Abstract
Daily temperature and precipitation data from Japanese stations are studied to estimate climate noise, an unpredictable part of long-term variability. The climate noise is compared to the interannual variance of monthly mean temperatures, and monthly precipitation totals, to access the potential predictability. Potential predictability is that part of the interannual variance that exceeds climate noise. Primarily during July and January, it is found that potential predictability of temperature can exceed 40% at some stations. Less potential predictability is present during April and October. The results are similar for monthly precipitation totals.
