Uplift of low-level air in deep convection

1
Uplift of low-level air in deep convection

• Geraint Vaughan

2
Example of CO profile
Convective outflow plumes

Mean CO out of cloud above 10 km
High variability at low levels (biomass burning)
Median CO 1-3 km from Dornier

3
Intercomparison flights
Dornier minus Egrett CO, ppbv
25
20
AD12 AE10
15
10
Difference, ppbv
5
0
AD28 AE30
-5
12.6
13.0
13.4
12.2
Time, h
4
Biomass burning period (1) and pre-monsoon(3) CO
decreases strongly with height
5
Sharp fall in upper-level CO during December
6
Monsoon period (4) no vertical gradient strong
vertical mixing
7
Monsoon break (5) reversed vertical gradient
8
Can we explain background changes by long-range
transport?

• 3-D kinematic back-trajectories calculated from
  each flight section out of cloud above 200 mb
• Most trajectories had ascended from low level in
  the past 10 days
• Point where trajectory ascended through 500 mb
  surface plotted together with measured background
  CO from Egrett

9
Sharp decrease in December due to change in air
mass origin
10
Flights through anvil outflow
November 16 (left hand panel) CO enhanced in
anvil outflow February 6 (right hand panel) CO
reduced in anvil outflow
Green/Yellow denotes cloud
11
Where does the anvil air come from?
Anvil peaks in CO above 10 km all less than 110
ppbv
Boundary-layer outflow at 8 km?
Boundary-layer air all gt 100 ppbv
12
Same for AE25
Anvil peaks in CO above 10 km around 55 ppbv
Boundary-layer air gt 60 ppbv
13
Definition Anvil enhancement
Inferred out-of cloud CO concentration, by linear
interpolation of measured background values
A
A Enhancement of CO in the anvil (negative for
this flight)
14
Definition Vertical Difference

• Calculate average CO in 2 km blocks on Egrett
  descent profile, from 0 8 km
• Calculate difference between these averages and
  the mean background above 10 km
• Call this the vertical difference

Anvil Enhancement

Vertical difference for height 2 km
Mean background gt10 km
Use Egrett descent profile rather than ascent
because of problem with Dornier intercomparison.

Mean from 1-3 km
15
Method

• Calculate anvil enhancement for each flight
• Calculate vertical difference for seven vertical
  intervals 0-2, 1-3, 6-8 km, for each flight
• Plot against each other
• Calculate correlation coefficient for each set of
  vertical differences

16
Mean anvil enhancement vs. vertical difference,
all flights
Mean anvil enhancements ltlt vertical differences
suggests mixing
1 1 line
Line of symbols represents one flight
17
Correlation coefficients
Correlation coefficients increase slowly to a
peak at 4-6 km then decrease sharply
18
Maximum anvil enhancement
Maximum anvil enhancements similar in magnitude
to vertical differences Max correlation again
around 5 km
19
Conclusions

• Trajectory calculations consistent with
  background upper-level CO measurements impact of
  local convection small except in monsoon
• Maximum anvil enhancement suggests that uplift
  from mid-levels dominates anvil composition
  during pre-monsoon and monsoon break