Weather conditions during the flights

This work is ongoing and multiple flights are not described yet. For instance there are no descriptions of the Lufthansa flights.
Please let me know if you would like me to describe a specific flight in detail!

• 11 December 1999 to Colombo
• 12 December 1999 return flight from Colombo
• 18 January 2000 to Male
• 19 January 2000 return flight from Male
• 8 February 2000 to Male
• 9 February 2000 return flight from Male
• 25 March 2000 to Male
• 26 March 2000 from Male
• 14 April 2000 to Colombo
• 15 April 2000 return flight from Colombo
• 18 May 2000 to Windhoek : PV filaments, a southern hemispheric fold and convective outflow
  The flight data for the ascent are missing.
  The aircraft first flew through PV filaments at 36 N (20h40, PV~2 PVU) and 20 N (22h35, PV < 1 PVU) and at the end of the flight sampled the bottom of a fold at 20 S (4h55, PV ~1.4 PVU) according to the map of PV, the PV X-section and PV versus time. The observed water vapour indicates that the first filament was encountered probably slightly later at about 20h45. The observations show a small minimum water vapour at the end of the flight probably corresponding to the fold, not well reproduced by the model.
  The high cloud map shows high clouds associated with deep convection in the ITCZ in a band between 3 S and 10 N over the Atlantic. Over Africa the deep clouds are less organised, mainly centered over Central Africa. The infrared cloud photograph gives a better indication of the location of convective cells over continental Africa.The convective cloud clusters west of the flight track over west Africa and the Atlantic can be found also in the map of vertical wind at 500 hPa and 250 hPa (yellow). The cloud clusters over Central Africa (e.g. the high/cold ones at 6S-3N, ~16 E and the lower (warmer) ones at 5-12 N, 15 E) are not so evident in the model vertical wind maps. A smaller isolated cloud cluster is visible at about 2N, 12 E quite close to and east of the flight track, on the border between Cameroun and Gabon.The cell at 8N,8E over Nigeria in the model does not seem to correspond to an observed cloud cluster at 0 UTC. The cloud cover interpolated from the model to the aircraft location (as a function of time, X-section) shows partial high clouds between 23h45-0h15 and 1h-1h45 which seem to be outflow from the deep clouds further east (cloud cover does not extend downward). The trajectories ending at 0-3S do indeed show uplifting from below 500 hPa (red colour) over Central Africa east of 20E, 6-11N. Between 0h-0h30 (profiles 90-100, 7-9 N) the cross section of vertical wind shows evidence of the convective cell over Nigeria that cannot be identified in the cloud photograph. The model is thus probably wrong there. Please do note however the observed and simulated maximum in water vapour just before midnight ... which might correspond to past convective influence. At about the same time (23h30) the aircraft went up to a higher flight level, with lower temperature where one would expect lower water vapour mixing ratios. The higher water vapour here thus points to convective outflow. Rapid fluctuations of water vapour or temperature (indicating nearby convection) were not observed during this flight.
  
• 19 May 2000 return flight from Windhoek: the same PV filaments and NH+SH folds as on the flight to Windhoek
  The PV distribution (map, X-section) of the return flight is very similar to that of the previous day. The southern hemisphere fold at 20S has been advected slightly to the east and its bottom sampled at a higher altitude just after ascent (at about 280 hPa) with a smaller maximum PV value of less 1 PVU at 7h (PV versus time) than during the flight to Windhoek. Also the filament at 20 N has been advected further east and a maximum PV value of less than 1 PVU was sampled in it at 12h20. The filament at 37 N is now much more pronounced and can be identified as a tropopause fold extending downto 400 hPa. The aircraft sampled its stratospheric part (PV values up to almost 5 PVU at 14h), with whole air sample 10 being taken in the southernmost part of it.
  The flight data are now complete so that also the vertical structure over central Europe can be analysed: just before and during descent the aircraft once more sampled the lower stratosphere up to PV values of almost 5 PVU at 15h05. The specific humidity (X-section, versus time) shows little variation in neither the southern hemispheric (7h) nor the northern hemispheric fold (14h). A decrease in water vapour was observed as the aircraft entered the lowermost stratopshere when the descent started (15h15).
  
  The water vapour versus time can be used to identify the part of the flight affected by tropical deep convection: the simulated water vapour is enhanced between 8h15 and 11h, while the observed water vapour is so between 8h45 and 10h45. The model simulation thus extends the convectively influenced latitude band too far south. Another latitude band with enhanced humidity was sampled between 12h30 and 13h45 north of the PV filament at 20 N. The cloud cover X-section and cloud cover versus time only shows local high cloud cover greater than 10 % there along the flight track. Some limited high cloud cover was also simulated between 10h (or 8h) and 11 h. This corresponds to the upward (yellow) velocities near profile 90 (WAS sample 6) in the X-section of vertical wind. The map of high cloud cover shows the deep convective cloud band over the Atlantic did not continue over continental Africa and the aircraft flew east of it with winds blowing towards the Atlantic between 8S to 8N (wind map, trajectories). Hence the enhanced moisture is probably not due to these clouds but rather to deep convection that should be located further to the east. The IR cloud composites and the Meteosat visible image show the cloud band at 3-4 N over the Atlantic indeed did not continue over the African content and makes one doubt if there really was deep convective cloud locally at the time of WAS sample 6. Over the continent there are a few, more shallow, convective cloud clusters (warmer tops) present here and there, e.g. south of the Tjaad lake (11N,15E) and over Sudan (north of the equator, east of 25E). Over southern hemispheric tropical Africa hardly any deep convection is found in the IR cloud photograph, except an even more shallow cloud over Congo at about 2 S, 23 E). The clouds over Morocco extending toward the flight track over Algeria are well simulated (these are not part of the ITCZ). Of the back trajectories only a few (ending near 2 N) show uplift from below 500 hPa. The water vapour observed between 8h45 and 10h30 exhibits strong rapid fluctuations that might be helpful to identify possible past convective influences. The observed temperature was much smoother in the tropics.
  
  
• 14 June 2000 to Male
  Yet to be described
  
  
• 15 June 2000 return flight from Male : wavy tropopause affected by deep convection
  The map of PV shows a cut-off low over Italy/Algeria very loosely connected to the main reservoir of high PV over northern Europe by a filamentary structures extending over Greece and Turkey.
  The cross section of PV shows a lot of undulations (waves) in tropopause height. At places where the tropopause is high the relative humidity and cloud cover maximize. It is likely that here the tropopause has been pushed upward by deep convection.
  • As a function of time the aircraft first entered the stratosphere as it made a south to north cross section through the PV trough over north Iran/east Turkey (10h45-12h45, see PV versus time). On both sides of this trough there is evidence of a tropopause fold, most pronounced and with very dry air on the southern side over Iran. In the middle of this trough (at about 11h30) the tropopause is higher and there are deep convective clouds at about 52 E, 32 N according to the high cloud cover map. This cloud is also visible in the Madison IR photograph. The actual location of this cloud was perhaps a little bit more to the north than in the model. If we look at the measured and simulated humidity versus time we see that the pronounced fold in the observations (green line) was between about 10h45 and 11h30 (low humidity between seems to be wider and extending further northward and in time). The weaker northern fold was observed between about 11h55 and 12h30. The cloud corresponds with the small maximum in the green line at about 11h45. The deep cloud is also visible in the cloud cover X-section.
  • Next the aircraft passed into a region with tropopause undulations but not much high clouds (12h - 13h45). The variability in observed and model simulated water vapour versus time do not correspond very well here. If I would have to say something about previous influences by deep convection here, I would locate it where the observed humidity is relatively higher.
  • Between 14h and 15h45 the aircraft was again in deep clouds according to the model (see cloud cover versus time and cloud cover X-section). This was from the Black Sea coast in Rumenia to Austria (high cloud cover map). At about 14h30 the aircraft seems to have been forced to ascend to a higher flight level due to deep convection (one can even see evidence of turbulence in the pressure versus time). The air did not become dryer as the aricraft went up, rather there seems to be a small maximum in observed humidity versus time (green line) as the aircraft did this maneuver. Note that the model shows a minimum in humidity (red line, 14h35) here. The geographic location was probably the Carpate mountains in Rumenia (at about 25 E). The Madison IR cloud photograph shows a deep convective cloud there.
    From the wind versus time (max of about 20 m/s at 14h30) and the wind map I conclude that at this point the aircraft was in the outflow of the jet over Sicily on the eastern side of the cut-off low centred on Sardinia. Sample 11 (trajectories) seems to correspond with this, but sample 10 (trajectories) was also in the outflow of the cut-off low. Trajectories for sample 9 probably passed too far south from the centre of the cut-off low to have been influenced by convection inside it.
    
    The Madison IR cloud photograph also shows some active deep convection over the Alps in Austria.
    
• 11 July 2000 to Male
• 12 July 2000 return flight from Male
• 27 July 2000 to Windhoek: stratospheric and deep convective influences
  This part of the flight where aircraft flight data are available was in the troposphere. The tropopuase was above 150 hPa and sometimes above 100 hPa during most of the flight, so the aircraft remained quite far below it. However the aircraft intercepted stratospherically influenced air twice. First over the Meditterannean at about 300 hPa an upper level filament/trough was sampled, and later during descent a tropopause fold at about 600 hPa in the lower troposphere (see PV X-section versus time). In both cases PV remained below 2 PVU (PV versus time). The filament was rather broad and did not cause a pronounced feature in the observed water vapour (H2O versus time). There is evidence in observed water vapour for the fold over southern Africa: a strong decrease of observed water vapour at flight descent. The stratospheric features are associated with the subtropical jets in the northern and southern hemisphere (eastward wind X-section). The fold during descent is associated with a weak baroclinic zone (the subtropical front, an enhanced vertical gradient in the X-section of eq. potential temperature) at 600 hPa. At 700 hPa and at the surface this front is located more to the north at about 8 S (map of eq.pot. temp. at 700 hPa).
  The X-sections of ECMWF cloud cover , relative humidity and vertical wind indicate very deep convection in the ITCZ between 23h40 and 0h30. Of course, the aircraft will not have flown through this, but rather around it. It does coincide with the maximum observed water vapour, which however does not show such a strongly localized signal, but just the broad water vapour increase typical of the ITCZ. The ITCZ deep convection is visible as a cloud band between 10 N and 2 S in the map of high cloud cover. In reality the deep clouds were not a coherent band but patchy (see IR cloud photograph). The aircraft flew along the west coast of Gabon and Cameroun, whereas the deep convective cloud clusters were mostly concentrated around the Central African Republic, and another more to the west over west-Nigeria. The observed temperature and water vapour do not show the strong fluctuations expected in the vicinity of deep convection. The vertical velocity of ECMWF at 250 hPa indicates partly resolved deep convective cloud clusters between 2 and 11 N over Africa but not at exactly the correct positions. This indicates the limitations of ECMWF in locating individual deep convective cloud clusters. Some of the trajectories ending at aircraft locations between 0 and 10 N show evidence of uplifting from the lower troposphere (red= below 500 hPa, purple below 850 hPa) as they passed through the resolved cloud clusters over Central Africa.
  
• 28 July 2000 return flight from Windhoek: southern hemisphere tropopause folds+ deep convective influences
  The return flight shows pretty much the same PV distribution (map 250 hPa). At the ascent in the southern hemisphere there is again the tropopause fold which the arircraft crossed at 6h20 at about 600 hPa (PV X-section). Note that the PV values (versus time) remained below 1.6 PVU again indicating a mixture of tropospheric and stratospheric air. In the northern hemisphere the aircraft first sampled the upper tropospheric filament over North Africa (13h-14h) and then a pronounced fold with PV up to 6 PVU at 15 h. At 14h the upper tropospheric filament in the vertical X-section looks somewhat like another fold (yellow colour at about 230 hPa near profile 140 in PV X-section).
  ECMWF indicates deep convective influences first between 9h40-10h20, at about 4-7 N, and later from 11h40-12h20, at about 20 N (cloud cover versus time, cloud cover X-section, vertical wind X-section, RH X-section, high cloud map, vertical wind map). In the first southernmost deep convective cloud the interpolated cloud cover even goes up to about 100%, the second one looks more like outflow of convection. The comparison of simulated and observed water vapour versus time shows near the southermost deep cloud cluster coinciding peaks in observed and simulated water vapour. The northernmost deep clouds seem to be too far north in the model. The observed water vapour peak is at about 11h20 (the modelled one at about 12h). Note also the observed temperature fluctuations between 10h50 and 11h20.
  Temperature fluctuations corresponding with the intense southernmost cell were observed briefly at 9h45. The ECMWF model seems to indicate that the aircraft flew through this deep convective cell at 5 N, 10 E. Another cell is present in the model at 8N, 17E. In the IR cloud photograph at 12 h these cells are visible, but it seems likely that the aircraft flew in the more or less cloud-free air between them. The trajectories of WAS5 which show the impressive uplift of air from the lowermost troposphere in the cell at 5N (purple) then probably exagerate the upward transport. The map of all trajectories does not show other uplifting events of this magnitude.
  
  
• 16 August 2000 to Male
• 17 August 2000 return flight from Male
• 17 October 2000 to Male
• 18 October 2000 return flight from Male
• 4 November 2000 to Male
• 5 November 2000 return flight from Male
  
  
• 2 December 2000: some influences from tropical deep convection
  This flight was entirely in the troposphere (X-section of PV, PV versus time).
  The situation in the tropics is similar as on the return flight on 3 December which is more extensively described below. According to the model deep convective clouds were present over the African continent between 10 N and 15 S (map of high cloud cover, X-section of cloud cover). Based on the model cloud cover versus time would say we passed into the area influenced by tropical convection at 17h, at the same time when the aircraft went up to the highest level. The plot of humidity versus time and the X-section of RH and the map of RH show that the upper air was already moist north of this (12N -25 N, from 14h15 onward) but there were no high clouds there, perhaps because the air was descending slowly. Over the northern Sahara the air was very dry. There were also clouds at the beginning of the flight over the Mediterannean.
  At the surface the ITCZ was between 10 N and 25 S according to the map of equivalent potential temperature (see explanatory text for 3 December 2003). However, upward motions related to convection are only found over Africa south of the equator (5-15 S east of the flight track) and thus mostly downstream of the aircraft. This is consistent with the map of convective precipitation. Combining this with the wind leads me to expect convective influences rather between 5 N and 15 N and south of 5 S. The trajectories seem to corroborate this (curved trajectories with red colours ending at 5-15 N and originating near the equator, and a few trajectories ending near 15-20 S).
  
• 3 December 2000: mid-latitude and subtropical fold, tropical deep convection
  Before landing in Germany the aircraft passed through a subtropical and a mid-latitude tropopause fold (PV-map, PV X-section). The wind map shows that the mid-latitude jet stream (from France to Tunesia) merges with a subtropical air stream (from West-Africa to Libia) over Libia and Egypt. The two folds are along these jet streams. The northernmost (mid-latitude) fold is most pronounced and reaches downto about 700 hPa. The aircraft passed through the stratospheric upper part of the mid-latitude fold and the lowermost part of the subtropical fold. Air sample 11 may contain air from the subtropical fold. PV was over 5 PVU in the mid-latitude fold at 16h30. PV went up to slightly more than 1 PVU in the lowermost tip of the subtropical fold at 14h15 (PV versus time). The remainder of the flight was in the troposphere.
  Both folds were on the polar (northern) side of wind maxima (at 13h45 and 15h45), the jet streams mentionned above (wind versus time).
  
  The X-sections versus time of cloud cover and vertical wind show that ECMWF indicates that air sample 7 was near the top of a strong deep convective cloud system at about 0 N, 10 E (see maps of convective and large scale precipitation and of high cloud cover and column integrated water vapour). Samples 3 and 4 were also in a region with deep convection, albeit less intense, according to ECMWF. The cloud cover is consistent with the X-section of RH versus time. The trajectories for samples 3, 4 and 7 (trajectory 6) also show ascent from below 500 hPa (red part). The high cloud cover map shows deep convective cloud systems west of the flight track between the equator and about 14 N, and over central Africa east of the flight track between the equator and 14 S. The quality of the ECMWF model in representing the influence of the deep convection can be judged by comparing measured and modelled specific humidity versus time. The deep convective system at the equator, 10 E (sample 7) seems to be less well represented by the model than the larger area influenced by deep convection in the southern hemisphere (samples 3 and 4).
  The location of the intertropical convergence zone (ITCZ) near the surface can be determined form the map of equivalent potential temperature at 700 hPa. The red shading and strong gradients in eq. pot. temperature delimit the ITCZ ( I chose the shading colours to accentuate this).
  
  I also made vertical X-section versus latitude at 10 E for this flight at the request of Tae Siek Rhee. Note the differences between cloud cover X-sections versus time and versus latitude. The PV is more similar than a quantity like cloud cover for these 2 types of X-section. I think we should not use X-sections versus latitude because 10 E or any other longitude does not exactly coincide with the flight track.The CARIBIC flight tracks are so long that they they would need to be split up in a larger number of straight line segments (X-sections versus lat/lon). For previous aircraft experiments I supported the flight tracks were shorter and it made more sense to approximate them by a line segment in the lon-lat plane.
  
  
• 18 January 2001: trough, subtropical fold
  The flight started in a low (upper level trough), then went through an area with still relatively low tropopause in between highs and lows, then crossed the subtropical fold to remain in the subtropical middle troposphere (PV map, Z250 map, PV X-section) From about 4 PVU at mid-latitudes the PV increased to about 8 PVU in the fold (22h45-23h30) and then dropped to less than 0.1 PVU in the subtropics (PV versus time). The subtropical fold was quite pronounced and descended downto about 500 hPa, see also the specific humidity X-section. Dry air was also present on the equatorward side of the fold according to the model. The measured water vapour mixing ratios (erroneously?) show little indication of the fold.
  There were a few high clouds over Syria (high cloud map, cloud cover X-section, cloud cover versus time) but the aircraft probably flew over or around them. They are hardly recognizable in the cloud IR photograph.
• 19 January 2001
  On 19 January the situation was pretty similar to 18 January (PV map) ..... unfinished

• 31 March 2001: filament, subtropical fold, deep clouds on northeastern side of a cut-off low.
  The aircraft started in a ridge (high), then crossed a deep PV-filament (18-23 E, 21h30-22h15) that connected a Mediterannean low/trough (over south Italy, Tunesia,Greece) to the polar reservoir of high PV (PV map). The map of Z250 shows that the low was more or less a cut-off low. Later on (2h-3h) it passed through the subtropical fold over the Persian Gulf (53E-58 E, 30N-23N) (PV X-section) In the filament the PV attained values of about 6 PVU, and in the fold above 4 PVU (PV versus time). Deep convective clouds associated with the cut-off low were present over Greece and west-Turkey. Another cloud band was over Kurdistan/ west Iran (IR cloud photograph, high cloud cover map). The latter one was narrower in the observations than in the model. Perhaps the aircraft only penetrated the deep clouds over Turkey (23-24 h, cloud cover X-section, cloud cover versus time) and did it fly over the other cloud systems.The most significant differences between observed and modelled temperatures occurred between 2 and 4 h in the subtropical fold and on its southern side. The modelled and observed water vapour mixing ratios clearly show the subtropical fold. In reality the fold may have been slightly more to the south than in the model.
• 1 April 2001: subtropical fold (weaker) and cutt-off low.
  On the return flight the Mediterranean cut-off low has moved slightly north. The aircraft passed through its northern boundary over Hungary, Rumania and the Black Sea (PV map, PV X-section). Here the aircraft was for some time in the stratosphere and PV values exceeded 7 PVU (PV versus time). Over the Gulf of Oman (10h45-12h) it also passed again through the subtropical fold which was less pronounced at this crossing.The PV there only slightly exceeded 3 PVU. What seems to be another elongated (filament-like) part of the subtropical fold was crossed near Teheran (south of the Caspian Sea) with PV values remaining below 2 PVU (13h30-14h). The observed and modelled water vapour mixing ratios clearly show the presence of the main fold between 11 and 13h (i.e. longer than the PV). The relative displacement of the dry air relative to the high PV air can also be seen in the specific humidity X-section.The fold seems to continue quite deep, down to 500 hPa.
  There were (shallow) high clouds on the eastern side of the cut-off low, over the Black Sea, that the aircraft perhaps sampled (14h30-15 h). It flew well over the deep convective clouds in the cut-off low itself (high cloud cover map, cloud cover X-section, cloud cover versus time, IR cloud photograph).
  WAS samples 8, 9, 10 and 11 have been influenced by the cut-off low. WAS sample 5 was probably taken in the subtropical fold, and perhaps also WAS sample 4 which shows descent from above 200 hPa.
  
• 13 May 2001: troughs (lows), deep fold, subtropical filament, subtropical high
  This flight started in a PV ridge (high) over Europe, then it passed through the northern sides of a big trough (8 W-43 W) and a small trough (49-57 W) , then, in a subtropical ridge, a PV filament (60 W-65 W) was crossed (PV map). The flight was mostly in the troposphere, though close to the tropopause in the troughs. Only at the eastern (11h45-12h15, PV > 7 PVU) and western (14h-14h45, PV > 4 PVU) edges of the big trough it passed into the stratosphere.On the eastern side of the big trough this was due to the presence of a strong fold south of Ireland. The aircraft went up from 329 to 287 hPa near the western boundary of this fold. On the western side there was a weak fold only. In the small trough PV attained up to almost 3.5 PVU (15h30-16h). The subtropical filament was crossed around 17h15 (PV X-section, PV versus time). At the northern sides of the PV troughs the aircraft took advantage of westward winds (wind map). The measured minima in water vapour mixing ratio were less pronounced than the modelled ones (H2O versus time) but clearly present although slightly shifted westward (later). High clouds were encountered in the ridge over Europe and deep high clouds were in the big trough just west of the deep fold. Furthermore there were high clouds at the end of the flight in the subtropical high (high cloud map, cloud cover X-section, cloud cover versus time, IR cloud photograph). The dry slot of the big trough, i.e. the deep fold south of Ireland, is visible in both the the modelled high cloud cover and the IR cloud photograph. The specific humidity and relative humidity maps at 250 hPa and the specific humidity X-section also show this feature.
• 14 May 2001: subtropical PV filament (bottom), aged trough (stratosphere), 2 cold frontal cloud bands
  The return flight passed again through the subtropical PV filament (49 W-46W) which now is much closer to the small PV trough. The aircraft now flew parallel to the southern edge of the small PV trough (35-42 W) and passed more or less through the middle of the large PV trough (27 W- 0 W) ( PV map). In the big trough the aircraft was in the stratosphere all the time (3h15 - 5h45) with PV values between 7 and 9.5 PVU. On the southern edge of the small trough (1h30-2h15) PV reached only up to about 2 PVU. In the PV filament (1 h) even less. (PV versus time, PV X-section). Between 3h15 and 4h45 there are relatively large deviations between measured and modelled temperatures (in the western part of the big trough). The only possible explanations I can think of is gravity waves or filamentation (unresolved sub-structure).
  Between 29 and 35 W a cloud band associated with the cold front of the small trough was crossed, and from 15 W the aircraft flew over the clouds associated with the frontal zone around the big trough. Only near the descent it entered these clouds. The dry slot is still clearly visible in the high cloud map, but was north of the aircraft track. There were also some high clouds near the departure at Cuba (cloud cover X-section, cloud cover versus time).The observed high clouds (whitest in the IR cloud photograph) seem consistent with this. Please note that the model cloud top heights are quite uncertain in general.
  
• 19 May 2001: wide cloud band, highs and lows, weak fold.
  This flight started in a cut-off high over Europe, briefly passed into a trough/low I, then again in a high over the Atlantic, then crossed a bigger low II to finish through a subtropical high in Isla Margerita (PV map, PV X-section, PV versus time). In low I (about 10h10-10h50) the aircraft remained very close to the tropopause (PV<3 PVU). In low II (13h15-15h30) PV reached values above 6 PVU. Everywhere else the aircraft stayed in the troposphere. Low II had a not very pronounced fold on its western side. In low II the measured temperature fluctuated significantly (turbulence?). There was a wide deep frontal cloud band (11h30-13h) on the eastern side of low II (high cloud map, cloud cover X-section, cloud cover versus time, IR cloud photograph). In this cloud band the vertical winds at 500 hPa and at 250 hPa were strongly upward locally. There were more patchy clouds around 17 h. Around 11h on the western side of low I there was also a shallow cloud over England according to ECMWF.
• 10 June 2001: deep frontal cloud band, ridge
  This flight started in in the stratosphere (PV > 9 PVU) in an upper level trough, the rest of the flight, west of 3 W, was in an upper level ridge. (PV map, PV X-section, PV versus time). Just before descending air with somewhat elevated PV (> 1 PVU) was sampled. Most of the flight over the Atlantic was outside but parallel to a frontal cloud band located to the north (high cloud map, IR cloud photograph). Between 13h30 and 15h45 the aircraft crossed this deep frontal cloud band (cloud cover versus time, cloud cover X-section). Associated with it there were strong upward winds between 18 and 26W at 500 hPa and 250 hPa (X-section of vertical wind). The trajectories ending there also indicate uplift.
• 11 June 2001: deep frontal cloud band, jet stream, uplifted WAS samples:
  During the return flight the aircraft flew more to the north over the Atlantic along the jet stream (wind map), mostly slightly south of the PV gradient (tropospheric air, see PV map) and on the northern boundary of the cloud band (high cloud cover map, cloud cover X-section, IR cloud photograph). Only between 4h30 and 5h30 the flight was in the stratosphere (PV up to 6 PVU) in the Atlantic trough north of the jet stream(PV X-section, PV versus time). The flight ended in the stratosphere in an upper level trough over Europe (after 6h30). Trajectories for WAS 5, especially WAS 6 and also WAS 7 show evidence of uplifting along the cloud band. The origin of the uplifted trajectories is over Central America.
  
• 24 June 2001: highs and a low, weak folds, weak cloud band.
  This flight started in a ridge (high) over Europe, then crossed a trough (between 17 W and 38 W) and the rest of the flight was in a subtropical high (PV map). There was a weak fold on the western side of the low (12h45) and an even weaker one on the eastern side (11h30, PV X-section). Most of the flight was in the troposphere (PV versus time), except perhaps in the westernmost fold where PV attained about 2.5 PVU. Also in the subtropical high at the end of the flight after 15h45 the aircraft was just below the tropopause. The specific humidity X-section also nicely shows the two folds.
  There was a shallow cloud band with 80 % cover on the eastern side of the low. Within the low, some shallow patchy clouds were also present (perhaps remnants of deep convective clouds). On the western side of the low there were patchy shallow clouds above the aircraft. Near the destination on Cuba the aircraft flew near the base of some patchy shallow clouds (high cloud map, cloud cover X-section, IR cloud photograph, cloud cover versus time). Hence, the aircraft probably flew most of the time in cloud free air, except for the cloud band on the eastern side of the low.
  
• 25 June 2001: jet stream, highs and lows, cloud bands.
  The situation was of course quite similar to the previous flight. The aircraft flew much more to the north to take advantage of the jet stream (PV map, wind map). It now crossed the low/trough between 35 and 10 W, 2h30-4h (PV up to 9 PVU). Also between 65 and 55 W (1h-1h45) it was in the stratosphere (PV> 5 PVU) in another trough (PV X-section, PV versus time). Outside the troughs there were deep cloud bands during much of the flight (cloud cover X-section, high cloud map, IR cloud photograph, cloud cover versus time).
• 8 July 2001: low with fold, subtropical high, filament
  The aircraft started in a trough with a fold on itswestern boundary. From 15 W over the Atlantic it flew in the subtropical highs. Between 40 and 45 W it flew parallel and north of a filament, taking advantage of the maximum in westward wind there(PV map, wind map, PV X-section). Only in the fold at 10h25 the aircraft entered briefly into the stratosphere (PV > 5 PVU). The rest of the flight PV was less than 1 PVU (PV versus time).The flight was relatively cloudfree, only at 40 W (13h15-14h15) as the aircraft reached the filament some deep clouds were encountered (high cloud map, cloud cover X-section, IR cloud photograph). Around 16h30 again some clouds were encountered (cloud cover versus time).
• 9 July 2001: subtropical high, clouds
  The situation is very similar to the one during the flight to Cuba on 8 July 2001. On this return flight the aircraft flew more in the middle and at the northern edge of the subtropical high parallel to the jet stream (where the wind was eastward). Over Europe (east of 7 W, after 4 h15) the aircraft entered the stratosphere for the first time in a trough. For the rest of the flight the aircraft flew in the troposphere (PV<1 PVU). (PV map, PV X-section, PV versus time, wind map). The aircraft probably flew near or in patchy shallow clouds from time to time up to 40 W, 1 h45 (cloud cover X-section, high cloud cover map, IR cloud photograph, cloud cover versus time). WAS4 shows some evidence of uplifting from the lower troposphere over and near Florida.
• 11 August 2001: decaying filamentary and vortex structures, some ITCZ air.
  This flight sampled some decaying filamentary and vortex structures (PV map). Most of the flight was in the troposphere with PV < 2 PVU(PV X-section). During the first half of the flight when the aircraft sampled the filamentary structures, it was close to the tropopause from time to time. At 11h45 a small vortex at 15 W, 42 N (PV slightly more 3.5 PVU) was crossed (PV versus time). From the observed and modelled temperature versus time it can be seen that ECMWF very well analysed its position (only a small phase error).
  At the end of the flight near South America the aircraft was in the northern part of the ITCZ (see map of high clouds and convective precipitation, and IR cloud photograph). The map of eq. potential temperature at 700 hPa shows that the tropical band of warm moist air was centred at about 10 N (in between the red contours).The clouds near the destination were above and away from the aircraft (cloud cover X-section)
• 26 August 2001: a low with 2 folds
  This flight crossed through a low or trough (high PV) between about 48 and 58 W (see PV map and PV X-section) with PV influences down to the surface. Both the PV and specific humidity X-section show that shallow folds were present on both sides of the low. This is also clear from the PV as a function of time that shows a double maximum between about 9h45 and 11h15. Perhaps this will also be visible in ozone. The measured temperature exhibits strong fluctuations of more than 2 K during this period. These might be due to clear air turbulence generated by the strong wind shears (see wind speed and X-section of northward wind v). A deep frontal cloud band was present on the eastern (before 10 h) side of the low (see cloud cover X-section and high cloud cover map). The IR cloud photograph is consistent with this, demonstarting a cloud band encircling the low and passing over the southernmost tip of Greenland. The photograph shows only a small patch of deep convective clouds in the centre of the low. There's also some evidence for this in the specific humidity map which shows a small maximum in the centre of the low.
• 18 November 2001: a low with a deep fold on its eastern side
  A low with a deep fold on its eastern side (40 W) was crossed between 40 W and 60 W, 12h30-15 h (see PV map , PV X-section and PV as function of time). On the same side there was a pronounced front in the lower troposphere up to 500 hPa (strong gradients in the eq. potential temperature X-section). There was also a pronounced deep frontal cloud band on the eastern side of the low (see cloud cover X-section, high cloud cover map and IR cloud photograph at 12 h and the vertical upward velocities at 500 hPa and 250 hPa).
• 19 November 2001: a frontal cloud band
  The return flight was partly parallel to the eastern boundary of the low crossed in the flight of 18 November 2001 (see PV map). Only in the short period between 0h15 and 1h15 the aircraft was in the trough (PV as a function of time, PV X-section near profile 80). There was still a deep frontal cloud band on the eastern side of the low (after 0 h 45) and the aircraft flew more or less along it (cloud cover X-section, cloud cover as a function of time, high cloud map, IR photograph at 0h, and the vertical winds at 500 and 250 hPa). WAS sample 7 seems to have been influenced by the upward transport in the cloud band according to the trajectory calculations (actual upward transport will have been even faster). Such features are also known as warm conveyor belts. The aircraft moved to a lower altitude near the middle of the low at about 0h45 when it also entered the cloudy region according to ECMWF. It benefitted from the high wind speed along the front from then on (map of wind).
  WAS samples 9, 10 and 11 also show uplift.
• 26 January 2002: stratospheric air
  This flight was mostly in the stratosphere at PV values of 5-7 PVU (PV map, PV X-section, PV as a function time). The plane flew through an extensive low/trough (map of Z) north of the jet stream (wind map). There were regularly deep clouds in the troposphere below (cloud cover X-section, high cloud map, IR cloud photograph).
• 21 february 2002: Atlantic blocking high, a small low, a cloud band and a filament
  • In the first part of the flight over the eastern Atlantic the aircraft flew in a blocking high (map of Z). The anticyclonic motions are also evident in the trajectories in the eastern part of the flight.
  • The aircraft flew through furthermore across a small low between 46 and 54 W, approx. 13h55-14h35, with folds on both sides(PV map, PV X-section, PV as a function of time). The plane went up from about 300 to 250 hPa in the centre of this low. The low had fold-like structures on both sides well below the aircraft but not at flight altitude.The low specific humidity in the low (stratosphere) simulated by ECMWF is not reproduced (unsurprisingly?) by the measurements. There is some evidence of weak deep convection in the low itself (green patch at 50W, 30N in the high cloud map) but the IR cloud photograph is dominated by a cloud band on the eastern side of the low, well in the blocking high, that was sampled by the aircraft (cloud cover versus time, X-section of cloud cover).There are somewhat enhanced measured temperature fluctuations (turbulence?) inside the low. Turbulence might be caused by the large wind shear (see wind speed and wind direction versus time and X-section of northward wind) in the low.
  • A smaller fold-like structure was crossed at 12h20 (PV X-section, PV as a function time). It is due to the filament at 40 N, 35 W in the PV map.
• 22 february 2002: northern parts of the low and blocking high of 21 february 2002 and cloud band
  Obviously the return flight more closely followed the jet stream (wind map) than the flight to the Dominican Republic. It sampled the northernmost parts of the low (22h15-23h15) and blocking high (map of PV, PV X-section, PV versus time). The part of the flight between 16 and 30 W, 2h-3h, was at about 2 PVU (PV versus time), so more or less at tropopause level at the boundary between the blocking high and the polar low (along the jet).
  The cloud band in the blocking high (east of the low) was also sampled again (high cloud cover map, cloud cover X-section, IR photograph).
  WAS samples 10 and 11 show uplift a.o. from the south-Caribic.
• 7 March 2002: blocking high, filament, tropopause region (3 PVU)
  • The plane flew through a blocking high over the east Atlantic (maps of PV and Z).
  • At about 12h10 the plane flew through (the tropospheric part?) of a filament at 30 W. West of the filament (12h15) there may have been a filament of uplifted air (cloud cover X-section).There was also a cloud band in the IR satellite imagery and high cloud cover map.
  • Finally in the western Atlantic the plane flew at about 3 PVU in what seems to be an intrusion of lower stratospheric air into the subtropical troposphere (X-section of PV).
    
• 16 March 2002: troughs and a narrow ridge, and a decayed cut-off low in the subtropics, a cloud band.
  The aircraft first flew near the tropopause in a trough over Europe, then in the troposphere through a narrow ridge east of Ireland (10-20 W, centred around 11h30), then for a considerable time in the stratosphere a pronounced trough with tropopause below 400 hPa (20-50 W, about 12h15-14h30), finally in the troposphere towards Cuba (see PV map and PV X-section; also see map of Z). In this last flight segment it touched the tropopause in a region influenced by what is probably a decayed cut-off low (air of northern origin) in the subtropics. The PV along the flight track in the pronounced trough reached up to 8 PVU, in the decaying cut-off low only up to 1 PVU.
  There were deep clouds on both sides of the pronounced trough (cloud cover X-section), most intense at its eastern side. Both belong to the same frontal cloud band encircling the big low (high cloud cover map). There was also some "deep" convection in the middle of the trough which however did not extend above about 400 hPa, the tropopause, well below the aircraft. These clouds are visible in the map of medium cloud cover. The modelled cloud fields are consistent with IR cloud composite and the Meteosat IR photograph at 12 h.
  When passing from the big trough into the subtropical tropospheric high the aircraft almost perpendicularly crossed the jet stream (at 50 W, about 14h30; map of wind, wind speed versus time) where the wind was more than 70 m/s even though it was just in between two jet stream maxima. Right there the airvraft briefly went up to a higher cruise level.
• 17 March 2002: jet stream, troughs and a ridge with a frontal cloud band.
  The return flight was very similar to the flight of 16 March, except that in the beginning, in the subtropics, between 75 and 45 W, the aircraft flew more or less along the jet stream, or equivalently the PV-gradient (PV map, wind map, PV X-section). The decayed cut-off low remained far to the south. Afterward again the big trough, the ridge and the smaller trough over Europe were subsequently crossed. There was still a cloud band around the big trough however less intense than on the 16 March flight and this was partially sampled (high cloud map, cloud cover X-section, IR cloud photgraph).
  
• 13 April 2002: ridges (highs) and filament crossings.
  There was a trough over Europe and a ridge (blocking high) over Britain. The aircraft straight across this blocking high(5-25 W) and then across a PV filament at the place where it was attached to a trough. From 40-60 W the aircraft flew parallel to the filament but west of it, again in a high, profiting from the maximum in southwestward wind there (see map of PV). At the end of the filament crossed its southwesternmost tip. The aircraft was just above the tropopause as it crossed the filament near its base (3.8 PVU, 11h15) and also as it crossed its tip (3 PVU, 15h30) (PV X-section, PV versus time).The flight was relatively cloud free (high cloud map, cloud cover X-section, cloud cover versus time). There was a cloud band on the eastern sied of the big trough and filament, but the aircraft just passed at a place where it had dissolved (IR cloud photograph). The temperature measurements do not show increased variability (turbulence) during the filament crossings.
• 27 April 2002: troughs
  This flight started in a big trough (PV 6-8 PVU) and continued in this trough up to 40 W. Between 50 and 70 W the aircraft flew in a second big trough (PV about 8 PVU) (PV map, PV X-section). In between (12h-13h15, 40-50W) it was briefly in the troposphere. After 16 h, west of 70 W, it was in the subtropical troposphere (PV versus time).
  The flight was mostly above the clouds, except briefly (15h40-16h45) after passing into the tropical troposphere (cloud cover X-section, IR cloud photograph, cloud cover versus time).
• 28 April 2002: along edges of troughs along jet stream
  The situation was more or less the same as on 27 April 2002, but the aircraft now flew along the frotnal zone at the southern edges of the two troughs (PV map) to take full advantage of the jet stream (wind map). Between 47 and 23 W it was in a ridge (high). While flying along the westernmost trough the aircraft remained slightly south and below the tropopause(PV< 2.5 PVU), when flying along the easternmost trough it entered the stratosphere (7-8 PVU) (PV X-section, PV versus time).
  As during the flight to Cuba, there are hardly any cloud influences (cloud cover X-section, high cloud map, cloud cover versus time, IR cloud photograph).Only between 40 and 50 W (0h15-1h15) the aircraft crossed a frontal cloud band associated with the westernmost trough. In the frontal zone along the eastside of this trough the air moves northward and slowly upward causing this band. The vertical winds at 500 and 250 hPa there are not very pronounced.