Comparisons of temperature response to solar forcing in the pre- and post periods of satellite data assimilation



On the basis of both composite and multiple linear regression analysis, three temperature datasets from two reanalyses and one set of satellite observations have been used to evaluate the different responses in the winter [December–February (DJF)] period as they relate to solar forcing between the pre- and post periods of satellite data assimilation. The two periods are defined as 1958–1978 when no satellite data was available to be assimilated and the 1979–2002 period when satellite data was assimilated in the operational forecast models. The solar forcing signal for the two periods can be identified as significant in the extremes of the solar cycle (maximum and minimum). The composite analysis shows that the solar response of the DJF temperatures in the three datasets shows large-scale similarities although there are differences over the southern middle-high latitudes and some tropical areas. The stratospheric response showed the strongest DJF temperature anomalies related to solar variability occurring over the Arctic, but its sign is negative in 1979–2002 and positive in 1958–1978 for solar maximum activity. The stratospheric temperature response can be confirmed using a multiple regression analysis. The temperature features may be partially explained by the impacts of heating via ozone absorption, and is reinforced by the solar cycle, El-Niño Southern Oscillation (ENSO), stratospheric quasi-biennial oscillation (QBO), volcanic eruptions and other factors. In contrast, the tropospheric response, with a dynamic wavelike structure, occurs over the middle latitudes. The tropospheric differences between the two periods are not clearly resolved and raise questions about the efficacy of the observations and our ability to use the observations effectively. Copyright © 2010 Royal Meteorological Society