FIGURES

• Figure 1
• Figure 2
• Figure 3
• Figure 4

A global analysis of convective cloud near the analyzed tropical tropopause is presented. The goal is to understand how convective processes affect the tropopause region (from 14--18km) and the direct role of convection in stratosphere-troposphere exchange. The analysis is based on high-resolution global cloud imagery (GCI) of cloud brightness temperatures from satellites and from reanalysis tropopause temperature data.

The frequency of deep convection above the tropopause increases with decreasing tropopause temperatures,as illustrated in Figure 1 for two months.ÝThe relationship is stronger in February (Fig 1A) than in August (Fig1B). In February (Fig 1A) as the tropopause is colder, the depth of convection above it increases.The deepest convection is rarely found at tropical tropopause temperatures warmer than 200K. In addition, convection in the tropopause region is different over land and ocean with deep convection and the tropopause temperature having a stronger diurnal cycle over land.

FIGURE 1: Distribution of GCI brightness temperatures as a function of coincident tropopause temperature for tropical points between 20 deg S and 20 deg N. A) February and B) August. Solid line is the 1:1 line where GCI temperature equals Tropopause Temperature.

In active convective regions clouds penetrate the tropopause less than 1--2\% of the time, and penetration generally does not extend more than 2km above the tropopause. The distribution of convection in the tropopause region and above the tropopause is shown in the zonal mean in Figure 2, also for February and August. Note that there is more convection above the tropopause in August (Fig 2B). The role of convection increases nearly exponentially below the tropopause.

FIGURE 2: The zonal mean vertical distribution of convective events as a function of latitude and temperature for A) February and B) August. Convection is expressed as cloud fraction (h). Thick dashed lines are the monthly averaged zonal mean NCEP thermal tropopause temperature. Contours at 0.005, h = 0.01,0.02,0.05 and 0.1.Ý

Figure 3 illustrates the location of convection above the tropopause during each season. Consistent with Figure 2, the maximum convective activity is during Northern Hemisphere summer (July to September). Convection above the tropopause follows convective activity throughout the year. 1986-7 shown is an ENSO warm event, so the maximum convection in January- March is in the central (not western) Pacific. In Figure 3 there is a strong relationship between cold tropopause temperatures and convection above the tropopause. Correspondance is not direct, but cold tropopause temperatures lie near regions of deep convection.

FIGURE 3: The distribution of convective events colder than the tropopauseÝ for each of four 3 month seasons. October--December (top left), January--March (top right), April--June (bottom left) and July--September (bottom right). Plot range is 30S to 30N. Thin black lines are seasonally averaged NCEP thermal tropopause temperature. Contour interval is 1 deg K up to 196 deg K and 2 deg K for warmer temperatures.

The upward convective mass flux across the tropopause can be calculated by using the temperature difference above the tropopause at each point to estimate a height, and then multiplying volume this height represents by a standard atmospheric density and integrating across events. The upward convective mass flux across the tropopause has a strong annual cycle illustrated in Figure 4. The maximum convective mass flux. in the summer in each hemisphere, and a global maximum in July--September.Ý

FIGURE 4: Total upward convective mass flux by month into the stratosphere (in units of 10¹⁶ kg per month) estimated by the volumetric method (see text) for the Northern Hemisphere Tropics 0-20N (dashed), Southern Hemisphere Tropics 0--20S (dotted) and the Global tropics 20S--20N (Solid). Error bars represent the range of monthly values over 3.5 years.

On an average basis, convection is sufficient to overturn the tropopause region once a month. However,
from the analyzed tropopause and above, convection turns over the air mass in the lower tropical stratosphere only every 6 months, compared to 3 months for the background diabatic ascent.
Convection is sufficient to ventilate the tropopause region, but not sufficient to supply all the mass into the lower tropical stratosphere.

For a complete copy of this work please visit: http://www.asp.ucar.edu/~andrew/papers.html

Ý