58 AIR MASS ANALYSIS 
of Brunt (Phys. Dyn. Met., p. 45), a lapse 
rate of 11.3°/gkm between surface and 1 gkm 
would be enough to cause (absolute) instabil- 
ity in the layer. Even up to 1.5 gkm condi- 
tions of lapse rate favorable for instability 
exist on this day, the actual rate being 
10.7°/%km, whereas the critical (neutral) rate 
is 10.3°/gkm. In any case, the layers up to 
1 km are definitely unstable. These conditions 
would provide a “trigger’’ of sufficient magni- 
tude to initiate displacement of the lower 
layers. 
It is seen that a particle from A starting 
on account of instability in that layer will 
gain energy in the initial portion of the 
ascent up to B (794 mb). The gain is 
AFBMA. After B, the ascent up to D would 
involve a loss of energy BCDEB. It is only 
if the particle can reach D (500 mb), that it 
will come into the environment of latent in- 
stability and begin to gain energy; and it 
ean reach D only if BCDEB is less than 
AFBMA. But actual measurement of these areas 
shows that BCDEB is slightly greater than 
AFBMA, being roughly in the ratio 14:11. 
This is what one would expect, if one re- 
members that the meteorograph was sent up 
only 5 minutes before the occurrence of the 
third squall, there being a comparative lull 
between the second and third squalls. During 
five minutes the meteorograph could have 
ascended only about 1 km. So, it is only up 
to about a km above ground that the tephi- 
gram can be considered to be representative 
of the conditions before the squall; above it, 
the tephigram shows conditions during the 
occurrence of the duststorm. As the tempera- 
tures rise in the upper levels after a dust- 
storm, the area BCDEB is perhaps greater 
than it would be before the occurrence of the 
phenomenon; for the same reason the tem- 
peratures along BF may have been lower 
before the squall, making the area AFB 
slightly greater. 
The possibility of descending cold air from 
a dust- or thunderstorm which oceurred at 
neighboring stations having travelled to Agra 
and lifted up the “latently unstable’? mass of 
air is not entirely ruled out, but in the ab- 
sence of positive evidence to this effect and 
in view of the fact that the superadiabatic 
lapse rate existing is itself sufficient to cause 
the phenomenon, it seems reasonable to attri- 
bute the duststorm on this day to the trigger 
action of insolation leading to a superadiabatic 
lapse rate. The latent instability in this case 
is apparently built up by the moisture brought 
in by the passage of the low-pressure wave 
referred to above in the description of the 
synoptic situation of the day. 
The fact that dust or thunderstorms do not 
occur more frequently than they do, although 
the vertical temperature gradients in the lower 
levels may be adiabatic or superadiabatic day 
after day in the summer months over Agra is 
probably due to the want of the necessary 
condition, viz., latent instability. 
In this sounding all particles up to about 
460 mb are latently unstable, an extreme case; 
the level of greatest latent instability is at 
about 880 mb, but since the convection is 
obviously taking place from the heated sur- 
face, the proper analysis is to consider the 
energy areas for a rising surface particle. 
The decrease of wet bulb below 900 mb indi- 
eates that a surface particle has considerably 
less latent instability than a particle say at 
880 mb whose condensation level would be at 
750 mb giving a smaller negative area and 
far larger positive area than for the surface 
particle. The same would be true for particles 
up to around 620 mb. Mixing in the lowest 
300 mb layers will increase the surface wet- 
bulb, however, and tend to distribute the 
latent instability more equally among the dif- 
ferent levels, in these layers. If this sounding 
were an early morning one, the latter consi- 
deration would certainly deserve some weight 
in estimating the realizable latent instability 
for later in the day. 
In many cases, however, the wet bulbs are 
much lower, relative to the dry bulb, at the 
upper levels than in this sounding, with only 
a shallow moist layer (but not necessarily 
a temperature-inversion) near the surface 
which will often appear to have considerable 
real latent instability. But here it is well to 
allow for the effect of convection in lowering 
the surface wet bulb by mixing with the 
drier layers aloft, for in this way the amount 
of latent instability for the surface particles 
can be rapidly decreased before the convection 
reaches the condensation level. The method 
of assuming an average wet bulb for the 
lowest layers, in the same sense as Mr. 
Namias suggests using an average specific 
humidity, is often called for in such situa- 
tions. On the other hand, evaporation often 
keeps the surface wet bulb from falling or 
even increases it in spite of mixing, and then 
the depth of the moist layer is increased at 
the expense of the drier upper air—which 
means the latent instability is increasing. 
Likewise advection of moist or dry currents 
aloft will alter the state, as explained in the 
article on Isentropic Analysis. Isentropic 
cross-sections and charts are not conveniently 
compared with estegrams since the former 
contain only A and q (or condensation pres- 
sures). 
