Water Vapour Imagery

(This note wouldn't pass muster as an answer to a question in a degree-atmospheric physics paper, so don't use it a such! It's simply an attempt to put into words for the intelligent layman the basic concepts of some powerful principles used in current operational and research meteorology)

The advent of high-quality water-vapour imagery (WVI), received from geostationary satellites, offer meteorologists a "new" way of visualising developments in the atmosphere as they occur ('real time').

At its most basic, viewing areas of moist (usually presented as white, or very light-grey shades) as distinct from dry (dark/black areas), is obviously useful. [ However, it is important to remember that the sensors are detecting radiance from a broad (altitude) band, with diffuse boundaries between 400 and 200hPa (circa 7 to 11km)]. Now, not only can tangible clouds of water droplets or ice crystals be observed and followed, but also the potentially cloudy areas can be observed - at least in the upper troposphere.

Once frequent images were obtained (at least hourly), then these moist and dry regions could be followed. One use is to monitor the advection (movement) of dry (dark) areas over the top of humid and potentially unstable lower tropospheric air-masses (detected by other traditional synoptic methods). Such situations can lead to "explosive" convective activity.

In the field of aviation meteorology, the sharp discontinuity (light vs. dark/black) between dry & moist regions are associated with jet-stream shear; not only useful to track the position of the jet of course, but also areas of potential clear air turbulence (CAT).

However, allied to theories associated with vorticity-forced cyclonic development, looped WVI really comes into its own.

In adiabatic, frictionless flow, potential vorticity (PV) can be defined as the product of two variables:
the [ absolute vorticity ] of the air (on an isentropic* surface) and ...
its [ static stability ];
or more loosely, a measure of the tendency of air to 'twist' as it flows downstream and the degree of 'damping' to vertical motion. The resultant variable is strictly called 'Isentropic Potential Vorticity' or IPV.
(* isentropic: a thermodynamic process not involving change of entropy - crudely equivalent to adiabatic in meaning. )

Wherever these two quantities have a high & positive value, then PV will be high.

So, how does all this fit in with the use of WVI?

Stratospheric air in association with a powerful jet (and less dramatically with short-wave troughs), have high stability (see the definition relating to the stratosphere), and high, positive absolute vorticity - provided the air is being sampled on the cold side of the strongest flow aloft, or in the area in or just ahead of the axis of an upper trough.

Stratospheric air is also dry. WVI detects such dry regions very efficiently. If we 'see' a dark/dry slot in the imagery, then we can infer that stratospheric air has descended to lower levels, dominating the 7 to 11km column that the satellite is monitoring; these must be areas of high (stratospheric) PV - and "PV anomalies" (discrete dark slots associated with marked cyclonic development) can be used to monitor the vigour, location, phasing etc., of developmental regions.

Why should such discrete areas of dry, stratospheric air be so important? One way of trying to understand what it going on is to understand that if stratospheric air has descended, (revealed as the dark slot on WVI), then there must be motions causing this descent, and where air is going sharply down, there must be air going up! The greater the 'vigour' of the air going down (sharply darkening dry slot over a relatively small area), the greater must be the effect of the air going up - leading to thick cloud cover, rapidly falling pressure, and associated developments involved in cyclonic development.

A more rigorous view, at least in terms of IPV development theory, is to regard well-formed upper-level IPV anomalies as the initiator of cyclonic circulations in the column below it. [(This is analogous to the effect you get when you stir the surface of an initially still bath of water - a small circulatory motion at the surface of the bath will eventually translate to motion deeper within the water: thus it is with the atmosphere (only another fluid after all)].
If this 'stirring' at high levels (the PV anomaly) is coincident with a low-level marked baroclinic zone (i.e. a classical frontal boundary), then, other factors being allowed for, the induced circulation will lead to poleward warm-air advection (and thus falling pressure), equator-ward cold-air advection, and atmospheric 'development' (formation or enhancement of low pressure areas) will occur. (On operational charts, the warm front moves north, the cold front sweeps south, (reverse directions if viewing in the Southern Hemisphere) and the wave depression deepens.)

If the PV forcing is marked and is co-incident with sharp baroclinicity (large gradient of temperature with horizontal distance), then 'explosive deepening' of a depression will occur.

Modern NWP diagnostics packages can output 'pseudo' WVI - this is overlaid on 'real' WVI and any mismatch in model analyses readily seen and allowed for. Forecasters have become used to using water vapour imagery to diagnose and monitor development in mid-latitudes - a powerful tool indeed.
If you want to know a little more about water vapour imagery, see here.
And a classic example of WVI associated with a major NE Atlantic autumn storm is shown here.

WVI - Some additional notes


 **  Spacecraft sensors integrate radiant heat energy through the column of the atmosphere within the field of view (FOV), NOT from a fixed level.
 **  WV radiances are detected using Channel 10, which is tuned to 6.7 microns, and allows a resolution of around 5 km (Meteosat & GOES) at the sub-satellite point (SSP)
 **  Radiation "seen" is biased towards nearest WV layer in the FOV, or the underlying (relatively warm) surface, if no cloud or very dry air is present.
 **  Maximum response: 80% of radiation from 620 to 240 mbar, with notional maxima of response for "standard atmosphere" at around 400 mbar. (However, this 'fixed' level is mis-leading: it is best to remember that the radiation is integrated through a layer having diffuse upper & lower bounds.)
 BLACK  : "warm" (low altitude / near-surface radiation source; small amounts of water vapour (in the sub-satellite column) in the FOV)
 DARK GREY  : "cool" (slightly colder than low altitude - typical of AC / thin, low AS clouds, therefore typical of low or mid - troposphere around or just below the level of non-divergence (LND)
 LIGHT GREY  : "cold" ( typical of thicker AS levels, or tops of NS, or thinner (but low) CI/CS OR areas of high humidity [ but no clouds evident ]-- around and just above the LND, and at the theoretical maxima of radiation that the sensor is responsive to.
 NEAR WHITE  : "very cold" ( typical of thick / high CI/CS)
 BRIGHT WHITE  : typical of CB clusters which poke out above the general moist / cloudy levels. (Useful for detection of MCS)
 ** Importance of WV imagery lie not just as an instantaneous image, but with loops over time.
 Blacker-with-time: (implies) >  warming / lowering WV content.
   Either (or combination of):
   (a): Descending air or
   (b): Advection of drier air or
   (c): Clouds being replaced by non-cloudy air
 Whiter-with-time: (implies) >  cooling / increasing WV content.
   Either (or combination of):
   (a): Ascending air or
   (b): Advection of moister air or
   (c): Non-cloudy zones turning cloudy

( : but slight changes between moist / non-cloudy and moist / cloudy zones cannot be inferred from WVI, especially at CI levels.)

 **  WVI patterns take up the character of the flow they are in. In the FOV, a WV signature, especially in developmental situations, will almost certainly not be at one level. It is therefore a "tracer" of horizontal and vertical atmospheric motions. Even at jet-stream altitudes, where we tend to assign a 'fixed' level to the jet core, the jet often wanders up and down through several hundreds of metres, and where marked cross-contour flow is involved, then at the entrances and exits to such jets, changes in altitude of at least 1500 m are not unusual.
 **  There are often abrupt changes between "black" and "light-grey" areas: i.e. as between dry / descending and moist / ascending zones. [ US & Canadian Met. sources point to these regions as most likely ones for Clear Air Turbulence (CAT) - indeed they quote a successful detection rate of some 80 % for CAT in these areas. The sharper the boundary, the more likely is CAT to be found, though no inference (as yet detected) can be made as to the severity of the CAT. ]
 **  Very useful in detecting vorticity patterns - which, with attendant moisture indicators, can give useful clues to development.

   **  Beginning of dry slot ("dry intrusion") cyclonically curved into the vortex centre -- associated centre is "closing / cutting off"
   **  Broadening / blacker dry slot -- indicates development is still in place -- maximum cyclogenesis about to begin. (Intrusion of high tropospheric / lower stratospheric air)
   **  If dry slot is ill-defined / not-warming (darkening) with time -- development of parent system possibly arrested early.
   **  System is weakening when dry / descending air totally encircles the vortex.
[ Sometimes, the dry slot encircles the vortex several times.]
   **  Well defined "hook" on dry imagery: defines classic PVA areas -- indicating strong development.
   **  Short-wave troughs (SWT): active troughs, generating "+SHRA/+RA/TS" rather than just an enhancement of the general shower regime - often marked by a noticeable dry-line (or abrupt change from 'moist' (white) to 'dry' (dark grey/black); found either along or just to the rear of the lower tropospheric trough axis (i.e. ~700hPa).



 standard analysis at 06Z  This is the conventional surface analysis with the familiar fronts, lows etc. After this point, the depression deepened smartly (but NOT rapidly), to a value of 988 hPa (or mbar) in the Oslo area of Norway some 36 hrs later. 'A' indicates the area influenced by the warm-air conveyor ahead of the driving upper trough - showing how warm, humid air is thrown well ahead of the developing centre to produce the 'white' area seen on the WV image: 'B' indicates the cold low-level air cutting in behind the development.
 image of developing dry slot here  A: 'Dry' slot tucking in behind developing depression: darkening of the dry intrusion indicated that development (falling surface pressure) was to be expected.
 B: High water vapour content detected in the 'conveyor' streams associated with the baroclinic development - leading to thick cloud & rain/snow etc.
 C: The Polar Front Jet lay just to the north of the black / white discontinuity along this clear-cut edge ... however note comments below, as the PFJ can only be placed like this in certain circumstances.

   **  Beware placing jetstreams along WVI boundaries (see introductory note). Water vapour patterns trace the level within which the WV maxima lies, which varies with both space [ all three dimensions ] and time -- e.g. the conveyors associated with extra-tropical depressions.
   **  Some general rules:
> a strong / non-buckling jet can be located by the sharp, well-defined poleward cloud edge of an associated cirrus shield (but see comments above re: placement of jet).
> the stronger the jet -- the better the definition (the greater the ageostrophic forces).
> the location of the dark / descending zone relative to the jet depends on the curvature of the flow:
>> CYCLONIC: darkest zone equatorwards of jet core
>> ANTICYCLONIC: dark zone is poleward of jet core.
(NB: however, I have not found this to be a very true statement, and some reservations are held about this.)
> jet streaks can be traced at the head of a dark zone which is known (from independent analysis) to be within the jet region.
   **  For release of Convective instability -- decrease of ThetaW with height -- need dry, mid-level air over-running moist low level air. Use (a): WVI loops to trace mid-level moisture and (b): NWP ThetaW fields to trace low-level moist plumes.
   **  Preferred location for maxima of such development is along the leading edge of a dark zone.
   **  Use loop to infer movement of edges of moist zones -- i.e. does movement agree with appropriate NWP frames at the same time -- if not, adjust NWP derived output as required.

 Some other notes that might be useful .....
 **  Use change of "whiteness" to ascertain ascent / development or descent / decay.
 **  Broad troughs often have a sharp WV (dry/moist) boundary immediately to the rear of the axis - i.e. along the line that marks the abrupt reversal of the sign of vertical motion / change of sign of relative vorticity.
 **  Sub-tropical jets (STJ): usually show a marked edge which enables the translation / shape change to be monitored, thus confirming, or amending NWP ideas.
 **  NWP relative humidity fields can be cross-checked with WVI with high degree of correlation. Any deviations due to the model 'atmosphere' deviating from the "real" atmosphere can be seen and allowed for.
 **  Sharp WVI discontinuity along jets caused by sensing of lower stratospheric air (low WV content) and dry / descending polar maritime air.


WVI example

This led to high rainfall totals and 'damaging' winds in the NE Atlantic/NW European region in late autumn 2002 (The "Prestige" Storm).


Water vapour example & 300 hPa



 A  Dry slot of rapidly descending stratospheric air to the immediate rearward of the synoptic feature. Strongest gusts occurred as the leading edge of this feature encountered landfall over northern Portugal/NW Spain.
 B  Area of broad-scale ascent associated with the warm conveyor of the developing low.
 Other notes:  This magnificent image in the 'water-vapour' channel from the ESA/EuMetSat 'Meteosat' platform neatly captures the major development that occurred through the 13th November 2002, which in the next few hours was to produce high-winds and heavy rain over NW Iberia, with gusts reported to at least 60 knots. One ship in Sea Area 'FitzRoy' reported a mean wind of 65 knots as the leading edge of the 'dry slot' swung east. (Incidentally, this storm was responsible for the severe damage to the Tanker "Prestige", which subsequently broke up on the 19th, with consequent loss of its contents - a major pollution episode ensued.)

During the evening/night to come, the storm swept NE to bring heavy rainfall (with flooding) to many parts of NW Europe (including Britain), along with high winds. Squally winds were reported for the west, and later the northern Departments of France, and in the period 14/0300 to 14/0900 UTC, gusts of at least 60 knots were reported just inland of the central and eastern English Channel, with 'on-coast' gusts to at least 70 knots (possibly more in isolated spots).


 1  Note the implied accumulation of air just below the jet-core as the high-speed air on the NW'ly jet has to slow abruptly as it encounter the weaker & markedly cyclonically curved gradient in the trough axis: the air has to spread out and descend, encouraging the downward penetration of high-PV/stratospheric air. This action brings air having a high velocity to low altitudes, mixing with the already 'perky' low-atmosphere flow.
 2  If the air is going down sharply (1), then it must be going up somewhere, and this region, just forward of the highly diffluent trough and associated PVA-max region, is just one such, where heavy rainfall was reported.
 3  The equatorward 'tail' of the cirrus outflow shield, curling back around the dry intrusion, where in former times a classical occlusion might have been drawn, but is really an upper-air feature, and not necessarily frontal either. Various workers have found that the strongest gusts (possibly tornadic-enhanced) occur near/just in front of the 'tail' of this feature in the dark slot.