Strong evidence exists for a widespread decrease in the DTR over the past several decades in many portions of the globe. The climatic factors that can affect the DTR are many, but indications are that cloud cover, including low clouds, has increased in many areas where DTR has decreased. These increases in cloud cover could be indirectly related to the observed global warming and increases of greenhouse gases (or perhaps even to the indirect effects of increases in aerosols), could be simply a manifestation of natural climate variability, or could reflect a combination of all three.
Obtaining a robust answer regarding the cause(s) of the decrease in the DTR will require efforts in several areas. First, an organized global effort is needed to develop relevant and homogeneous time series of maximum and minimum temperature, along with information on changes of climatic variables that influence the DTR, such as cloudiness, stability, humidity, thermal advection, and snow cover. Second, improvements in the boundary-layer physics and treatment of clouds within existing GCMs is critically important. Third, the treatment of both anthropic tropospheric aerosols and greenhouse gases must be realistically incorporated into GCMs with a diurnal cycle. Fourth, measurements need to be made to help clarify the role of aerosols. These include (1) data on diurnal, seasonal, and geographic changes in solar radiation, in areas influenced by aerosol forcing and areas free of it; (2) simultaneous data on key aerosol properties (mass-scattering efficiency, wavelength dependence, backscatter fraction/asymmetry factor), concentration, and composition; (3) data related to changes in all the relevant cloud parameters and their coupling to aerosol changes (if any); and (4) improved chemical models to allow more accurate calculation of the sulfate aerosol effects and to provide an estimate of the time course of this aerosol forcing. Finally, imaginative climate-change detection studies that link the observed climate variations to model projections will be required to convincingly support any relationship between anthropogenic changes and the DTR.
It will be difficult to satisfactorily explain the observed changes of the mean temperature until an adequate explanation for the observed decrease in the DTR can be determined. Moreover, the practical implications of projected temperature changes and whether they are likely to continue will be even more difficult to assess.
This work was supported by a U.S. Department of Energy/National Oceanic and Atmospheric Administration (NOAA) interagency agreement and NOAA's Climate and Global Change Program.
We thank the following scientists for providing us with additional data: Reino Heino for the Sodankylä data, and Takehiko Mikami for the Japanese data.
GOODRICH: Tom, were you able to use the ISCCP data sets to test your hypothesis that the decrease in the annual range of cloudiness might be associated with temperature?
KARL: I think there might be problems with both the longevity and the quality of those data, and the solar radiation data network is not as comprehensive as one might wish. We really need to get cloud and temperature experts together to see whether relating the two makes sense.
CHARLSON: To determine how cloud cover affects heat balance you would really need to know cloud liquid-water content and mean droplet size, as well as cloud area. All three govern the system albedo. The ISCCP data don't have those, and I'm not sure our current remote-sensing systems have that capability either. If we found a correlation, we wouldn't know whether it was caused by cloud microphysics or dynamics.
SHUKLA: Tom, one of your most dramatic differences was that between the rural and urban areas of Japan. Would it be naive to ask whether radiative forcing associated with urban pollution could be involved?
CHARLSON: Urban haze around most cities actually extends for hundreds of kilometers. Also, neither the soot aerosols nor the light-absorbing gases in smog cause any great amount of heating.
KARL: Various studies have shown that maximum temperature is not much affected by urbanization. But it is a significant point to keep in mind, especially given how it dominates the record in Japan.
KEELING: Many of you have probably heard of the Arctic and Antarctic borehole data that suggest that the heat flux from the interior of the earth is not steady state. Is it possible that these borehole temperatures are showing the effects of nighttime radiative warming?
CHARLSON: The next set of NCAR model calculations will be simulating diurnal temperature range, and a model in Hamburg is doing that too. It will be interesting to hear their results.