最后,morton 1990年的部分转载如下: The third lesson is thal we know surprisingly little about hydrology. In fact we know so little that we think we know a lot. Until such time as this illusion is abandoned, the financial resources and scientif ic skills will continue to be directed away f rom the kind of original research that is needed to jolt hydrology out of its long period of stagnation.
The fourth lesson is that a similar situation exists in many of the related environmental sciences. For example, ldso (1984a, 1984b) and ldso and Brazel (1984) have criticized the predictions of the Carbon Dioxide Assessment Committee of the U.S. National Research Council (1983) that a doubling of the atmospheric carbon dioxide content would cause a near surface temperature rise of approximately 3 degrees Celsius and would be associated with a 40 to 75 percent reduction in the streamflows for the major watersheds in the western United States. The grounds for these criticisms are (1) that the prediction is contrary to the available evidence; (2) that the models used to produce the predictions ignore the existence of a number of confutative feedback mechanisms, one of which concerns the way that increased carbon dioxide concentrations decrease transpiration by causing the stomala to contract. Moreover, after discussing feedback mechanisms that could affect the relationshio between over- grazing and desertif ication, ldso (1 981) concluded by stating: "Consequently, more basic research into boundary layer processes appears necessary to provide the additional and improved algorithms required to make the (general circulation) models more descriotive of the real world".
The fifth lesson is that there is a real need for an alternative approach to environmental research. Such an approach should involve a study of natural systems as they actually exist after feedback mechanisms have modified, or even reversed, the effects of small-scale orocesses. lt is exemplified in the Gaia hypothesis (Lovelock, 1979, 1986), the accomplishments of the Bergen School of Meteorology (see Jewell, 1981)and the development of the complementary relationship models (Morton, 1983a, 1986a) for estimating actual areal evapotranspiration and lake evaporation. Although not used nearly as often as the linear cause-and-effect approach, this alternative has a much better record of success.