6.3                                   Long waves  

6.3.1                                Some general concepts

6.3.1.1                          Scale

The space scale associated with "long waves" is typically considered to be in the order of 10,000 km per wave, or, typically between one (360º wave–length) and five (~70º wave–length) waves around the Earth. The longevity and speed of movement of these features is typically in the order of days to weeks. And in the case of some very persistent features, the effective life–span of the feature may persist for weeks to months. While outside the scope of this handbook it is noted that seasonal weather forecasting in the Antarctic may have a basis in the characteristics of Southern Hemisphere long waves (Ryzhakov, 1983; Ryzhakov et al., 1990; Ryabkov, 1999).

6.3.1.2                          Zonal versus meridional flow

In the case where the long wave pattern has waves of small amplitude, the individual waves tend to move eastwards with relatively high speed through the pattern that is said to be of high zonal index. (Commonly, such an index might be based on the geopotential height difference at the 500–hPa level between two latitude bands). In the case of "low zonal index", the waves tend to have much larger amplitude (that is, are greater in meridional extent) and move more slowly, may be stationary or may event retrogress (move to the west).

In very pronounced meridional situations the westerly flow may become split such that one branch of the westerly flow moves to the north, and the other branch to the south, of a "blocking pair" of a low–high–pressure system combination. Often such blocks will persists for many days if not weeks.

An excellent summary of the relationship between cyclone–frontal systems and the large–scale flow is given in Schultz et al. (1998, p. 1,768).

6.3.1.3                          Identifying and using the long wave pattern

With improvements in the global numerical models one might argue that the imperative that the long wave pattern be monitored is not quite as pressing as it used to be. Occasionally this might be the case, however, there are many instances when the various global models offer differing solutions to the extended period forecast. And so the forecaster might give extra credence to the model that offers a prognosis that is consistent with long wave theory and experience. Typically the long wave pattern has its practical use in helping to assess if the short wave features (highs/lows/fronts) will strengthen; decay; become blocked; or be steered in a particular direction, and so on.

The conventional tools used in monitoring the long wave pattern include daily and 5–day mean 500–hPa contour charts; filtered 500–hPa charts in which typically waves less than about 70º wavelength are filtered out to reduce the effect of the short wave features; Hövmoller charts (for example, see Figure 6.3.1.3.1); and looped satellite imagery. In this last case, particularly where the loops span several days, the forecaster can gain an appreciation of any systematic movement of short wave features around the broader scale steering flow. Moreover, loops of water–vapour imagery can be very revealing in identifying persistent areas of descent/ascent or of jet–stream location.

6.3.2                                Some specific examples of long waves in the Antarctic context

It was mentioned above that systems tend to be progressive if the long wave pattern is of a high zonal index; in other words, the waves are of small amplitude and mean wind speeds are relatively high. On the other hand, if the long wave pattern is meridional then systems tend to be slow in movement or even retrogress. And in the case of blocking, low–pressure systems approaching the high–pressure part of the block can be steered southeastwards due to the upper split flow.

In the examples (strong wind events that occurred during the early summer of 1983‑84 between Mawson and Casey Stations) presented below it was found that when major depressions approached the East Antarctic coast, they moved closer to a station when there was long–wave ridging to the east of the station. Conversely, the lows remained well offshore when there was a long wave trough east of the station and a long wave ridge to the west. Moreover, stations lying to the west of a blocking ridge experienced prolonged periods of cloud bands crossing the coast with gale–force winds. Ridging lying to the south over the continent advected these cloud bands to the west and back towards the upstream long–wave trough.

6.3.2.1                          A prolonged period of bad weather at Casey

Once the blocking ridge became established east of Casey around 10 December 1983 (Figure 6.3.1.3.1) flying conditions in the area were poor. There were five events between 17 December 1983 and 12 January 1984 that produced gales at Casey Station. In each case the synoptic scale lows slowed down and weakened as they approached the long wave ridge (block) to the east. Their associated cloud bands were then steered towards the coast.

As an example of these events the period 16 to 19 December 1983 is shown as Figure 6.3.2.1.1. In this example the depth of the approaching trough (Figure 6.3.1.3.1) was sufficient to erode the downstream block and so the short wave 500–hPa low was able to move east of Casey but the block quickly re–established itself. This example, although relevant to the Casey area, is also typical of blizzard development in the Mawson and Davis areas where ridges downstream of the stations steer lows close by.

6.3.2.2                          A major low passes harmlessly to the north

In situations where the long–wave trough is east of a station with a ridge to the west, or where the tropospheric flow is basically zonal, the major storms tend to stay away from the station. As an example an event is examined that occurred over 3 to 5 January 1984 when the long wave trough lay to the east of Mawson and Davis Stations (Figure 6.3.1.3.1). A major developing cloud vortex appeared to be moving towards Mawson and Davis on 3 January 1984 (Figure 6.3.2.2.1). However it kept on an eastward track and passed harmlessly well north of both stations during 4 and 5 January 1984.

The 500–hPa pattern (Figure 6.3.1.3.1) shows only weak ridging associated with this event, and significantly, westerly 500–hPa flow to the east of Davis during 3 and 4 January 1984, as the cloud vortex developed.

6.3.2.3                          Patterns at 500 hPa associated with winds in East Antarctica

Phillpot (1997, p. 93–163) discusses the use of 500–hPa charts in forecasting surface winds for many stations between the Molodezhnaya area and Leningradskya. In particular Phillpot presents 500–hPa composite height contour fields associated with surface winds reaching at least storm force in East Antarctica (Figures 4.23 and 4.24(d) on pages 93 and 98 of his study). Common to most situations presented by Phillpot are 500–hPa lows northwest of the station with coastal downstream ridging. This pattern is quite similar to that which is described in the specific examples above. In the absence of numerical guidance, the pattern recognition approach to be inferred from Phillpot's work provides very useful guidance.

6.3.2.5                          Implications for planning operations

Baba (1993) reports optimism in being able to make predictions of the general trend of the weather a few weeks ahead through the prediction of the long wave pattern. It is inferred from Baba that these types of predictions can influence operational planning when he reports that, during twelve seasons of ocean bottom surveys, winds of force 4 or less occurred on about 80% of occasions in good years and only about 38% of days in bad years. Baba's results support the view in Section 6.3.2.1 that a high frequency of bad weather may occur when a long wave trough is located just west of an area, and by implication there would be a long wave ridge just east of the area.