Upper Air Meteorology

Some Preliminaries

The mid-latitude westerlies blow in response to the global-scale difference in temperature between colder/high latitude regions and warmer, lower latitude zones. The earth does not receive its heat uniformly over the globe: much more heat is available at equatorial latitudes, than at the poles. Circulations must be in place to distribute the excess on/near the equator to other regions north and south of the tropics: the meandering westerlies, often concentrated into ribbons of fast-moving 'jets' of strong winds play an important part in maintaining the overall global heat balance.

The patterns are not uniform of course, as anyone looking at upper air analyses/forecasts will readily appreciate. Variations occur in response to seasonal changes in hemispheric temperature differences. In winter, with a generally strong north-south temperature gradient, the resultant thermally driven winds are much stronger than in summer, when temperature differences are somewhat less pronounced, due to the summer hemisphere receiving more insolation. The situation is complicated at times in that disturbances running through the long wave pattern in the upper flow can in themselves lead to advection of warm air polewards, and cold air equator-wards, offsetting the temperature differences which produced the broad-scale westerlies in the first place! There are also variations when the conservation of absolute vorticity theorems are applied, which govern the shape/propagation of waves in the upper westerlies. These complex 'feedback loops' are a common feature in meteorology.

The long waves - also known as Rossby waves: What produces these?

A pure west to east flow, perhaps implied by the simple model of warm/low latitudes and cold/high latitudes outlined above, is difficult to sustain in the real atmosphere, with orography, and differential thermal capacities of land versus sea playing a major part in causing perturbations in the broad-scale westerly flow. On a mean (e.g. monthly) analysis, and often on individual days, long wave trough/ridge systems can be found.

On average, there are between 3 and 7 such waves in each hemisphere, and considering just for the moment the Northern Hemisphere, there are two that are regarded as semi-permanent features: one roughly downstream of the North American Rockies, and another downwind of the Himalya range. The position of these features, plus their semi-permanence, suggest that the massive mountain chains involved are the primary reason for their existence. The mechanism is almost certainly a need to conserve absolute vorticity in a column of the atmosphere that is forced to climb abruptly in this fashion.

The number of long waves is largely influenced by the magnitude of the north/south temperature gradient - when the thermal gradient is weak, typically in summer, then the westerlies are able to meander rather more than in winter and early spring, when the flow is stronger. The number and position of the major (Rossby) long waves is very important in the study of hemispheric weather patterns, and even slight changes in amplitude/longitudinal position can have major repercussions thousands of kilometres downwind. 

For more information on the long/Rossby waves, see here.

How and why does the pressure change in the upper air?

Imagine the atmosphere consisting of an infinite number of slices in the vertical. As heat is pumped into each slice, by whatever method (advection from somewhere else; heating from below etc.), that slice expands (gains energy), and as well as expansion sideways, and downwards (against a net expansion upwards by the layer below), there is a general expansion upwards. The definition of atmospheric pressure is the force per unit area due to the weight of the atmosphere above any defined point. Thus, for any level in the atmosphere, if heat is supplied to a column of these infinite number of slices, the whole column expands effectively upwards, and because more of the atmosphere is now above any one point, the pressure at that point increases.(There must be a net expansion upwards, because the earth's surface forms an effective block to net downward expansion.) The opposite occurs for cooling of a column.

Why do meteorological charts show contour heights etc., and not isobars?

In meteorology, we don't take a fixed height, and measure the pressure of that height (though in the early days of upper air meteorology, that was in fact done), but instead define a pressure level ( 850 hPa, 700 hPa, 500 hPa etc.), and find the height (amsl) at which that pressure value is found. There are several reasons why this is so: It was easier when balloon sounding using pressure sensors was the only method of upper air measurement to note the height at which pressure levels were found; and in computation of upper dynamics (e.g. wind vectors), by using this scheme you can eliminate density from the equations ... an obvious variable. There are also advantages when relating the upper air patterns to the readings on a pressure-altimeter.).

On a constant pressure level map then (say 500 hPa), you will have heights plotted as, for example 5340, 5432, 5286, 5567 metres etc., and then contours are drawn (just like isobars, which in fact they are), and so we get to the often smoothly flowing patterns with ridges/troughs etc., familiar to those used to MSLP charts of isobaric plots.


                                Example of 500 hPa product

What is the significance of high versus low contours?

From the first paragraph, in cold air the 'column' of the infinite number of slices collapses downwards, and in warm air, the 'column' expands upwards: So, in summary: a cold troposphere is associated with low contour values. a warm troposphere is associated with high contour values. And further, as a very rough/first approximation: higher contour values are associated with warmer weather; lower contour values with colder weather, or Polar/low contours vs. Sub-Tropical/high contours. (This doesn't always hold true right at the bottom of the atmosphere of course: it is not unusual for cold, low level air to undercut warmer air aloft with high contour values, and vice-versa.)

Why do we get zones of very strong winds aloft/jetstreams?

The gradient of temperature from cold/low contour areas to warm/high contour areas tends not to be gentle in mid-latitudes, but often abrupt for a variety of reasons (continental/ocean thermal contrasts; convergence of airstreams etc.), and temperature contrasts are enhanced from time to time - drawn on our maps as fronts. It is important to remember, that although we show fronts as surface features, they extend right through the Troposphere, as a sharp contrast in temperatures in the horizontal, and this gives rise to the strong winds aloft, as I shall try to explain below.

The effect of these temperature contrasts is cumulative. Therefore at lower levels, say around 850 hPa, the temperature differences across a frontal surface, which then lead to pressure differences (see opening paragraphs), give winds say around 30 knots. However, by the time we get to 400 or 300 hPa, the warmer column of air will have a greater cumulative expansion that the colder column of air, and the consequent pressure difference will be greater, hence the resultant wind will be stronger: typically at these levels 70-110 knots, and just below the Tropopause (top of the troposphere/"weather zone"), maybe up to 150 knots or more.

Only where you get these marked thermal contrasts do you get the potential for strong winds aloft, and so, jetstreams are associated with active frontal systems in the mid-latitudes. We call the jet associated with the Polar Front, the discontinuity between ex-Polar air and ex-SubTropical air, the Polar Front Jet (PFJ)

Below is a simplified diagram of the relationship of the PFJ to a classical/mid-latitude frontal system in the Northern Hemisphere. Note that the 'arrangement' also applies in the Southern Hemisphere, but it would be inverted.

Why does the jet vary in strength/direction etc?

The PFJ lengthens/shortens; weakens/strengthens etc., in response to complex changes in the structure of the related air mass boundaries. Jet strengths are usually strongest across the Atlantic during the autumn (as the first signs of polar outbreaks collide with still warm sub-Tropical air masses to the south), and early spring (as the northern air is still cold after the hemispheric winter, but the tropics/sub-tropics start to warm in response to the increasing insolation further south).

Many irregularities occur along the jet length: There are feedback mechanisms at play, whereby a developing cyclonic system can both temporarily strengthen, and eventually decay the vigour of a jet. In general terms again:

Shorter wave-length trough/ridge systems are also found on a day-to-day basis running mostly through the very long wave pattern: the number and speed of translation governed by synoptic-scale balances in development/baroclinic instabilities: 

What is the significance of the position of the jet stream?

The position of the PFJ governs the overall weather type at all seasons of the year. If it sweeps from SW to NE between Iceland and Scotland, then the BI/NW maritime Europe is generally mild/windy with occasional warm anticyclonic periods; If it comes due west to east around 55N, then a stormy run of weather, with frequent, but well distributed rainfall; if it comes 50 N or below, then generally a cold or cool type, with storms running along the Channel or over northern France etc., with wintry/or cold weather to the north. Further, the wave patterns in the upper westerlies, in very large measure, govern weather events on a variety of time scales ranging from a day to weeks, or perhaps even months.

Why does the jet stream change?

The position/orientation of the jetstream, and its strength/continuity is governed by such things as the distribution of Pacific SST anomalies; Pacific/North American temperature constrasts; injection of energy from active tropical disturbances etc. (These are deeper matters which are inappropriate for this article.) Why the PFJ in some years runs well to the north (mild winters/potentially very warm summers) or well to the south (cold winters/unsettled summers) or becomes blocked is still a subject of debate. My own opinion is that a primary trigger must be Pacific SST anomalies allied to the strength/depth of the Polar vortex, which in turn is coupled to the Stratospheric flow and the switch from summer to winter circulation type. If you talk to someone else, you will get other answers.

Sometimes, upper air patterns change only slowly - whats going on?

At mid to high latitudes in the upper part of the troposphere (above roughly 5 km ), the mean wind flow exhibits a broadly west-to-east motion - this applies in both hemispheres, and is a consequence of the global-scale heating differential between the cold/polar regions and the warm/tropical latitudes, coupled to the deflection of air on a spinning sphere: the earth.

On many occasions, particularly in mid-latitude/temperate zone regions, the flow is directed more or less directly from west to east, crossing few latitude zones within the same longitude range: this is a 'highly zonal' type - any short-wave disturbances embedded in the flow will be carried quickly along and the weather is ever-changing as a succession of frontal systems, interspersed with transient ridge conditions cross any one point.

diagram of a zonal jet arrangement
diagram of a zonal jet arrangement


However, on both average (e.g. monthly) pressure maps and on individual days, long-wave trough/ridge patterns can be found - some having large amplitude, i.e. the airflow meanders a long way north and south around the loops of the pattern, crossing many parallels of latitude in a relatively limited longitudinal range: a 'meridional' type; Usually, some west-to-east progression of the looped pattern can be seen over a 24hr period, and the associated surface weather type changes, albeit more slowly than the zonal type described earlier.

diagram of a meridional pattern


However, if the 'loops' in the pattern become locked in one geographical area, then depending where you are in relation to the upper flow, the associated surface patterns are often little changed from one day to another, and in extreme cases, from one week to another - the pattern is said to be 'blocked'. Some examples of blocked patterns are shown below:

Omega and diffluent blocks

In, and just to the east of a slow-moving trough in the upper flow, the surface weather will tend to be of a low pressure/convective/showery type, and perhaps cool for the time of year (but not necessarily); In, and just to the east of a static ridge in the flow, the surface pressure will tend to be high, with settled conditions lasting until the block is destroyed. This latter case is responsible (over land) for prolonged dry/hot weather in summer, but cold/sometimes grey conditions in winter, and considerable pollution build-up can occur at all seasons due to the stagnation of the lower level air and high air-mass stability encountered.

What's all the jargon associated with upper air features?

The flow in the upper Troposphere, roughly above 500 hPa / 18000 ft / 5.5km is much smoother and behaves like a vast river, not of water, but of air. As with rivers, the flow can take up a variety of patterns; from swift/straight to slow/meander and variations in-between. When it is a swift and straight flow from a largely westerly point to a largely easterly point, this is known as ZONAL flow (because the flow crosses many time/longitude zones in a short time), and if the flow is directly from west to east, this is a HIGHLY ZONAL flow. Disturbances/short wave upper troughs embedded in this flow (and the attendant surface disturbances below 500 hPa), move quickly from west to east. When the flow buckles into large amplitude meanders, the flow is said to be MERIDIONAL (because the net movement is north/south along meridians of longitude over time), and blocked situations are said to be HIGHLY MERIDIONAL: this sort of situation gives us either persistent wet weather (we are locked into the trough portion of the loop), or persistent dry weather (we are locked into the ridge portion of the loop). Large amplitude troughs/ridges forming part of the upper flow can move steadily eastwards (progressive motion), stay sensibly stationary (blocked), or move backwards against the general trend (retrogress), though the latter case is usually accompanied by disruption of a trough/spawning of a high cell, where the general flow gives birth to a closed low/high which then lives a life away from the general upper flow, or alternatively, the flow is diverted around the low/high (bifurcation).

How are the upper features related to surface weather patterns?

Broadly speaking, msl low pressure is found in and forward of a marked upper trough, and msl high pressure is found in and forward of a marked upper ridge. These troughs and ridges need not be large amplitude affairs that sweep over hundreds of kilometres of the globe: ~~ A minor RIDGE in the upper flow over a synoptically 'small' area such as southern Britain, or northwest Germany, can act to dampen any tendency to convection: the ridge will be an area of descent, which will produce subsidence warming/drying of the air, and act against vigorous shower activity. ~~ A minor TROUGH, (or short-wave length trough, 'shortwave' for short!) perhaps only just detectable on a chart, can be the cause of much heartbreak for forecasters! The trough may be a slight hesitation in the upper flow, but combined with adequate moisture, and perhaps enhanced by land heating, can lead to notably more intensive shower activity, and operational forecasters are always on the look-out for such irritations at such levels as 500 hPa or 300 hPa. Operational NWP models don't always capture the vigour of such features correctly - perhaps not sharpening them sufficiently, or slightly mis-locating them, and human-adjustments, coupled with analysis of satellite imagery (particularly water vapour imagery), can be very beneficial. The shape of the trough/ridge is also important:

Troughs can be:

Ridges can be:

Diagram of various trough/ridge types