Introduction

The ability to measure cloud parameters has advanced considerably over the past fifty years, in terms of instrumentation and analysis techniques. Earlier (e.g., Squires 1958), drops were captured on a media such as glass slides exposed to the airstream from aircraft platforms; after suitable corrections, this gave a measure of cloud droplet number concentrations. Those measurements were used to develop the concepts that maritime clouds had droplet concentrations smaller than continental clouds and typical values were assigned to each. That type of characterization still exists today, very often using textbook-type values of microphysical parameters. In this chapter, we explain why it is necessary to get away from such simple approaches and measurements and start to consider clouds as more complex, stressing the use of the latest in-situ and remote-sensing probes.

First, it is necessary to consider the accuracy to which cloud properties must be measured to simulate them properly in numerical model simulations of our climate. There have been a few sensitivity studies, such as those by Slingo (1990) and Rotstayn (1999).

The top of the atmosphere radiative forcing by doubled carbon dioxide concentrations can be balanced by modest relative increases of ~15-20% in the amount of low clouds and 20-35% in liquid-water path, and by decreases of 15-20% in mean drop radius. This indicates that a minimum relative accuracy of ~5% is needed.. ..to simulate these quantities in climate models (Slingo 1990, pp. 49).

The total indirect forcing is -2.1 W m 2...resulting.from a 1% increase in cloudiness, a 6% increase in liquid water path, and a 7% decrease in droplet effective radius (Rotstayn 1999, pp. 9369).

These studies suggest that cloud properties need to be measured with significant precision; Slingo suggested to within 5%. However, this is a very demanding, if not unrealistic goal with today's instruments.

Because of the increasing complexity in obtaining measurements and analyzing them, often the user obtains data from institutional data banks and is only vaguely aware of all the associated problems. Problems such as instrument calibrations, expected accuracy, software analysis problems, and scale effects, for example, are rarely understood. This chapter will hopefully provide a short primer to allow appropriate questions to be asked before the provided data is used.

0 0

Post a comment