MIKE MOSS SAYS: Mike, At first blush that would seem to make sense, and for some applications it may actually be a better approach. However, the use of thirty-year averages ending with the most recent "zero" year for climatological "normals" was agreed upon by the World Meteorological Organization as a means of:
1. Establishing comparative climatological values that are standardized from one region or location to another, so that you don't end up trying to compare likely conditions at station A with a 37-year history against station B that has been around for 72 years, for example, and
2. As a compromise between wanting the "normals" to include a significant period of time long enough to be reflective of a full or nearly full range of potential year-to-year variability in weather conditions for the location of interest while at the same time being short enough to be sensitive to real changes in the climatology of a given location, whether those changes might be due to local/regional urbanization and building patterns, natural cycles like the Pacific Decadal Oscillation or multidecadal sea surface temperature variations in the Atlantic, or large-scale variations associated with anthropogenic gas and particulate emissions.