APPENDIX 3 – "REPRESENTATIVE" ANTARCTIC ATMOSPHERES

On Standard Atmospheres

A "standard" atmosphere serves as a reference for the relation between pressure, height, and temperature in the vertical. There have been several such atmospheres adopted in the past, for example: the USA National Advisory Committee for Aeronautics (NACA) Atmosphere; the International Commission for Air Navigation Atmosphere; and the International Civil Aviation Authority (ICAO) standard atmosphere. Each of these is similar in many respects. (Saucier, 1955, p. 53). The ICAO standard atmosphere is now generally in use and its chief specifications are:

·                         MSLP: 1013.25 hPa;

·                         Mean sea level temperature: 15°C;

·                         Temperature (virtual) lapse rate 6.5°C up to 11 km (where T = –56.5°C);

·                         Isothermal lower stratosphere above 11 km (to at least 20 km).

NACA (1954, p. 2) indicates that the ICAO standard atmosphere is in "fairly good agreement" with the annual average values of pressure, temperature and density for the North American atmosphere near 40º N up to an altitude of about 20,000 m (~65,000 ft). Being a hypothetical approximation to the "real" atmosphere found in mid–latitudes of the Northern Hemisphere, the ICAO (or any of the others mentioned above) is likely to be a poor representation of the Antarctic atmosphere. The practical implication of this is that while the ICAO atmosphere is used, among other purposes, for the calibration of altimeters allowance needs to be made for the cold Antarctic environment. As mentioned in Section 3.4.2.7 to adjust for the real atmosphere's departure from 15 °C the following equation may be used:

Z  =  Z_ind(1 + (Ts – 15)/288)                                                                                          Equation A3–1

It may be seen that when the surface air temperature (Ts) is below 15°C the altimeter readings (Z_ind) will be higher than the estimated "true" height Z. For example, if the surface air temperature is –20°C the indicated height (Z_ind) has to be multiplied by 0.88 to give a more accurate (lower) estimate of altitude.

Suggested "Representative" Antarctic Atmospheres

It is evident from the discussion above that ideally a "standard atmosphere" might usefully be derived for the Antarctic. While a rigorous derivation of such a model atmosphere is beyond the scope of this handbook the following describes related work steered by the editors but with I. Barnes–Keoghan and D. Shepherd of the Australian Bureau of Meteorology and S. Colwell and S. Harangozo of the British Antarctic Survey performing most of the data manipulation and processing and data display work. The aim of this work is to suggest "Representative Antarctic Atmospheres" (RAAs) that might more closely reflect the vertical atmospheric profile of high southern latitudes than does the ICAO standard atmosphere. As will be seen there appears to be merit in defining at least two "RAAs": one for summer and one for winter, and, RAAs for the various latitude–based geographical regions of the Antarctic and sub–Antarctic.

Data

In deriving the ideas for the suggested RAAs the mean January and July vertical temperature traces (along with extreme daily and standard deviations of daily temperature values) were constructed for:

·                         Mount Pleasant Airport, (Tables A3–1 (a) and (b) and Figures A3–1 (a) and (b));

·                         Bellingshausen Station, (Tables A3–2 (a) and (b) and Figures A3–2 (a) and (b));

·                         Halley Station, (Tables A3–3 (a) and (b) and Figures A3–3 (a) and (b));

·                         Amundsen–Scott (South Pole) Station, (Tables A3–4 (a) and (b) and Figures A3–4 (a) and (b));

·                         Vostok Station, (Tables A3–5 (a) and (b) and Figures A3–5 (a) and (b));

·                         Casey Station, (Tables A3–6 (a) and (b) and Figures A3–6 (a) and (b));

·                         Macquarie Island Station, (Tables A3–7 (a) and (b) and Figures A3–7 (a) and (b)).

The data used to compile these figures and tables were from available standard pressure level data from all radiosonde flights during the period 1980 to 1999 inclusive. The surface data, however, were from other available long–term means and are shown in Tables A3–8 (a) and (b). The seven stations used were chosen because not only do they provide good sectional data across the Antarctic ((Mount Pleasant Airport and Bellingshausen are almost on the 60°W longitude, and Vostok/Casey are approximately on the 110° E longitude, and being south of 75º S, Halley is physically not far from this transect) but because they are believed to represent various regions of the Antarctic. Mount Pleasant Airport would be representative of the sub–Antarctic climate with Macquarie Island also being used to average out any local bias, due, for example to the proximity of the South American orography. Bellingshausen is representative of the Antarctic maritime climate being at the tip of the Antarctic Peninsula; South Pole Station and Vostok represent continental stations, while Halley and Casey are considered to be typical of coastal Antarctic stations more exposed to low level easterlies than the more frequent westerlies that, for example, Bellingshausen experiences.

Rationale behind suggested separate summer and winter and latitudinal RAAs

The mean January and July temperature profiles for each of the above stations are displayed in Figure A3–8 (a) and (b) respectively. Given that the data used to compile these figures, apart from the surface data, were from all available flights during the period 1980 to 1999 inclusive, it is considered valid to compare the traces one with the other. Features to note are:

·                         (a) the trend below about 250 hPa, summer and winter, for the temperature to be colder as one moves further south;

·                         (b) this trend reverses in summer above about 250 hPa, at which level all the traces and the ICAO standard atmosphere apparently have about the same temperature;

·                         (c) the traces all depart further (become colder) from the ICAO standard in July;

·                         (d) the Mount Pleasant Airport and Macquarie Island traces are, apart from the lower levels in summer, almost co–incident with each other;

·                         (e) the South Pole and Vostok traces are almost co–incident with each other in both January and July (to the extent that both are sampling the same pressure levels;

·                         (f) Casey and Halley have similar profiles in January (apart from the lowest levels) although in July these profiles separate a little;

·                         (g) below 250 hPa the trace for Bellingshausen remains intermediate between the sub–Antarctic traces (Mount Pleasant Airport and Macquarie Island) and the coastal stations of Halley and Casey;

·                         (h) the traces do not adequately show the radiation induced surface inversions as only standard pressure level and surface data were used.

There may well be many reasons for the above characteristics however, prima facie, it seems reasonable to assert that:

·                         Most of the observations ((a), to (g)) are a consequence of the seasonality that comes from the Antarctic being mostly shaded from the sun in winter. Point (b), for example, probably reflects that in summer, the lower level Antarctic environment causes tropopause heights to be lower the further one goes south, whereas in winter, being in darkness, the whole atmosphere above the Antarctic cools, with the destruction of ozone is a major factor in the cooling aloft.

·                         There is sufficient shift (cooling) between January and July for there to be separate RAAs for each of these months.

·                         With Casey and Halley, being almost ten degrees of latitude apart in location relative to the South Pole, the separation of their profiles in July is probably due to the effective increased "continentality" of Halley as the sea –ice zone increases, and Halley being more completely in the Earth's shadow.

·                         However, given the closeness of the profiles for these two stations in summer it seems pragmatic to average their profiles even in the July case;

·                         A sensible characterisation of the high southern latitude atmosphere might be to have January and July atmospheres constructed according to Table A3–9

Mean upper winds

Upper wind conditions, while generally not considered to be part of the specifications of a representative atmosphere are, however, useful in depicting the average state of an atmosphere. Figures A3–9 (a) and A3–9 (b) are, respectively, wind roses for January and July for the above stations for which the mean temperature profiles have been calculated. The periods of record for each wind rose were:

·                         Mount Pleasant Airport                       1988 to 1999

·                         Bellingshausen Station                        1980 to 1998

·                         Halley Station                                     1957 to 1999

·                         Amundsen–Scott Station                    1980 to 1999

·                         Vostok Station                                                1980 to 1991

·                         Casey Station                                      1959 to 1998

·                         Macquarie Island Station                    1994 to 1999

Features such as the polar vortex in winter and the variable nature of the upper winds over the pole are quite evident.

Table A3–9     A suggested geographical distribution for the characterisation of the

vertical structure of the high Southern Hemisphere atmosphere.

An application of a Representative Antarctic Atmosphere

Quite apart from altimeter settings another practical application to which a "standard" atmosphere may be put is the "correlation" between height (Z) and pressure in order to gain an appreciation of the pressure altitude at which certain cloud types might be usually present. In practice, in the absence of other objective guidance (for example, NWP) a forecaster might construct a "pressure height line" on the latest radiosonde trace in order to ascertain the geopotential heights of clouds or moisture band/temperature inversions that may be present to assist with a "nowcast". According to the WMO International Cloud Atlas cloud ètages (ranges of levels at which clouds of certain genera occur most frequently) vary from the tropics (highest upper limits) to the polar regions (lowest upper limits) according to Table A3–10.

Using the Halley July data (Figure A3–3(b) and Table A3–3(b) and Table A3–8(b)) (perhaps as a proxy for a "Representative Coastal Antarctic Atmosphere" for the purposes of this discussion) leads to the pressure boundaries corresponding to the Polar ètages given in Table A3–11. For comparison, the pressure levels for an ICAO Atmosphere are also given assuming the ètages for polar regions and it may be seen that the ICAO values overestimate the pressure both the upper and lower boundaries of the ètage.

Table A3–10     Ètage of low, middle and high cloud for various temperature regimes.

Table A3–11     Pressure limits of Polar Ètages of low, middle and high cloud assuming an ICAO standard atmosphere (right hand column) and a Halley July mean temperature profile (left–hand column).

The authors have not, in fact constructed actual "RAAs" as described above for two reasons: firstly, it may be prudent to process all available radiosonde data for all stations before averages are made; and secondly, given the latitudinal and seasonal effects evident in the profiles presented there may be scope for a model that had latitude and season as input parameters. Obviously, on a day–to–day basis, NWP could provide these types of output, but for the purposes of characterisation of the "average" atmosphere a Representative Antarctic Atmosphere seems to have merit. This is work that might be worth attempting once the SCAR‑sponsored Reference Antarctic Data for Environmental Research (READER) project data set has been compiled (see http://www.antarctica.ac.uk/met/READER/ ).

Table A3–1(a)            Upper–air data for Mount Pleasant Airport (WMO: 88889) for January.

                                     (Based on data for the period 1980–99.)

Table A3–1(b)            Upper–air data for Mount Pleasant Airport (WMO: 88889) for July.

                                     (Based on data for the period 1980–99.)

Table A3–2(a)            Upper–air data for Bellingshausen (WMO: 89050) for January.

                                    (Based on data for the period 1980–99.)

Table A3–2(b)            Upper–air data for Bellingshausen (WMO: 89050) for July.

                                    (Based on data for the period 1980–99.)

Table A3–3(a)            Upper–air data for Halley (WMO: 89022) for January.

(Based on data for the period 1980–99.)

Table A3–3(b)            Upper–air data for Halley (WMO: 89022) for July.

                                    (Based on data for the period 1980–99.)

Table A3–4(a)            Upper–air data for Amundsen–Scott (South Pole) (WMO: 89009) for January.

                                    (Based on data for the period 1980–99.)

Table A3–4(b)            Upper–air data for Amundsen–Scott (South Pole) (WMO: 89009) for July.

                                    (Based on data for the period 1980–99.)

Table A3–5(a)            Upper–air data for Vostok (WMO: 89606) for January.

                                    (Based on data for the period 1980–99.)

Table A3–5(b)            Upper–air data for Vostok (WMO: 89606) for July.

                                    (Based on data for the period 1980–99.)

Table A3–6(a)            Upper–air data for Casey (WMO: 89611) for January.

                                    (Based on data for the period 1980–99.)

Table A3–6(b)            Upper–air data for Casey (WMO: 89611) for July.

                                    (Based on data for the period 1980–99.)

Table A3–7(a)            Upper–air data for Macquarie Island (WMO: 94998) for January.

                                    (Based on data for the period 1980–99.)

Table A3–7(b)            Upper–air data for Macquarie Island (WMO: 94998) for July.

                                    (Based on data for the period 1980–99.)

Table A3–8(a)            Surface data for January for the stations shown.

(Based on data for the period 1980–99.)

(* indicates that for these coastal stations sea level pressure values are given whereas for those elevated stations (marked **) station level pressure is given.)

Table A3–8(b)            Surface data for July for the stations shown.

(Based on data for the period 1980–99.)

(* indicates that for these coastal stations sea level pressure values are given whereas for those elevated stations (marked **) station level pressure is given.)