The Ingredients of an Observational Strategy

To begin to determine the magnitude of the aforementioned potential cloud perturbations, the overall objective of the observational approach must be to quantify the respective susceptibilities of cloudiness or, more specifically, cloud radiative properties, 8c, to changes in the general circulation and to internal microphysical changes induced by the aerosol. This approach is presented in Stevens and Brenguier (this volume) as:

8m JJUa L UG

where c represents cloudiness, M is the large-scale meteorology, G stands for greenhouse gases, and A for aerosols. A specifically refers to the subset of the aerosols that may impact cloud microphysical properties, namely cloud condensation and ice nuclei (CCN, IN) and absorbing aerosol particles scavenged in hydrometeors.

The term in the braces is the large-scale meteorological forcing, SM, related to the aerosol and greenhouse gas forcings, both of which are the result of integrated radiative forcings on large spatial scales. Thus, from Equation 21.1 we can represent cloud perturbations as:

where and XM are the sensitivities (susceptibilities) of clouds to perturbations in Aand M, respectively.

In designing a strategy for an observational program to assess the impacts of aerosols upon clouds, it is thus essential to ask whether variability in observed cloud radiative properties, Sc, will be dominated by variability in aerosols or by the varying meteorological forcing. Equation 21.2 shows clearly that to determine the sensitivity of clouds to aerosol perturbations XAji through observations, we must first understand the impacts of meteorology upon the cloud system, as expressed by XM.

The meteorological sensitivity, XM, is poorly known in many cases, and thus it is a major challenge to use observations to determine it. For marine stratocumulus clouds, there has been considerable success in relating cloud properties (e.g., cloud fractional coverage) to the large-scale meteorology using observations (e.g., Klein and Hartmann 1993; Klein et al. 1995; Wood and Bretherton 2006). As Stevens and Brenguier demonstrate (this volume), the meteorological sensitivity is very high in many cases, and this will limit our ability to attribute perturbations in clouds to those in aerosols.

We can further quantify the uncertainties in determining the sensitivity to aerosols by considering the variability across a set of measurements that would be made in a particular observational campaign. If we define a 2 as the observed variance in parameter x across this set of measurements, Equation 21.2 can be used to show that:

where r is the correlation coefficient between the A and M Equation 21.3 demonstrates that not only do we need to understand the meteorological variability and its impact on the clouds, we also need to understand to what extent the meteorological variability covaries with the aerosol variability. In other words, accounting for the meteorological variability in a dataset (i.e., the second term in Equation 21.3, through knowledge of 1M) is not sufficient to determine fully the aerosol sensitivity, if one does not also understand how the aerosol properties are tied to the large-scale flow. In addition, we note that such a formulation assumes a one-way cause-and-effect relationship, at least on short timescales, between aerosols and clouds and is a simplistic representation of a tightly coupled system in which clouds and aerosols interact on all scales. For instance,

0 0

Post a comment