distributed regions (metro-agro-plexes) that depend sensitively on the particular blend of industrial activity, agricultural product, moisture level, temperature, and national and local infrastructure and priorities. The observational requirements are specific for the species and the trajectories, though the specifics may vary for specific national boundaries. The canonical suite of simultaneous in situ observations obtained with 0.1-km resolution is O₃, NO, NO₂, OH, VOCs, HONO₂, HO₂, CH₃OCH₃, PAN, DMS, SO₂, H₂SO₄, HCl, aerosol composition, number, size, and mass as a function of size, and tracers, specifying the region of origin of the air mass.

This imperative requires definition of the production and loss mechanisms, distribution, and optical properties of aerosols. Observations must be directed at processes that control aerosols from the fine scale to the global scale. Specifically, observations must clarify the following: (1) mechanisms controlling the rates of production of aerosols from those gases that are relevant to both direct and indirect forcing; (2) processes controlling the evolution of aerosols, including growth, activation to cloud drops, and wet and dry removal; (3) relations between aerosol optical depths and aerosol properties; (4) roles of specific chemical classes of aerosols, such as organics, in direct and indirect forcing; and (5) cloud-activating properties of different classes of ambient aerosols.

The character of scientific analysis in addressing the aerosol problem is critical to achieving progress because of the close but complex linking among chemical, biological, and physical processes. In particular, a critical strategy is to establish the relationship between key dependent variables (such as aerosol light scattering and absorption coefficients, number concentration of cloud condensation nuclei [CCN], etc.) and the major independent variables and then to test that functional dependence over a large dynamic range of variables. This strategy shares much in common with the approach of analyzing the structure of stratospheric ozone photochemistry, taking the form of the systematic analysis of partial derivatives linking dependent and independent variables. The explicit observation of these derivatives, or response function, provided the key evidence that overturned central tenets in ozone chemistry; it is the approach required in the field of aerosol chemistry.

This point is directly addressed in another report, Aerosol Radiative Forcing and Climate Change,¹⁷ which casts the problem in terms of carefully designed “closure experiments”—experiments in which an overdetermined set of observations is obtained, and the measured value of a dependent variable is compared with the value calculated from measured values of the independent variables. This approach requires fundamental restructuring of both the observations and the architecture of the modeling effort. The key point is that, through a sequence of these analyses comparing calculated and observed variables and their associ-