## Lab: Climatological Structure of the Atmosphere

### I. Introduction

The average of many individual weather events over time periods of a month or more defines the statistical characteristics of the atmosphere; this is referred to as climate.

For example, some summer days are warm and dry, others are oppressively hot and humid; this is a weather statement. But summer in New York is always hotter than winter; this is a statement about New York's climate .

In this lab we are going to examine the monthly mean vertical and latitudinal structure of some of the basic variables that describe the atmosphere's climate and general circulation: temperature, humidity, zonal wind speed, and meridional wind speed. The term "zonal" refers to the west-east direction, and the term "meridional" refers to the south-north direction. Recall that any two dimensional vector can be resolved into two components that are perpendicular to one another. Zonal and meridional winds are the zonal and meridional components of the observed wind. All of these are presented as average values over all longitudes (called a "zonal mean" since it is an average in the west-east, i.e. "zonal" direction) and over fifteen-year periods for each of the months of the year. The results can be thought of as representing typical cross sections of the atmosphere at one latitude.

A note about pressure coordinates: During our lab this week we will continually refer to the variation of meteorological field with pressure or on pressure surfaces. More specifically, we use pressure instead of height. This is common practice in meteorology. As we have seen when we examined the hydrostatic balance, pressure decreases with height. The decrease is not "linear" (that is the pressure change is not proportional to the height change) but "exponential". Moreover, it also depends on the density (or temperature) of the air: the denser (or colder) the air, the faster is the change of pressure with height. This is precisely why it is convenient for meteorologists to use pressure as a vertical coordinate rather than use height. When we do so, we do not need to continue and use density in our equations. The difference between two pressure surfaces is directly related to the mass of air between them and this is important when we look at fluid motion and need to consider balances of forces, such as the geostrophic balance. When we use pressure surfaces, the isobars are replaces by lines of constant height (in other words, we measure the changes in the height of the pressure surface at different points) and it is the height gradient that is driving the motion on a pressure surface. On a pressure surface air will initially flow from the regions where height is large (equivalent to high pressure) to where it's low (equivalent to low pressure). Coriolis force will deflect the flow such that it is parallel to the height contours moving in a clockwise (anti-clockwise) around low height centers in the Northern (Southern) Hemisphere and opposite around high height centers, just as with pressure on a height surface. Standard pressure surfaces in meteorology are 1000, 850, 700, 500, 300, and 200 millibars. In midlatitudes this is roughly equivalent to 100, 1500, 3000, 5000, 9000, and 12000 meters above the surface.

The purposes of this lab are to:

1. make you familiar with how meteorological parameters vary vertically, latitudinally (equator-to-pole and between the hemispheres), and seasonally,
2. make you think about how convective and baroclinic instability, and hence the occurrence of thunderstorms and synoptic-scale storms, vary over the world,
3. allow you to see the effects of differential insolation and the climate system's response to it,
4. help you develop a deeper understanding of why different parts of the world have different climates, and
5. introduce you to the use of the spreadsheet program Excel for the analysis of small data sets. If you have never seen Excel, please go through the Introduction to Excel before coming to lab. The introduction to Excel is also located in the Class Files section of Courseworks.

The data that you will view were compiled from all available radiosonde (weather balloon) profiles, ship and weather station measurements made over the fifteen year period 1958-1973. The horizontal coverage is global, with 2.5 degree resolution. Vertical measurements extend from the surface (1000mb) to the lower stratosphere (50mb), with measurements every 50mb. More information about this data set can be obtained from the OORT dataset documentation.

The OORT data set is different from the NCEP/NCAR Reanalysis set that we used in last week's lab. In the latter, a global climate model was used to assimilate the data while here we are looking at the actual observations. This direct averaging is faithful to the data but has some drawbacks. The data set contains numbers at all pressure levels and latitude bands because at least some observations have been made at all latitudes over this long a period of time. But the time average hides some important differences between one part of the world and another. In northern midlatitudes, where much of the latitude circle is covered by land and many of the nations have long been industrialized, the radiosonde network is dense and data are plentiful. But near the equator and in the Southern Hemisphere, where ocean covers a greater part of most latitude circles and many nations are just now developing, the radiosonde network is sparse and the averages are less reliable indicators of the true mean state of the atmosphere. In fact, the "Roaring 40's" of the Southern Hemisphere are almost entirely ocean, almost perpetually stormy, and are therefore rarely visited by humans. This part of the world is virtually unobserved by radiosondes. Thus, some parts of the climatology are more reliable than others. This sampling problem is the stimulus for the development of Earth-orbiting satellites that can observe every location on the planet and give us a true global picture.

### III. Glossary

Baroclinic instability / weather system
Baroclinic instability is caused by the latitudinal temperature gradient. A column of warm air is taller than a cold-air column containing the same amount of air because warmer air is less dense than colder air (hydrostatic equation). On the Earth, low latitudes are warmer than high latitudes, and so the atmosphere extends to higher altitudes in the low latitudes. The resulting pressure gradient (high on the equator side and low on the poleward side) is responsible for the generally eastward current of air (westerlies) in the midlatitude. But this also means that in low latitudes there is more air at higher altitudes than in high latitudes, creating a situation where potential energy (the energy resulting from elevation above the Earth's surface) is available for generating motion. One of the ways in which this potential energy is released is the generation of weather systems - the large-scale "rings" of low and high pressure that perturb the underlying eastward flow in the midlatitudes. This happens when the buildup of potential energy reaches critical levels and the atmosphere becomes "baroclinically unstable". Potential energy embedded in baroclinic instability is dissipated by weather systems in two ways; first, potential energy is converted to kinetic energy used in the rotation of the storm, and second, by moving warm air to the north and cold air to the south, thus decreasing the temperature gradient. The way this happens is that the geostrophically balanced atmospheric wind (jet stream) begins to form meander which grow bigger until they either break of the main jet or loose their energy due to friction. These growing meander are the storm systems that we observe every other day in the winter season in the middle latitudes.
Geostrophic balance and friction
Geostrophic balance represents a balance between the pressure gradient force and the Coriolis force. For the reasons discussed above, pressure gradients are inherently unstable - the atmosphere will try to dissipate the potential energy associated with the pressure gradient by moving air from high to low pressure. However, because the Earth is rotating, the flow of air is deflected by the Coriolis force. When these two forces exactly balance each other, as is often seen in the middle and upper portions of the atmosphere, winds travel parallel to the isobars (lines of equal pressure) and the atmosphere is said to be in geostrophic balance. In the lower portions of the atmosphere, there is another force involved in the balance - friction. Friction acts to decrease the speed of the winds traveling parallel to the isobars. Since the Coriolis force is dependent on the wind speed, it too decreases, and there is no longer a balance between the pressure gradient force and the Coriolis force. Winds can then move across isobars from high to low pressure. Fig. 1 illustrates geostrophic balance and Fig. 2 shows how friction modifies geostrophic balance.
Thunderstorms / convection
convection in the atmosphere occurs when rising air cools more slowly than its environment. When dry air rises, it expands and cools at the dry adiabatic lapse rate, 9.8° C per kilometer. If the air is moist and condensation occurs, the rising parcel cools more slowly due to the release of latent heat into the parcel. The slower rate of cooling is called the moist adiabatic lapse rate. The amount of latent heat released depends on the amount of moisture in the air. A typical value for the midlatitudes is about 6.5° C per kilometer, but in the tropics, where the air is very warm and moist, the moist adiabatic lapse rate is only about 4° C per kilometer. If the rate at which temperature decreases with height in the environment is greater than the rate at which the parcel is cooling, the parcel will be warmer and less dense than the environment and will continue to rise. When the air near the surface is very warm and moist, there is a lot of energy stored in the form of latent heat. If a warm, moist parcel ascends in an environment with a steep lapse rate (i.e., one in which temperature decreases rapidly with height), the parcel will quickly become much warmer than the environment as condensation occurs and latent heat is released. The parcel will then be much more buoyant than its surroundings and will ascend even faster. This is what happens in thunderstorms - warm, moist air rises, cooling more slowly than the surrounding air. As it cools, more and more latent heat is released, making the parcel warmer and warmer with respect to its surroundings. You can see this by looking at the shape of thunderstorm (cumulonimbus) clouds - notice how tall they are. This indicates that there is strong vertical motion in the cloud.

### IV. Lab Instructions

#### A. Latitudinal temperature structure and baroclinic instability

Task1: Use the viewer to examine the zonal mean January temperature. Which is colder in January: the North Pole or the South Pole? Why? Looking at all pressure levels, where is the coldest spot located? Now make a time-latitude plot to display the seasonal cycle of temperature as a function of latitude. Select "Time" from the horizontal axis pop-up menu, and select "Latitude" from the vertical axis pop-up menu; click "Redraw".

• In what season is the equator-pole temperature difference largest/smallest in the Northern Hemisphere?
• Is it largest/smallest in the same season or a different season in the Southern Hemisphere?
• In which season then do you expect the most vigorous synoptic-scale storms, given the nature of baroclinic instability and the Earth's need to transport excess heat poleward to where the largest radiative deficits exist?

Transfer the January temperature and January humidity tables to Excel. Open the January temperature Excel table. Each column of data corresponds to a different latitude band, in 2.5 degree latitude increments, and each row gives the temperature (in degrees Celsius) at a different pressure level. The rows are arranged in order of decreasing pressure (increasing altitude) because that is the sequence in which data are recorded by a rising weather balloon; thus the first row of data at the top of the table corresponds to the Earth's surface. We are used to having the surface at the bottom, so sort the data according to pressure in ascending order.

Task 2: Make charts (line graphs) of temperature vs. latitude at the 200 mb, 500 mb, and 1000 mb pressure levels. When using Excel, always select the "Scatter" plot option rather than the "Line" plot option, even when you are asked to make a line graph. You can easily connect the dots on a scatter plot to make it a line graph. Notice that the latitude values in row 1 contain both letters and numbers (i.e., T(-90)). Excel does not recognize these as numerical values, so you will need to add a row containing only the latitude values. Do this by inserting a blank row above the temperature data and entering "-90" in the cell corresponding to the temperature measurements at -90. Then click on the next cell in the row, type the appropriate formula in the formula bar, and then use autofill to paste the formula into the remaining cells in the row. To make all 3 graphs at the same time on one chart, select the row you created which contains the latitude numbers for the x-axis, and then, while holding down the Control key, select each of the rows for the three pressure levels, before clicking on the Chart Wizard. Based on your graph, answer the following questions:

• At the Earth's surface, is the meridional temperature gradient larger in magnitude in the Northern or Southern Hemisphere?
• At approximately which latitude is this temperature gradient largest?
• Does the magnitude of the temperature gradient increase or decrease as one moves from the surface to the tropopause?
• Considering that baroclinic instability needs strong meridional temperature gradients to enable storms to grow, are synoptic weather systems (storms) more likely to be found in the lower/middle, or the upper, troposphere? Excluding thunderstorms (a different type of storm formed by strong atmospheric convection - see also B.2), how high in general should airplanes fly in order to get above the weather and have a smooth flight?

#### B: Vertical temperature and moisture structure and convective stability

Task 3: Make new line graphs showing temperature vs. pressure at three latitudes: One near the equator, a second in northern midlatitudes, and a third at the North Pole. (As in the previous step, put all three graphs on a single chart to make them easy to compare; but this time you have to select columns rather than rows.) On these graphs, altitude increases toward the left, i.e., Earth's surface is on the right and the lower stratosphere is on the left. Locate the tropopause (defined as the pressure level of minimum temperature) near the equator and in the midlatitudes. Focus now on the vertical temperature structure near the surface. At which latitude is near-surface air warmest relative to air at slightly higher altitudes? What part of the world should therefore be most prone to thunderstorms? Do you notice anything qualitatively different about the near-surface vertical temperature structure at the North Pole compared to the other two latitudes?

Task 4: Now make a chart in which you plot temperature vs. pressure at two locations (again on the same chart for ease of comparison): One at a Northern Hemisphere midlatitude location, and the other at the identical latitude but in the Southern Hemisphere. Comparing the vertical temperature structure in the two hemispheres, in which season are thunderstorms most likely to occur?

Task 5: Select and copy the equatorial temperature profile. Now open the January humidity Excel table; this shows water vapor mixing ratios (in grams of vapor per kilogram of dry air) in the same format as for the temperature table. Sort the data into ascending order as you did for temperature. Paste the equatorial temperature profile into the humidity table and make a scatter plot of temperature vs. humidity at the equator. Be sure that both the temperature and humidity profiles are sorted in ascending order; both variables should decrease with height.

• Is moisture positively-correlated, negatively-correlated or not-correlated with temperature?
• Is the behavior linear, or does moisture change with temperature faster at either warm or cold temperature (i.e., for a given difference in temperature, is the difference in humidity larger at the warm end or the cold end of the diagram)?
• Considering that warm, moist air near the ground is conducive to thunderstorms, does the correlation of water vapor concentration with temperature help or hinder the formation of thunderstorms?

#### C. The zonal mean general circulation

Task 6: Return to the viewer and display the January Zonal wind field. At what pressure level and latitude do the maximum mean zonal winds (the jet streams) occur? Does the latitude of the jet stream correspond more closely to that of the maximum or minimum temperature, or the latitude of maximum or minimum latitudinal temperature gradient? Consider the tropics (equator to 30 degrees latitude). What is the direction of the zonal wind near the surface? Zonal winds are positive if they move west-to-east (also called "westerlies" because the wind is coming from the west) and negative if they move east-to-west ("easterlies"). Similarly, meridional winds are positive for south-to-north motion and negative when moving north-to-south.

Task 7: To make the zonal wind direction easier to see, display the data as a black-and-white contour plot; click and hold on the pop-up menu currently displaying "colors"; select "contours"; click "Redraw". The thick zero contour separates positive (eastward) zonal winds on one side of the contour line from negative (westward) zonal winds on the other side. Poleward of about 10 degrees latitude, what is the direction of the zonal wind near the tropical tropopause? The zonal wind in the tropics is produced primarily by the Coriolis force acting on air parcels moving toward higher or lower latitudes. Remember that the Coriolis force deflects parcels to the right in the Northern Hemisphere, and to the left in the Southern Hemisphere. For air moving poleward, in what direction does the Coriolis force deflect it? What about for air moving equatorward? Based on this, infer the meridional wind direction in the upper and lower tropical troposphere from the observed zonal wind direction at these altitudes. In other words, in what direction would the wind have been moving in order for the Coriolis force to have produced the observed zonal winds? Try using a globe to help you visualize the movement of the winds.

Task 8: Now return to "colors," choose "Time" and "Latitude" for a time-latitude plot, and click "Redraw." Now type 200 (ie. 200mb) in the pressure selection box and click "Redraw" again. This will display the seasonal cycle of the zonal wind at the tropopause in color. In what season do the strongest jet stream winds occur? Given that weather systems are carried by the jet stream, do you expect that weather changes faster in general in winter or summer? Is the strength of the jet stream correlated primarily with maximum/minimum temperature or temperature gradient?

#### D. The meridional mean general circulation

Task 9: Use the viewer to display the January Meridional wind field. Are mean meridional (south-north) winds weaker, stronger, or about the same strength as mean zonal (west-east) winds? Compare the meridional and zonal wind speed at 20N at a pressure level of 200 mb. Which component is larger?

Task 10: The height gradient at the 200 mb level is largely oriented in the meridional (south-north) direction. You can convince yourself by looking at a map which shows the height of the 200 mb pressure surface in geopotential meter. When the contours are close together that means a large pressure gradient. Remember that the gradient is at a right angle to the contours and thus most everywhere in meridional. Which two places depart mostly from the meridional direction?

### VI. Lab Report Instructions

Write a lab report (as per the Lab Report Format) summarizing the major findings of your investigation. Incorporate your answers to the following questions into your lab report:

1. Are summer-winter temperature differences (an example of a natural, rather than an anthropogenic, climate change) generally larger in the Northern or Southern Hemisphere? What is the major geographic difference between the two hemispheres, and how might that explain this result? Now consider current attempts to detect the signature of global warming due to increasing greenhouse gas concentrations; the greater (and faster) the warming, the easier it is to detect. Should we expect to see the clearest evidence of global warming in the Midwestern U.S. and central Asia, in coastal locations such as New York City, or out over the open ocean? Explain.
2. Consider the qualitative difference between the lower troposphere lapse rate in January at the North Pole vs. midlatitudes and the tropics. What major thing (aside from the extreme cold) differentiates the North Pole in January from lower latitudes? How might this cause the observed difference in lapse rate? (Hint: Why is the surface in general the warmest part of the atmosphere? What would happen to the surface if the source of this warmth were removed?)
3. Consider the locations of sign changes of the mean meridional wind with latitude near the surface (in both hemispheres), and the sign of the tropopause meridional wind in the Northern Hemisphere tropics. Make a schematic pressure- latitude cross section of the Northern Hemisphere tropics and midlatitudes, indicating by arrows the direction of the meridional wind in the different locations. At what latitudes does the surface meridional wind converge? Diverge? What about the tropopause meridional wind? What must be the sign of the vertical velocity between the surface and tropopause at these latitudes? Indicate this by arrows as well. Following the arrows, will a parcel of air originating near the equatorial surface eventually return to its point of origin? Considering the effect of rising and sinking motion on the saturation of a moist air parcel, which latitudes do you expect to be cloudy and rainy? Clearer and drier? How does this correspond to the appearance of the satellite images and precipitation maps we saw in lecture?

• Peixoto, J.P., and A.H. Oort (1991). Physics of Climate. American Institute of Physics, New York.
• Ahrens, C. D., Meteorology Today. Brooks/Cole
• Legates, D.R., and C.J. Willmott (1990). Mean seasonal and spatial variability in gauge-corrected, global precipitation. International Journal of Climatology, Vol. 10, pp. 111-127.

### Contributors:

• Audrey Wolf, GISS
• William Kovari, GISS

Updated July 9, 2007
To report problems, email webmaster.