Elevated subsidence inversion

Subsidence inversions can occur in a warm sector between a retreating warm front and an approaching cold front, almost always at higher levels and usually above the freezing level. However although the observed dissipation of the altocumulus layer at 10-12,000 ft may have been caused by adiabatic warming associated with subsidence at this level, Tony Pallot considers that this is unlikely to have a bearing on conditions near the ground. The fact that surface pressure was actually falling at the time ahead of an approaching cold front suggests ascending (cooling) air at low levels rather than descending (warming) air, so a subsidence inversion at low level does not seem likely 1Email from Tony Pallot to Martin Shough, 25.07.07.

Nevertheless we pursued this question further on the strength of an extremely dry level on the noon Brest profile at about 2500ft (see Fig. 21). Might not such a layer of dry air indicate the top of a subsidence inversion? This is often the case; but not, we were advised, in this instance. The overall dynamics of the atmosphere are inconsistent with a subsidence inversion at this low level. Rather this layer is believed to be what is known as a "dry air intrusion caused by continental air being forced north parallel to the frontal zone, possibly from as far south as the Iberian peninsula 2 It is also possible that dry air is created ahead of a katabatic front, when the front is moving slower than the air ahead of it so that air tends to descend down the frontal surface. But this usually occurs from high altitudes down to about 3000m, and although the mechanism could be responsible for adiabatic evaporation of altocumulus cloud at this level the much lower Brest dry layer seems more likely to be associated with a deliquescent haze layer reported by observers at around 2000 ft and which Capt Bowyer described as "due to bad air from the continent. Tony Pallot would expect the Brest layer to be fairly localised but opines that the intrusion could have extended northward to the Channel Islands area by that afternoon (email to Martin Shough, 16.08.07) .

Fig. 21 Extract from the Brest 1200Z radiosonde ascent profile
1011.0 95 18.2 8.2 52 6.79 220 10 290.4 310.0 291.6
1000.0 188 16.8 6.8 52 6.23 220 11 289.9 307.9 291.0
968.0 463 12.8 2.8 51 4.86 224 13 288.6 302.7 289.5
935.0 754 12.6 -18.4 10 0.96 229 15 291.3 294.4 291.5
925.0 844 12.6 -5.4 28 2.78 230 16 292.2 300.6 292.7