conditions as its “forecast.” It serves as a skill baseline, as any forecast that provides value should be able to outperform the climatological average.
In the weather industry, a model is not considered “skillful” simply because it is often correct; it must be more accurate than the historical average for that specific date and location. For example, forecasting a high of 75°F in Los Angeles during July is likely to be accurate, but because that is the climatological norm, the forecast hasn’t added new information. Professionals search for “Skill Scores relative to Climatology” to see if a specific provider is actually capturing the unique weather anomalies of the current year.
As the forecast horizon (Days Out) increases, the skill of most numerical models tends to degrade toward the climatological mean. By comparing a Day 10 forecast against Climatology, analysts can determine the “predictability limit”—the point at which the model is no longer providing better information than a simple historical average.
Yes. While Climatology looks at what happens on average over decades for a specific date, Persistence looks at what happened yesterday. In stable patterns, Persistence is the harder baseline to beat, but over long periods, Climatology is the standard against which “Anomalies” (deviations from the norm) are measured.