What Do We Not Know and What Do We Need to Know?

The response of clouds and their radiative effects to global warming represents a longstanding and considerable area of uncertainty in our understanding of climate change. At present, it is not known whether changes in cloudiness will exacerbate, mitigate, or have little effect on the increasing global surface temperature caused by anthropogenic greenhouse radiative forcing. Another substantial uncertainty is the magnitude of radiative forcing resulting from the modifi cation of cloud properties by anthropogenic aerosols. Global climate models provide scant reliable insight regarding these issues because of their inability to parameterize correctly or otherwise represent the small-scale convective, turbulent, and microphysical processes that control cloud properties. It is therefore crucial to document and assess global and regional low-frequency variations in clouds and radiation fl ux that have occurred over the past several decades, a period marked by rapidly rising temperature and changes in anthropogenic aerosol emissions. This will enable us to estimate from observations how clouds and their impacts on the radiation budget are responding to global warming and aerosol changes. Moreover, a trustworthy observational record will provide a good constraint on global climate model simulations. Previous investigators have documented multidecadal variations in various cloud and radiation parameters, but no conclusive results are yet available. Problems include the lack of global and quantitative surface measurements, the shortness of the available satellite record, the inability to determine correctly cloud and aerosol properties from satellite data, many different kinds of inhomogeneities in the data, and insuffi cient precision to measure the small changes in cloudiness and radiation that nevertheless can have large impacts on the Earth’s climate. Many of these defi ciencies emerge from the 18 J. R. Norris and A. Slingo absence of an observing system designed to monitor variations in clouds and radiation on timescales relevant to climate; to compensate, observations must be assembled from a system originally designed for purposes of weather forecasting. Although we cannot go back in time to make more and better observations, we need to improve our processing of the available historical measurements to mitigate inhomogeneities, provide better retrievals of cloud and aerosol properties, and extend the record farther back in time. Furthermore, it is essential to construct an observation system with suffi cient stability and longevity to measure long-term variations in cloudiness and the radiation budget with improved precision and accuracy. Our present observing system has unfortunately little prospect of enhancement at this time and, moreover, is in danger of future deterioration since there are no defi nite commitments to replace several critical instruments when current satellite missions end.