UNCERTAINTIES IN GLOBAL CLIMATE CHANGE OBSERVATIONS AND NUMERICAL MODELING

摘要

It is extremely difficult to understand the laws of the present climate system and the more so to assess potential climate changes in the future. This is confirmed by lack of reliable estimates of the contribution of anthropogenic factors to the formation of the present climate with understanding that, for example, the anthropogenic enhancement of the greenhouse effect (due to the growth of GHG concentrations in the atmosphere) should cause certain changes in global climate. In this connection, a primitive understanding of the global warming as a general increase of temperature, growing with latitude, is rather dangerous. Analysis of the observed data obtained in high latitudes of the Northern Hemisphere [1] has shown that such judgements do not correspond to reality. To assess the reality of climate predictions, it is critically important to test the adequacy of models from the viewpoint of reproducing the present observed changes and paleodynamics of climate (from indirect data). The situation with the use of the present observed data is rather paradoxical: an experience in testing the adequacy is confined to the use of average temperatures while it is necessary to use different information and moments of a higher order. Goody [39] drew attention to the prospects of using the space- borne observations of the spectral distribution of the outgoing longwave radiation. Unfortunately, the problem of an adequate planning of the climate observing system has not been recognized, as it should be [7, 9, 52]. The present paradoxical situation is characterized by the fact that a huge amount of poorly systematized satellite observations is combined with the degradation of conventional (in situ) observations mentioned above.