An Analysis of Forced and Internal Variability in a Warmer Climate in CCSM3

Changes in the mean state and the modes of internal variability due to increases in greenhouse gas (GHG) and aerosol concentrations were investigated by comparing a suite of long-term integrations of A1B runs and the corresponding control runs with a constant level of GHG and aerosol concentrations in the Community Climate System Model, version 3 (CCSM3). The evolution of signal- [defined as the standard deviation (STDV) of ensemble mean anomalies] to-noise (defined as STDV of departures of individual members from their corresponding ensemble means) ratio (SNR) is examined. It is shown that SNR is sensitive to the amplitude of external forcing, and the sensitivity is variable and geographical location dependent. The time evolution of the SNR is largely due to the changes in the mean while little influence on the internal variability is found. Surface air temperature (TS) and geopotential height at 200 hPa (H200) responses are largely linear with an increase in GHG and aerosol concentrations and can be well reconstructed using linear trends. The spatial patterns and temporal evolution statistics of the leading modes of internal variability of seasonal mean TS, H200, and precipitation are similar between the A1B and control runs, suggesting that the leading modes are less affected by the increase in GHG and aerosol concentrations. However, the similarity of these spatial patterns between the two runs slightly depends on the variable and season. In the tropical Pacific Ocean, superimposed on a warming trend, amplitude of internal variability in the El Nino-Southern Oscillation regions is slightly suppressed in the A1B runs.