Search

In this paper, radiosonde temperature time series (RTT) from 1958 to 2005 collected by the 116-station Chinese radiosonde network are examined. Quality control and homogenization are used to obtain a reliable RTT. The homogenization results revealed significant discontinuities in the RTT. Analysis suggested that 70% data availability is the minimum data requirement (MDR) for these RTTs. A new dataset is built by meeting this MDR, which reduced the number of potential stations from 116 to 92. Analysis on this dataset reveals that warming trends in the troposphere and cooling trends in the stratosphere were weakened by reducing the stations. Averaged RTT trends for China were generally consistent with those of global scale, but with some discrepancies. During 1958–2005, averaged temperatures in China tended to decrease in the lower stratosphere and upper troposphere, in contrast to warming trends in the mid- and lower troposphere. The trends varied with two different subperiods. For 1958–78, cooling trends in the entire atmosphere were similar to trends at the global scale. For 1979–2005, warming occurred in the lower troposphere, with the amplitude of the warming tending to weaken with increases in altitude and shifting to a cooling trend above 400 hPa. Seasonal trend structures suggest that warming in the lower troposphere is attributable to temperature increases in December–February (DJF); cooling in the upper troposphere and stratosphere was found mainly in June–August (JJA). Unlike with results of a larger spatial scale, a robust cooling layer was found around 300 hPa.

There have been numerous attempts to propose past warm time periods as “analogs” for a future greenhouse warming. In this paper it is argued that, although paleoclimate studies may provide important insights into process operating in the climate system, there may be no warm time period that is a satisfactory past analog for future climate. The future greenhouse warming may represent a unique climate realization in earth history. This conclusion is based on the following considerations: 1 ) comparisons with Holocene (9000 BP) or Eemian climates (120 000 BP) may be inappropriate because much of the variations in these climates can be explained in terms of seasonal rather than mean-annual forcing it has yet to be demonstrated that increased warmth for these intervals involved mew annual temperature increases that were globally synchronous, 2) comparisons with older and warmer climates (ex., Pliocene, Eocene, or Cretaceous) can be misleading because these warm periods had reduced polar ice cover, whe...