Global Average Net Radiation Sensitivity to Cloud Amount Variations

Abstract

Time series analysis performed using an autoregressive model is carried out to study monthly oscillations in the earth radiation budget (ERB) at the top of the atmosphere (TOA) and cloud amount estimates on a global basis. Two independent cloud amount datasets, produced elsewhere by different authors, and the ERB record based on the Nimbus-7 wide field-of-view 8-year (1978?86) observations are used. Autoregressive models are used to eliminate the effects of the earth's orbit eccentricity on the radiation budget and cloud amount series. Nonzero cross correlation between the residual series provides a way of estimating the contribution of the cloudiness variations to the variance in the net radiation. As a result, a new parameter to estimate the net radiation sensitivity at the TOA to changes in cloud amount is introduced. This parameter has a more general character than other estimates because it contains time-lag terms of different length responsible for different cloud-radiation feedback mechanisms in the earth climate system. Time lags of 0, 1, 12, and 13 months are involved. Inclusion of the zero-lag term only shows that the albedo effect of clouds dominates, as is known from other research. Inclusion of all four terms leads to an average quasi-annual insensitivity. Approximately 96% of the ERB variance at the TOA can be explained by the eccentricity factor and 1% by cloudiness variations, provided that the data used are without error. Although the latter assumption is not fully correct, the results presented allow one to estimate the contribution of current cloudiness changes to the net radiation variability. Two independent cloud amount datasets have very similar temporal variability and also approximately equal impact on the net radiation at the TOA.