Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
Community
Ask the community for help and clear up your study doubts
Discover the best universities in your country according to Docsity users
Free resources
Download our free guides on studying techniques, anxiety management strategies, and thesis advice from Docsity tutors
The coarse morphology of meteorological circulation governing the ingress of tropospheric air to the stratosphere in the tropics. It covers the egress of air in mid-latitudes via tropopause folding during upper frontogenesis, and the annual average of air exchange for one hemisphere of the stratosphere. The document also explores the use of radiosonde temperature and outgoing longwave radiation measurements to measure high cloud amount, and the occurrence of tropopause deformation as evidence for rising motion and potential air exchange.
What you will learn
Typology: Lecture notes
1 / 92
This page cannot be seen from the preview
Don't miss anything!
TlYL o n
Panel Members
A.F. Tuck, Chairman
- CHAPTER tropopause, suggest that for some purposes it may be necessary to draw a distinction between this transi- tion region and the middle and upper stratosphere above it.
Indeed, if one assumes a mixing ratio of 2.5 and 6.5 ppmv at the hygropause and tropopause, respec- tively, between 30 °N and 30 °S, and of 4 and 25 ppmv poleward of these latitudes, it is a simple matter to show that something like a quarter to a half of all stratospheric water vapour is between the tropopause and the hygropause. Of course, in the mean this fraction probably has a much lower residence time than the remaining amount above it.
It has become apparent during the last decade that the transport of mass and tracers by the general circulation is conceptually simplified by using entropy (potential temperature) as the vertical coordinate. This point of view, originated by Shaw (1942), has been revived by Dutton (1976) and Johnson (1980) for the troposphere, by Tung (1982) for the stratosphere and consistently advocated by Danielsen (1961, 1968; Danielsen and Hipskind, 1980) for studies at tropopause level. The isentropic perspective suggests that past estimates of global cross-tropopause mass flux, made by zonal mean Eulerian streamline calcula- tions relative to pressure surfaces, may not be particularly reliable.
It is necessary to obtain good data on the covariance of mass and the mixing ratio of species whose transport across the tropopause is of interest. This is so because while potential vorticity P0 may in future be calculable from the global analyses produced by high resolution primitive equation numerical weather prediction models, it will still be necessary to know the correlation between P0 and the various chemical species in order to compute fluxes; such knowledge is derived currently from a small number of high quality case studies using aircraft and balloons. At present, global estimates of downward cross-tropopause fluxes rely on a very crude count of upper tropospheric cyclogenetic events to give the case study data a global dimension. The estimates of the upward flux in the tropics are even cruder, since the knowledge of the detailed physical characteristics and scale of the meteorological processes responsible is less secure, although substantial progress has been made recently. It remains true that almost all local, high quality knowledge of cross-tropopause flux is confined to the Northern Hemisphere, in the North American and British Isles/Western European sectors. Most of these data, moreover, have been obtained in the March- May period.
In this chapter, the tropopause is defined both statistically and in a local, synoptic sense by the value P0 = 1.6 × 10 -5 K m2kg-'s -', taken from an objective analysis of 8 years of zonal, temporal mean cross- sections of potential temperature, wind and potential vorticity by Danielsen (1984), see Figure 5-1. The definition applies from the pole to within 5 ° latitude of the equator, where P0 changes sign, and is coinci- dent with the conventionally defined tropopause. The analyses are consistent with those obtained in the FGGE year by the ECMWF analyses, see Figure 5-2. A difficulty of isentropic coordinates for work on longer time scales is that the motion of 0 surfaces with respect to geometric heights has to be established.
5.1.1 Meteorological Processes
There is a wide spectrum of circulation features in the tropical troposphere which involve vertical motions and which may be of importance in the transfer of air from the troposphere to the stratosphere and therefore in the ozone budget. The long-term mean motions are dominated by the planetary scale Hadley and Walker circulations, which are essentially statistical entities; temporal variability occurs on a wide
(km) 30-
25
mb
- ,
a)
,5O 7O 100
20O 30O
- - i 000 80 70 60 50 40 30 20 10 0 LATITUDE
(kin) 30 mb 20
-
b)
I- (^15)
10-
o 90 80 70 60 50 40 30 20 10 LATITUDE
- - -
c)
(kml 30.
25
20-
15-
10-
5"
0
mb
-1 n (^). v --_ .-- **=** _ 10 '----'- =1"
- ", ;-^ : -5-^500 f------^70
,:,,:., ,i ,} "i 200
700
90 80 70 60 50 40 30 20 10 0 LATITUDE
Figure 5-1. Zonal-annual mean distributions of: (a) ozone mixing ratio, ppmv; (b) potential vorticity, 10 -4 cm 2 s -1 K g-; (c) potential temperature, K, and westerly wind velocity, m s -"
variety of time and space scales down to the scale of the individual cumulonimbus clouds embedded in the larger circulation features. The annual cycles of the Hadley circulation and to a lesser extent the Walker circulation are evident in the north-south and east-west excursions of the tropical convergence zones which accompany the Asian and Australian monsoons as well as the cycles of rainy and dry weather elsewhere in the tropics.
Over much of the tropics, the nonseasonal variability of the Hadley and Walker circulations is primarily associated with the Southern Oscillation which is an interannual phenomenon although it is becoming ap- parent that there are prominent circulation changes at the subseasonal time scale of 40-50 days as well (Madden and Julian, 1971, 1972a; Anderson and Rosen, 1983; Lorenc, 1984). Synoptic scale disturbances in the tropical convergence zones include easterly waves and monsoon depressions which modulate the large-scale environment for the development of tropical storms and hurricanes and mesoscale disturbances such as squall lines and mesoscale cloud clusters. There are also important mesoscale circulations tied to localized interactions between the diurnal variation of solar heating and surface features; these include, for example, cloud clusters in the winter monsoon region and sea breeze circulations.
The long-term mean global scale flow in the upper tropical troposphere is well represented by the winds at the 150 mb level. Figure 5-3 shows the horizontal winds for January-February and June-August for the FGGE year 1979. The cross-equatorial flow in the western Pacific and the Indonesian region dur- ing northern winter and in the eastern Indian ocean in northern summer are the most important local con- tributions to the zonally symmetric meridional overturning known as the Hadley circulation. The zonal mean pattern of rising motion shifts from south of the equator in January-February to north of the equator in June-August.
Vertical motion is an extremely difficult quantity to estimate. At the synoptic scale it can be estimated kinematically from the divergence of the observed wind field; operational analyses now make use of reports from commercial airliners as well as cloud motion vectors to augment the conventional upper air network. Satellite measurements of tropical outgoing longwave radiation have also proved useful in identifying the spatial and temporal variability of the occurrence of the cold cloud tops associated with deep convection.
The zonal variability of rising motion in the tropics is part and parcel of a set of east-west overturn- ings which are known collectively as the Walker circulation. These overturnings are most easily seen in the divergent wind field and it has become customary to display the divergent winds in terms of the veloci- ty potential from which it is derived. Figure 5-4 displays the fields of 150 mb velocity potential X for the northern winter and summer of 1979 derived from ECMWF analyses of the FGGE data set. Negative values of X may loosely be associated, in the large scale temporal mean, with rising motion, but should not be identified with vertical velocity. Relative minima in these maps correspond to regions of widespread rising motion; in January the negative center in the Indonesian and west Pacific region is associated with winter monsoon convection and in July the center over southern Asia is evidence of the rising motion due to the summer monsoon. It is worth noting that, as the contribution to the velocity potential from each wave component is inversely proportional to its squared wavenumber, a velocity potential field em- phasizes the larger scale patterns of divergence and attendant rising motion; thus in these maps the Pacific- Indian Ocean Walker cells dominate.
The global wind analyses necessary to derive the transport potential fields implied in Figure 5-4 de- pend to a great extent on high level cloud motion estimates which are nominally applied to a single level only in the upper troposphere; unambiguous identification of penetration of the tropical tropopause in regions
STRAT-TROP EXCHANGE Z O O
A,,| o¢O ,!_ ¢N _ i 4
o_ "!N
I
J /_ I' '
0 ' T-- J :L "
° _ o _., /
, R L _
0
r
° g
(^00)
Z
0
of inferred vertical motion requires higher vertical resolution than is available in the conventional analyses.
Other indicators may suggest the passage of air upwards into the stratosphere. One readily available indicator is radiosonde temperature; cold temperatures are maintained at the tropical tropopause by the adiabatic expansion of air rising from below in deep convection. In the time mean the most vigorous con- vection is in longitudes 80 °E to 180 °E, as shown by the monthly 100 mb temperatures for 1979 in Figure 5-5, which was plotted for this report from radiosonde data. A study of 100 mb monthly mean temperatures by Newell and Gould-Stewart (1981) showed that these longitudes in November-February were cold enough to account for the dryness of the stratosphere, although this leaves open the question of the fate of the necessary ice crystals; in July the coldest areas are associated with the Indian monsoon although the 100 mb temperatures in this season are somewhat warmer than those in January. Temperatures at 100 mb below -82.4 °C are low enough that the saturation moisture content is 2 x 10 6 mmr such as is observed in the stratosphere (see Chapter 9). Frederick and Douglass (1983) and Atticks and Robinson (1983) have come to the same general conclusion from studies of daily radiosonde data but find a considerably larger areal extent of the region of potential exchange.
As a means of measuring high cloud amount, outgoing longwave radiation measurements suffer from contamination with radiation from lower levels in the atmosphere; Barton (1983) has estimated high cloud frequency using two channels from the NIMBUS 5 radiometer data which are preferentially sensitive to high clouds. His results from the period December 1972 through February 1975 are shown in Figure 5-6. The main tropical regions where high clouds occur are the monsoon areas of the west Pacific and India with secondary regions over South America and Africa. From the three Januaries sampled, Barton found that the cloud was less confined to the west Pacific during the E1 Nino January of 1973. A similar El Nino dependence was found for rainfall (Rao, 1984), velocity potential (Climate Analysis Center, 1983,
Finally, another indicator of large-scale motion that influences the stratosphere is total ozone; this will clearly be lower where ozone-poor tropospheric air enters the lower stratosphere. The ozone data from the Nimbus 7 TOMS for the FGGE year 1979, January and July are shown in Figure 5-7. Lowest values occur in January in the west Pacific and South A;nerica with generally higher values occurring everywhere in the tropics in July. Ghazi (1980) has presented a series of total ozone maps that show a minimum in January in the west Pacific and over India in July, essentially in accordance with the findings of Newell and Gould-Stewart, and consistent with the annual cycles of tropopause temperature and total ozone at Gan (1 °S, 73 °E), Figure 5-8.
The network of upper air and surface stations in the tropics at best provides a grid of data that resolves motions on scales of several hundreds of kilometers; this data source as well as evidence from rainfall, cloud and total ozone measurements can be used to identify regions in which rising motion is occurring but they fall short of defining the scales on which the vertical motions are organized. Nevertheless, the contoured radiosonde data show coherent temperature structure on large scales at 100 mb (Figure 5-5). At the level of the tropical tropopause in particular a number of mechanisms for an exchange of air across the statistical boundary between the stratosphere and troposphere have been proposed to account for the dehydration which must occur as air becomes dry enough to be considered 'stratospheric'. Is penetrative vertical motion organized at the scale of individual overshooting cumulonimbus turrets? Or are there mesoscale regions of moderate ascent, driven by cloud heating from below and cooling at the top? Or is there gentle rising motion over a large area? With these questions in mind, and the fact that many scales
a)
25"N^ o"^ 20", o'r 60"^ ' "^ 00"^ _ .... "^40 ^ '^0 Bo'w 40", 20"l. 00" .....^ o"^ 60"^ 40"^ O'W^ 0" 25+_ 5"N • • • 5"N WS 2, (^) .'--..• • " (^) ;-" • (^) • • " • •--<,5.o TT _ WS APRIL 1979 25N 0 20E 4(] 60" 80 100 120 140 100E 180 160W _40 120 100 80 EO' 40" 20"W 005.N
150 KN^50 •^ •-80"^ '^ "^ : -R_C^ •^ IS"N _5!o 25, (^) MAY 1979 25" 25 NF--_O" 20[ 40 00 80 100 _1_8 140 1_0 ( _eOr IBOW 140 120 00 80 60' 40" 20"W • • (^) • 0'25.N 15"N 5"NF/:77"C 5, _ **"'.** (^) R0" "" '. ./ " 5"N
25s[ .. **-------..** -.,.,....._.... JUNE 1979 .... ,5,25"
b)
25"N O"^ 20"E i^ 40" i^ 60" i^ 80"^ OQ"._^ 20" .....^ 40"^ GO'^ 80'^ O0"W^ 40"^^20 100"^ 80" ....^ 60"^ 40"^ 20"W^ O'25"N 15"N 5"N _-7TC ".. 15"N 5"N 5"5 5"S JANUARY 1979 o 2oE 40 0o" 0o" 0 o" 14o 6o" I o sow 40" 20" Io0 0o' 00" 40" 2o'w 0"
,50 p .^. :NO^ • C^. **-.**^80 C^ .•^ ,^ 5.N,/, 25'S " (^) FEBRUARY 1979 25"s 0" 20"E 40" 60" 80 100" 120 40' 60E q80 100"W 140 20 Io0' co' 00 40 20"w 0" 25N I _ _ '-'--""''-'" _ • , _ 125 TM 10"N S'N • -77" (^) :: • (^). •. **-ROC "•-,.Bn,'-.•."ao"** • (^) " 15"N5"N 15"'^ 5'0^ -83"^ -83^ "^ ?"^ K"^ '^^1 15"S 25.5_-- • -,,_ _ (^7) MARCH 1979 • ".-- • 125"
c)
25"N o"^ 20"E^ 40'^ 00"^ 80'^ - _00^^120 140 160E^100 160"w^140 120 100 00"^ 00'^ 40'^ 20"W^ 0+25.N
JULY 1979 25.N 0 "^ 20"E^40 60"^80 100 120 140"^ 160[^^180^100 W^140 120 I00^ BO^ 00"^ 40'^ 20"W^ 005" N"
t ..... 1
5N 5"0 (^) " " " (^) -7TC• ° • -ll C (^) t" 5"N5"S
AUGUST 1979 25'N {}"^ 20E^ 4O^00 80 ,..^ zOO^ • 120 ....^^140 150[^180 IUOW^140 120 100 80^60 40"^ 20"w|
_I(_I .... " .' (^) SEPTEMBER.- 1979 "" • 25"
d)
D" 20'E 40" 00" 80" 10(]" 120 140 ]66E 180 100w 140 120 00" 00 0 40" 20"W 0 25"N[, 77 "'''. rrl -. .... ., , ., • , , , 00"N
,50^ 5"0^ .-,,..--,...._."^ "^ _^ °^ •^ "^. "°J^ 720,15"S 75"' OCTOBER 1979 05"N 0"^ 20"E^ 4O"^ 00", 00" •.^^100 ,^ 2O^ 4^ **100**^ eo^ ...... 0o^ w^40 20 oo.^ 0o ¸.... 6o ¸^ o^ 2ow^ o25"N
ROVEM8ER 1979 o 0E 4(] 00 00 100 120 140 i_0 100 100w 140 120 100 80 00 ¸ 40 ¸ 20"W 0 ¸. 5'N -R0 -c;--- 80 • KN 25"S (^) OEUEMBER 1979 25"S
Figure 5-5. 100 mb tropical monthly mean temperatures for the FGGE year (1979), radiosonde data• (a) JFM, (b) AM J, (c) JAS, (d) OND. Dots are radiosonde stations• Note especially the fluctuation of the area enclosed by the -80°C contour over the year•
a)
b)
JANUARY 1979
400
(^25 ) 250 225
300 _ aoo I I I I I I l I I K I I JULY 1979
I I i i I i i _ I I I I I I I i
Figure 5-7. Monthly mean 03 column observations, FGGE year, 1979. Data are from the Nimbus 7 TOMS instrument; (a) January, (b) July.
a)
871^86
- -
TROPOPAUSE 83 TEMPERATURE °C
- **-
-** - -
GAN (1 °S, 73°E) 1964-
I I I I I x I I I I I I
MONTHLY MEANS
x
X X
\
274
272
270
268 X---K 266 OZONE COLUMN D.U. 264
262
260
258
256
254
252
-75 250 J F M A M J J A S O N D
b)
-9O
- - - - - - TROP TEMP - oC (^) **-
-**
GAN (1 °S, 73°E) 1959-75 MONTHLY MEANS
J I I L I I L L 63 64 65 66 67 68 69 70 71
I l L I L L J I 1 1959 69 61 62 72 73 74 75
Figure 5-8. (a) Annual cycle of tropopause temperature and ozone column density at Gan (1 °S, 73 °E). Data are monthly means, 1964-73. (b) Time series of monthly mean tropopause temperature, Gan, 1959-75. Note that there is a reproducible warming in February/March between cold peaks in December/January and April/May (on average). August is usually the warmest month.
Mesoscale cloud clusters have been observed in the tropics which seem to form independently of synoptic scale disturbances; over the South China Sea during Winter MONEX disturbances formed offshore in the early hours of each day apparently in response to convergence tied to a combination of local topographic features and the diurnal cycle of solar heating (Houze, et al., 1981). Johnson and Kriete (1982) have shown that forced mesoscale vertical motions are associated with these systems. Similar mesoscale convective zones were observed during the wet season over Panama (Danielsen, 1982). These clusters tend to form extensive anvils and stratus decks at middle levels that may spread over several hundred kilometers. Radiative calculations by Webster and Stephens (1980), Doherty et al. (1984) and others indicate that these exten- sive cloud layers should experience IR warming from below and cooling above to space; the net heating could drive a number of processes including mesoscale ascent and turbulent overturning. These processes are enhanced by the release of latent heat associated with the precipitation observed from these cloud decks which often show a radar bright band similar to those seen in middle latitude stratiform precipitation (see Houze and Betts, 1981).
Radiative destabilization may operate also in other extensive high clouds; in fact although individual cumulonimbus clouds reach up to the tropopause many of the anvils are found well below the tropopause. The origin of some common cirrus clouds near the tropopause is not known; examples have been given by Platt (1983) for the region of northern Australia. Potential sources are synoptic scale motions or some type of radiative instability, both of which may be triggered by cumulonimbus activity.
5.1.2 Cumulonimbus Clouds
It is appropriate to discuss the individual cumulonimbus, a scale on which there seems to be a good potential for troposphere-stratosphere exchange as proposed by Danielsen (1982). This section is designed to set the stage for the case studies of Section 5.1.3. The mechanisms involved in cumulonimbus convec- tion are discussed extensively in the textbooks of Riehl (1979) and Ludlam (1980). From the points of view of the present report it is desirable to know if this process can carry air into the stratosphere and if so, where and when. There is no question that large masses of air from the boundary layer are carried aloft to the upper troposphere by the cumulonimbus process. Although vertical motions in individual cells may range up to 30 m/sec, overall it takes an hour or more for significant quantities of air to be transferred to the upper troposphere. Whether the physical processes in the anvil then lead to penetration, as suggested by Danielsen, (1982) and reviewed by Holton (1984b) is an unsolved problem which is soon to be exa- mined experimentally in NASA's STEP project. For the present discussion it is assumed that there is as- sociation between penetration and anvils. Frequency of occurrence of cumulonimbus is shown in Figure 5-9 for two seasons and may be compared with that of cirrus from the same marine data set down in Figure 5-10. The two are clearly related, especially in the west Pacific and the Indian Ocean. Cumulonimbus distribution thus matches the large-scale motion distribution discussed earlier. The release of latent heat that maintains the tropical circulation occurs primarily within these phenomena, termed by Malkus (1968), the "firebox of the circulation". If it turns out that experiments verify Danielsen's hypothesis, the cor- respondence between the most extensive cumulonimbus activity and the greatest large-scale vertical mo- tion may allow stratosphere-troposphere exchange in the tropics to be monitored by means of the large- scale observables as we have discussed above.
We now consider some results of a case study which was designed to see how actual observations of large cumulonimbus tops compared with established climatologies. Airborne radar and cameras com- bined with horizon gyroscopes were used by Cornford and Spavins (1973) to study cumulonimbus tops during the April-June period in NE India. They concluded that tops extended to at least 20 km, and noted that Burnham (1970) had established that turbulence may extend into the clear air for 25-30 km around
ORIGINAL PAGE fS OF POOR QUALITY (^) STRAT-TROPEXCHANGE
FREQUENCY OF OCCURRENCE (PERCENT) OF CUMULONIMBUS DJF (1965-1976) OCEAN AREAS ONLY 30"[ 60E go£ 120£ 150"E 180" 150°w* 120"w go°w 60w 30°w 0 ° 30"[ go" N- go N**
60" N 60" N
30" $ 30" S
90* $ g0" $ 30"£ 60"[ 90"[ 120"E 150£ 180 150"W 120"W S0W 60W 30"W 0" 30"[**
JJA (1965 1976) FREQUENCY^ OF^ OCCURRENCE^ (PERCENT)^ OF^ CUMULONIMBUS OCEAN AREAS ONLY 30" [ 60 ° [ 90" E 120" [ 150 ° [ 180 ° 150" w 120 ° w 90" w 60" W 30" w O" 30" [ gO" N- .90" N
60 N 60 N**
30 N 30" N*
0 ° 0"
30 $* _ 30" S
$0" S 60" S
gO* $. gO" $ 30°E 80"[ 90"£ 120"E 1S0"[ 180" ISOW 120"W tOW** 60W 30"W O _,$OE_*
Figure 5-9. Frequencies of observation of cumulonimbus clouds over the ocean from Hahn, eta/. (1982) for December-February (top) and June-August (bottom).
a visible storm, and up to 11/_-3 km above its top. Since parcel theory was the best statistical predictor of the observed top heights (although it was not reliable on a day to day basis), it is clear that near-surface air may be carried, on a time scale of hours, largely undiluted to up to 20 km, with potential for further mixing above this altitude. Mattingly (1977) noted a statistically significant tendency for the vertical ex- tent of tops penetrating the tropopause in the Cornford and Spavins data to be correlated with the horizon- tal dimension: the bigger storms penetrated further, as shown in Figure 5-11.
It should be noted that upper level windshear frequently plays an important role in cumulonimbus development (Ludlam 1980), and the possibility of horizontal winds at the level of cumulonimbus tops and their associated anvils transporting water vapour and small ice crystals downwind must be considered. Aircraft studies in mid-latitudes (Barrett et al., 1973) have apparently shown that this can occur.
Shipborne radiosonde launches during the winter MONEX experiment have also shown mesoscale temperature structure just above the tropopause near Borneo which was sufficiently cold to be compatible with the low mixing ratio associated with the stratosphere at and above the hygropause (Johnson and Kriete 1982), see Figure 5-12.
The production of nitric oxide during lightning discharges in cumulonimbus storms was calculated theoretically by Tuck (1976), Griffing (1977) and Chameides et al. (1977): a recent review is given by Borucki and Chameides (1984). Measurements of NO2 production by Noxon (1976) confirmed this proc- ess, and he suggested that some of it could enter the stratosphere. Tuck (1976) calculated that there could be a flux of order 1034 NO x molecules per year into the lower tropical stratosphere from cumulonimbi, which is of the same order as the stratospheric photochemical production from O(_D) + N20 -- 2 NO.
x
2.5 2. z xx
xxx XX X I
0 0 0 0.5 1.0 1. LINEAR DIMENSION AT TROPOPAUSE (m) × 10 4
Figure 5-11. A diagram of the height of cumulonimbus tops above thetropopause versus horizontal dimension, from Mattingly (1977) using data from NE India obtained by Cornford and Spavins (1973).
5O
100 A JD
200 _/: 300 ¢#1Mi 400 n 50O 6OO 7O08O (^9001000) ze F 08 LST 14 20 02 16 14 A 12 E v I-- 10 "t"¢3 8
"I" 6 4 2 0
iz2 24 Zi 24 )i_ii )(25 25 ]_ )] 08 14 20 02 08 14 23^ 2Z/_ 24^ 2Z_ 25^ 2Z_ 25^ *23_F^ 2_F 25 20 02 08 100
o_ -1 0 -2 i_" 3 -1 '300 ul ee' 400 (/_ 1 500 _
/ , "°°',oo , _ ; 1000 08L 14 20 02 08 14 20 02 08 14 20 02 08 B 9 10 11
Figure 5-12. (a)Rawinsonde time series during winter MONEX for 9-11 December 1978. Stippling denotes regions of greater than 80% relative humidity. Solid contours are temperature deviations (K) from the 6-28 December mean. Wind speeds are in m/s (full barb - 5 m/s). Dashed line marks the tropopause. Bars at top indicate fraction of ship array covered by bright IR satellite cloudiness. (b) Ver- tical pressure velocity in units of 100 mb/day for the ship array computed by the kinematic method. Distance scale indicated represents advective length scale for 6 m/s motion of the anvil clouds. After Johnson and Kriete, 1982.
However, the degree of overlap between discharge channels and updrafts is unknown, and there is no reliable information on the partitioning of lightning-produced NOx between stratosphere and troposphere.
It is clear from the large scale wind fields in the tropics that any volume of air at or a few km above the tropopause will not be simply transported zonally. Dynamical constraints dictate that there will be meridional components to the flow, consistent with the correlation observed by Reid and Gage (1984) between tropical tropopause height and the angular momentum of the atmosphere. Variations in the inten- sity of the Hadley Cell also show this (Reed and Vlcek, 1969, Newell et al., 1974, Reiter, 1979); Section 5.2.3 also examines the connection between tropical and mid-latitude circulations.
5.1.3 Aircraft Studies near Cumulonimbus Anvils in Panama
In this section, we are concerned with the tracer aspect of water vapour. We show that from measurements of water vapour below and above the tropical tropopause it is possible to demonstrate that