Developing a Historical Precipitation Record - Satellite-based Applications on Climate Change - page 97

Fig. 7.1 Annual and 75 S-75 N average precipitation anomalies from the RCCA, GHCN, and
GPCP. Averages omit data where GHCN is not available
Fig. 7.2 Ocean-area averaged precipitation anomalies from the blended EOF-based reconstruc-
tion, REOF(blend), the RCCA, the AR4 ensemble coupled models, and the EOF-filtered GPCP,
REOF(GPCP). All are low-pass filtered to emphasize multi-decadal variations
interannual and multi-decadal averages. Testing showed that most of the RCCA
variance is derived from the SST predictor, including most of the land-area
variations. However, adding the SLP improves the skill and it is retained in the
analysis.
The RCCA was motivated by the need for better resolution multi-decadal
oceanic variations compared to the REOF(blend). Comparisons of these analyses
averaged over ocean areas show the differences (Fig. 7.2 ). Here the annual-spatial
averages are filtered using a weighted 7-year filter (see Smith et al. 2010 ). In
addition, the ensemble average of fourth Assessment Report (AR4) coupled models
is also shown through 1999, along with REOF(GPCP) since 1979. The AR4
coupled ocean-atmosphere models to help evaluate climate change from changes
caused by greenhouse gases and aerosols (Randall et al. 2007 ). Note that for the