The IGAC/ACAPS and MAGE committees chose the remote marine atmosphere for the first ACE process study. Non-sea-salt (nss) sulfate aerosol particles in the remote marine atmosphere are thought to have only one primary gaseous precur-sor, dimethylsulfide (DMS), thereby simplifying studies involving the formation and growth of the aerosol. This environment also provides an opportunity to establish the chemical, physical and radiative properties of the natural aerosol system and thus provides a background from which to compare and quantify any anthropogenic perturbation.

The specific goal of ACE-1 is to determine and understand the properties and con-trolling factors of the aerosol in the remote marine atmosphere that are relevant to radiative forcing and climate. To achieve this goal, the ACE-1 Science Team has defined three specific objectives:

To achieve these objectives the ACE-1 Scientific Steering Committee has identified six specific questions that must be answered. The science behind each question and strategies for answering it are discussed in Chapter 3, while the implementation of that strategy in terms of specific instruments and platforms is described in Chapter 5. The questions are:

2.2 ACE-1 Site selection

The mid-latitude Southern Hemisphere marine site near Tasmania was chosen for the initial experiment for several reasons:

2.3 Summary of regional climatology

Jasper and Downey (1991) have compiled a comprehensive discussion of the Cape Grim climatology. The brief summary here draws from that report and other Cape Grim data. Cape Grim (40.7S, 144.7E) is located on an exposed headland in the roaring forties. The strength and direction of the predominantly westerly flow is governed largely by the location of the subtropical high-pressure ridge. "Baseline" air, that is wind with a local direction between 190 and 280 degrees, typically is observed for several days following a frontal passage. The mean period between fronts and transition into "baseline" conditions during November-December over the period 1977-1992 was 5.8 days and the average length of baseline conditions following was 2.4 days (selected for occasions where baseline sector wind was observed for more than 24 hours). According to the Jasper and Downey climatol-ogy, during November-December "baseline" conditions occur on average 40% of the time during the day, decreasing to around 30% at around 0100 eastern Australian standard time (EAST). For "baseline" conditions the mean wind speed over the period 1977-1992 was 10.6 m s with a standard deviation of 1.1 m s.

For November-December the hourly mean temperature maximum (for all wind directions) is 15-16C with a nighttime minimum of 14C. The hourly average humidity reaches a minimum of 75% at around 1400 and a maximum of close to 84% in the early morning, around 0400. The mean monthly rainfall for the period is around 43 mm.

Monthly mean CN hourly concentrations in baseline air (1977-1991) for November-December are 427 cm with a standard deviation of 89 cm. Monthly means have varied from 259 cm to 651 cm. For CCN active at 0.23% supersaturation the monthly mean value for the period 1981-1992 was 81 cm. The November -December period is slightly before the average peak in DMS production. For example, for this period in 1991 atmospheric DMS concentrations varied from around 1 to 4 nmole m(Gillett et al., 1993). Radon concentrations in 1990 were around 104 mBeq m in baseline conditions and averaged 647 mBeq m in non-baseline conditions.

Average cloud cover can be ascertained from the Cape Grim solar radiation measurements (determined by B. Forgan, Bureau of Meteorology). For the period 1986-1992 monthly average cloud cover for November-December varied from 50% to 80% with a mean for this period close to 65%. The average for the general region to the south west of Cape Grim can also be determined from Warren et al. (1988) that was compiled from surface (ship) observations taken between 1952 and 1981. For the two cells west of Tasmania (40S to 45S and 135E to 145E), the frequency of occurrence of low stratus and stratocumulus clouds varied from 56% in spring (SON) to 66% in summer (DJF) although when cloud is present the fractional cover is slightly larger in spring (79%) than in summer (72%). The average total cloud cover is similar for both periods, 75% in spring and 72% in summer. In some respects, however, these averages may be misleading, since cloud cover is exceedingly variable as shown by time series of satellite observations for this region. For any given month the long term average can be a very poor estimator for the amount of cloud observed. This is a baroclinic region, and as such, variability and change must be expected and anticipated.