CS8 Fog Clearance Case Study - 6 February 2020

Figure 1: ASXX for 06 Z on Thursday 6 February 2020

Fog clearance, like formation, is incredibly challenging to forecast accurately, and because of the huge impact that fog has on airfield operations it has huge significance and there often will be a large amount of pressure riding on getting it right. The first question many forecasters get on a foggy morning is “when’s it going to clear then?” usually followed by another pilot saying “10 Z clearance?”

The Frank Barrow technique (F(t) = m if m ≥ 6; F = (12-m) if m < 6 where m = month and F(t) is fog clearance time in UTC/Z ) works well for a typical fog, but if you have a much more extensive or deeper fog then the clearance will be that much later.

On the morning of 6 February 2020 a large anticyclone was located over the Low countries (see Figure 1, right), extending north-westwards over the UK. Under the ridge axis across much of southern England and Wales areas of freezing fog had formed under radiation conditions overnight. The freezing fog extended in a narrow corridor from Northolt in the northeast, southwest-wards through to the Salisbury Plain, with dendritic patterns representing its penetration into the valleys forged in the Chiltern Hills (see Figure 2, below).

By dawn the fog had matured, expanded and deepened, and this was sampled well by the 0700 Z Larkhill ascent (Figure 3, below right) which showed the fog top to be at around 610 FT AMSL, or approximately 25 hPa in depth. This was also verified from the observation at High Wycombe (approximately 670 FT AMSL) being above the foggy air and the surface inversion.

Knowing the height of the fog top is incredibly useful for the forecaster as it can help with applying empirical techniques to determine a fog clearance temperature and time, however often there will not be a representative ascent and so an approximation must be made - this can be simply by looking at the visible satellite and approximating the elevation of the high ground not in fog, or by using a very rough approximation of 5hPa-10hPa for sky clear fogs and >20hPa for sky obscured.

MSG fog
Figure 2: High resolution MSG FOG imagery valid 0330 Z on 6 February 2020

In this case the fog clearance temperature can be approximated by taking a saturated adiabat (SALR) curve to the QNH, which gives T (fog clearance temperature) as 4°C. It is important to remember that in the depths of winter the fog can continue to deepen for several hours after sunrise, and there are cases (such as from December 2012) [ref: 2015] where a mature fog can continue to expand throughout the day.

There are many different processes that can help to dissipate fog, but the most straightforward and common one is that of insolation, and that is the process considered here. A useful technique to determine clearance of fog by this process is the Kennington-Barthram nomogram, which uses a few easily derived pieces of information to determine the fog clearance time. These include:

  • Fog depth (the most difficult to determine)
  • Fog clearance temperature (T2)
  • Dawn temperature (T1)

The Kennington-Barthram technique
In this example the dawn temperature T was approximately –2°C (see Figure 4), we have also already determined that the fog top was at approximately 610 FT AMSL (~25 hPa) and that the fog clearance temperature is approximately 4°C.

We can now use the nomogram (see Figure 5) quite easily to determine a clearance time. Start by finding a point 25 hPa up the horizontal y-axis on the left, and then move across until you reach the point T – T , which represents how much the air must be warmed by to clear the fog. Then from this point follow the curved lines down to the next diagram and move horizontally until you reach the point T +T , then follow the curves down again and finally move horizontally until you reach the curve that best represents the time of year. Note that in this case I have used both the 15 February and 15 January lines as the date is between the two, but nearer to the 15 February. This gave a clearance time of between 1130 and 1400 Z.

The UKV forecast of visibility (see Figure 6) was suggesting that the fog may have cleared a little earlier and certainly widely by 14 Z, and therefore may have been a little fast to clear the fog when compared with the empirical technique.

Looking at the actual observations (see Figure 7) and the satellite imagery from the day (see Figure 8), this suggested that the fog did indeed clear between 1130 and 1400 Z, though in places it lingered even beyond this – so an important caveat to this technique is to bear in mind your station’s position relative to the edge of the fog – as whilst you may clear your own fog, some fog from elsewhere could always advect in, or some upslope effect could occur, all acting to delay the clearance slightly.

This technique doesn’t always work, but it does allow the forecaster to quickly determine a useful rough approximation of the fog clearance time (within a 2-hour BECMG group range).

Larkhill ascent.
Figure 3: Larkhill 0700 Z radiosonde ascent representative of the boundary layer

Figure 4: METARs and SYNOPs from across central southern England at 0600 Z on 6 Feb 2020

Figure 5: Worked example of Kennington-Barthram fog clearance technique

Figure 6: (a) UKV forecast visibility T+5 frame valid 11 Z on 6 February 2020 (top) and (b) UKV forecast visibility T+8 frame valid 14 Z on 6 February 2020 (bottom)

METARs Boscombe
Figure 7: METARs from Boscombe Down (top left), Northolt (top right) and Brize Norton (bottom left) from 6 February 2020

Vis Sat METARs
Figure 8: Visible satellite and METARs 0900 Z on 6 February 2020

Barthram, J. A. A diagram to assess the time of fog clearance. Met. Mag. 93, 1964, 51–56

Price, J., Porson, A. & Lock, A. An Observational Case Study of Persistent Fog and Comparison with an Ensemble Forecast Model. Boundary-Layer Meteorol 155, 301–327 (2015)