1
Water vapor heterogeneity related to stratospheric intrusions
over the northern Atlantic revealed by airborne water vapor
lidar
H. Flentje, A. Dörnbrack, G. Ehret, A. Fix, C. Kiemle, G. Poberaj1 and M. Wirth
DLR Oberpfaffenhofen, Institut für Physik der Atmosphäre, D-82230 Weßling, Germany
Abstract. Airborne differential absorption lidar (DIAL) measurements of tropospheric water vapor
and aerosol/clouds have been performed along extended flight sections across the northern and middle
Atlantic on 13-15 May and 16-18 June 2002. Owing to intense dynamical activity over the Atlantic
complex atmospheric structures such as upper tropospheric/lower stratospheric intrusions to the lower
troposphere were observed. Those intrusions with H2O mixing ratios well below 50 ppm are a frequent
phenomenon rather than rare exceptions. They occur on the anticyclonic side of the polar jet and do
mostly but not always correspond to PV anomalies in ECMWF analyses. The signature in aerosol is
less marked due to small gradients at UT/LS levels but typically, narrow undulated aerosol streaks
occur along the intrusion axes reflecting the differential advection by strong strain acting on vertically
wind sheared air in the vicinity of the jet. The observed features cover a large range of scales from
synoptic wave disturbances to mesoscale fronts, gravity waves and convective cells further to turbulent
entrainment and flow processes at the small end of the cascade. Correspondingly huge is the dynamical
range of the associated water vapor mixing ratios, covering more than 3 orders of magnitude thereby
exhibiting very large gradients - typical scales of the intrusions are 1 km vertically and few 100 km
horizontally. ECMWF-based 3-D trajectories and simulations with the MM5 mesoscale model are
mostly capable to reproduce and explain the ongoing dynamical processes. This data set will benefit
the design of a satellite borne DIAL.
1. Motivation
Water vapor is the most important greenhouse gas [Möller, 1963;
Manabe and Weatherald, 1967]), its mixing ratio controls cloud
formation [Kiehl and Trenberth, 1997]) and radiative and chemical
properties of aerosols. Although water vapor constitutes only an
insignificant portion of the mass of the atmosphere (2x10 -6...2%),
moist processes play a fundamental role in atmospheric dynamics.
Furthermore, water vapor is notoriously difficult to measure. Its
concentration is highly variable in both space and time and varies
over four orders of magnitude in the atmosphere. Monitoring water
vapor with high accuracy, especially in the upper troposphere and
lower stratosphere (UT/LS) region is an important scientific issue.
In spite of its relevance, water vapor is one of the least accurate
parameters in the data assimilation for numerical weather prediction
(NWP) models. The adequate representation of the hydrological
cycle in numerical models will be certainly one of the major
challenges in the next few years. A necessary prerequisite is the
accurate measurement of water vapor fields in the atmosphere and
its assimilation in NWP schemes.
Most operational NWP analyses are based on lower and middle
tropospheric humidity from regular radiosonde soundings.
1
now at ETH Zürich, Institute for Quantum Electronics, CH-8093 Zürich, Switzerland
2
Radiosonde data are presently the only technique to provide vertical
water vapor profiles routinely for data assimilation. However,
assimilation schemes only use data for pressures p>300 hPa [Elliott
and Gaffen, 1991; Leiterer, 1997]) as the sensors do not provide
reliable water vapor at higher altitudes or in dry regions (<100
ppmv (0.0622 g/kg2)). Moreover, major “birth” regions of severe
continental weather (e.g. sensitive regions along the Atlantic or
Pacific storm tracks and at the east coasts of North America or
Asia) to which the short-term forecast errors are most sensitive are
only covered by a few radiosonde stations, see Marseille and
Bouttier, [2001].
In order to overcome these deficiencies weather services add
satellite products, such as data from the Tropical Rainfall Measuring
Mission (TRMM) Microwave Imager or the US Defense
Meteorological Satellite Program (DNMSP) special Sensor
Microwave/Imager (SSM/I) to their assimilation scheme [Moreau et
al., 2003]. These passive sounders usually have a large measuring
volume and reveal water vapor information between 200 and 500
hPa. In order to obtain humidity profiles complicated inversion
procedures (retrievals) have to be applied to the measured radiance.
Since in general the inversion procedure is ill-defined,
parameterizations have to be applied. The error characteristics of
these procedures are not well described and the accuracy of the
retrieved humidity is quite low. Even by modern assimilation
systems (ECMWF currently uses measured radiance directly in their
assimilation scheme) the background humidity error cannot be
reduced significantly below the range of 20-50% with tendency to
even larger errors in regions with large spatial gradients.
Meteorological services thus consider to add new techniques to their
assimilation like GPS occultation [Kursinski et al., 1996] or active
remote sensing by radar and lidar.
The latter involve less complex retrievals and provide independent
successive profiles. Particularly, Differential Absorption Lidar
(DIAL) measurements are able to provide spatially and temporally
highly resolved vertical water vapor profiles of high quality (<20 %
random error on specific humidity). Airborne DIAL data have been
used to investigate the aerosol and water vapor distribution in the
troposphere [Ehret et al., 1998; Browell et al., 1998; Bruneau et al.,
2001], the convective boundary layer [Kiemle et al., 1997], above
the Alps [Volkert et al., 2003] and in regions of very low water
vapour concentrations near the tropopause [Ehret et al., 1999;
Hoinka et al., 2003]. A routine assimilation of such data would be a
major step to overcome shortcomings of the humidity retrievals.
In order to acquire global water vapor fields operationally with high
spatial resolution and accuracy a space-borne water vapor DIAL has
been proposed. Presently, it is in the feasibility study process for an
“Earth Explorer” mission of the European Space Agency (ESA),
namely WALES (Water Vapor Lidar Experiment in Space). This
multi-wavelength instrument will be capable of measuring water
vapor profiles from the ground to the tropopause. Owing to the high
raw data spatial resolution, profiles can even be obtained through
km-scale cloud gaps in overcast regions. During the preparatory
phase the design and performance of the WALES instrument are
optimized. End-to-end simulations serve to select the set of H2OConversion of water vapor mixing ratio: wH2O [g/kg] =   10-3  wH2O [ppmv], wH2O being the mixing ratio in g/kg or
ppmv, respectively and =0.622 the ratio of molecular mass of water vapor to dry air. (1 g/kg  1608 ppmv)
2
3
absorption lines and transmitter/receiver parameters needed to reach
NWP impact requirements throughout the troposphere as defined by
the World Meteorological Organization (WMO, 2000). Also, the
impact of horizontal and vertical gradients in the water vapor field
on the DIAL performance, the typical signal-range and signalcoverage can only be estimated in advance using representative
atmospheric distributions of water vapor, clouds and aerosol optical
depths provided by long-range airborne lidar measurements.
Data similar to those anticipated by the future satellite generation
have been collected during two long-range flights across the
northern and middle Atlantic ocean by our DIAL system onboard
the DLR Falcon. Here, we survey the airborne DIAL measurements
of water vapor and particles which cover the middle and high
latitudes from the planetary boundary layer (PBL) up to the upper
troposphere in late spring and early northern hemispheric summer.
Meteorological analyses and model simulations are utilized to place
the observations in context with dynamical fields. Although the
paper’s original contribution is the general survey of the extended
DIAL water vapor profiles in their meteorological context, the data
challenge several scientific questions:
-
-
-
-
Do the observations increase our knowledge about the water
vapor field in data-sparse regions? What are the general
characteristics of the vertical water vapor distribution?
Do we get new insights into the structure (extent, depth) of
tropopause folds? Can the observations be used to constrain
current climatologies regarding the stratosphere-troposphere
exchange (STE), e.g. with respect to the depth of the analyzed
tropopause folds [Sprenger et al., 2003]? Do we find quasihorizontal layers ubiquitously present in the troposphere as
found in previous studies based on MOZAIC data [Newell et
al., 1999]?
What are the technical limitations (resolution, accuracy, low
water vapor concentrations) of the DIAL measurements under
the great variety of observed atmospheric structures?
Is there detectable injection of laminar aerosol layers from the
UT/LS region into the troposphere observed as depolarizing
particle striae?
The paper is organized as follows. Section 2 gives a brief
description of the experimental procedure, the DIAL and the data
evaluation, in Section 3 the DIAL measurements are presented and
meteorological interpretations are given for selected cases. In
Section 4, the scientific potential of the data set with respect to
stratosphere-troposphere exchange, data assimilation and their
implications for a water vapor DIAL in space are discussed. Section
5 summarizes and concludes the paper.
2. Experimental and Diagnostic Tools
2.1. Transfer Flights
The participation in the International H2O Project (IHOP)
campaign from mid May to mid June in Oklahoma (Central USA)
offered the opportunity to measure tropospheric water vapor and
aerosol backscatter ratio along two extended flights across the
northern and middle Atlantic ocean. The transfers took place from
13 to 15 May 2002 (E  W) and from 16 to 18 June 2002 (W  E)
and went from Germany via Iceland, Greenland and Canada to
4
Oklahoma (westbound ferry) and from Oklahoma via Maine
(USA), the Azores and Spain to Germany (eastbound ferry) as
shown in Figure 1. Following the sun, the westbound measurements
mostly took place in the dark at low solar zenith angles while the
eastward transit flights were predominantly performed during
daylight conditions. The typical flight level was 34000 ft (about 10
km altitude) and the total flight time for both transfers ( 18000
km) was about 26 h.
2.2. Water Vapor DIAL
For the IHOP campaign the DLR water vapor DIAL [Ehret et al.,
1998; Poberaj et al., 2002] has been installed onboard the DLR
research aircraft Falcon 20E (http://www.dlr.de/FB/OP) in nadir
looking arrangement. The transmitter is based on a Nd:YAG
pumped, injection seeded KTP-OPO (Optical Parametric
Oscillator). The Nd:YAG laser is operated in the single longitudinal
mode with 220 mJ per pulse at 100 Hz. Half of the fundamental
output at 1064 nm is converted to the second harmonic, serving as
the pump for the OPO, which produces 18 mJ per pulse at 925 nm.
The remaining laser radiation is used for atmospheric backscatter
measurements. A spectral purity of more than 99% is mostly
achieved during in-flight operation, the online frequency is
controlled for each shot by use of a water vapor absorption cell.
Three locking loops maintain a stable wavelength operation, the
exact wavelength information is later needed for the calculation of
water vapor concentrations. The back-scattered photons are
collected by a Cassegrain-type telescope with an aperture of 35 cm
and a field-of-view of 2 mrad. The received radiation is split into
separate polarization channels at 1064 nm and at 532 nm for the
aerosol measurement and the on-line and off-line signals at 925 nm
for the water vapor measurement. Background radiation is
suppressed by a 1 nm interference filter.
Water vapor when measured in the 925 nm spectral region allows to
cover the range of typical concentrations from the PBL to the upper
troposphere. The vertical and horizontal resolution is 500 m and 20
km in the PBL and about 250 m and 1 km in the upper troposphere.
Therefore, this instrumental setup allows to resolve micro- to
mesoscale structures and large humidity gradients e.g. occurring
near frontal zones.
The aerosol optical properties will be expressed in terms of the
particle backscatter ratio R, defined as the total divided by the
molecular backscatter coefficient R = (p+m)/m measured at 1064,
925 and 532 nm. Additionally, the colour ratio CR = 532nm/1064nm
as an indicator for the effective size of the optically dominant
particles is defined. From the parallel and perpendicular polarised
returns referred to the laser beam at 1064 and 532 nm the particle
depolarisation ratio p,532 = p,/p, (zero for spheres) for the
optically dominant particles is calculated, in case of mixed phase
clouds/aerosols few non-spherical particles masked by plenty of
spherical ones may be detectable only by the sole perpendicular
signal p,. The raw data resolution for the backscatter measurement
is about 100 m horizontally and 30 m vertically.
2.3. Meteorological Analyses and Mesoscale Simulations
The synoptic situations are characterized with the aid of T511L60
operational analyses of the European Centre for Medium Range
Figure 1
5
Weather Forecast (ECMWF). The analyses are interpolated onto a
regular 0.5°0.5° latitude-longitude grid. All the numerical
simulations are performed with the NCAR/PSU mesoscale weather
prediction model MM5. The non-hydrostatic compressible model is
used in its version V3.4.0 [Dudhia, 1993; Dudhia et al., 2001]. The
upper boundary is set to 10 hPa, altogether 115 vertical levels are
used (i.e. z250 m). The horizontal grid size is x=36 km and for
some simulations a local grid refinement scheme with x=12 km
has been applied. The simulations include parameterizations for
moist processes (Reisner), convection (Grell) and the MRF PBL
parameterization of vertical diffusion. Radiation effects are
neglected. The model is initialized with the ECMWF analyses and
the integration is forced by 6 hourly ECMWF-analyses at the outer
lateral boundaries. A modified upper boundary condition is used to
prevent the reflection of gravity waves [Zängl, 2001].
Zitate????
3. Observations
3.1 Meteorological Conditions
Figure 2 depicts the potential temperature  and the magnitude of
the horizontal wind speed VH on the dynamical tropopause (the 2
PVU surface; 1PVU = 10-6 m2s-1Kkg-1). Large -values correspond
to a high tropopause, small -values to a lower tropopause altitude.
Large -gradients (“fronts”) are related to a significant vertical
increase in VH which define the location of the jet streams in these
transition zones. Unlike constant-pressure charts, all tropopausebased jets appear here on one map (e.g., polar jet and subtropical
jet). Due to the dominating south-north -gradient the jets are
preferentially zonally oriented. Rossby waves appear as undulations
of these gradients, and coherent vortices appear as regions of closed
-contours (Morgan and Nielsen-Gammon, 1998, northern
hemispheric tropopause maps based on ECMWF analyses are
updated regularly at http://www.pa.op.dlr.de/arctic).
The period of the westbound ferry flights (13-15 May 2002; Fig 2a)
was characterized by a rapid cyclogenesis over the north-east of the
United States (around 80°W, 45°N) and above the Atlantic Ocean,
respectively. The troughs appear as southward excursions of low values, i.e. low tropopauses (stratospheric intrusions) accompanied
by strongly curved subtropical jet streams forming a nearly
sinusoidal - and VH-pattern from 90°W up to the Greenwich
meridian on 15 May 2002 00UTC. Other relevant features of this
period are the persistent ridge above Greenland (tropopause height
at about 240 hPa) and a mesocyclone (polar low) stretching
northward above the Norwegian Sea (tropopause height at about
540 hPa). For the eastbound ferry flights (16-18 June 2002, Fig. 2b)
one month later, the meteorological situation has changed
seasonally: larger -values (higher tropopause) extend further north
and the meridional -gradients are generally less pronounced
resulting in a weaker subtropical jet stream. However, an already
existing Rossby wave over the Atlantic amplifies while propagating
eastward during this short period. The amplification results in a
deep low west of British Isles and a significant southward excursion
of stratospheric air between the Azores and Spain.
Figure 3 details the synoptic situation around the respective ferry
flights with the aid of the geopotential height and the relative
vorticity at 500 hPa. The first segment of the westbound ferry
Figure 2
Figure 3
6
flights (Munich-Keflavik) intersected the southern arm of the polar
low above the Norwegian Sea, the second one passed the
anticyclone above Greenland, and the last two flight segments just
crossed the developing cyclone over North America while the final
part of the flight region was dominated by a large anticyclone over
the Mid-west of the USA. The first eastbound ferry flight from
Oklahoma City to Bangor, Maine encountered a similar but weaker
cyclone over the northwest compared to the last westbound ferry
flight. The second eastbound flight took place under nearly zonal
flow conditions with small dynamical components whereas the third
flight from the Azores to Spain passed the southern tip of the trough
centered west of the British Isles. The meteorological conditions
during the last flight were dominated by a high pressure system
over Central Europe.
3.2. Lidar Measurements
The west- and eastbound transits of the DLR Falcon were split into
4 segments each. The DIAL system operated rather stable such that
more than 16 h of water vapor and aerosol/cloud data could be
collected during the approximately 26 flight hours. There is huge
variability and complexity both in the water vapor as well as in the
aerosol backscatter observations as shown in Figures 4 and 5. The
water vapor profiles are dominated by strong vertical gradients:
high water vapor mixing ratios q>5 g/kg are mostly confined to
altitudes less than 4 km and q decreases exponentially with
increasing altitude, hence logarithmic color scales are used. Only
for the flight near the subtropics (17 June 2002 around the Azores)
higher q-values are measured at greater altitudes which are
associated with rising thermals of the tropical convection. It must be
noted that in some parts of the flights dense tropospheric clouds
strongly attenuated the laser beam, and occasionally, the beam was
even completely blocked. These regions without any backscatter
signal below these clouds (and also without water vapor data) are
left white in Figs. 4 and 5. Table 1 provides an overview of the
water vapor data coverage and the horizontal and vertical
resolutions for the individual flight segments given a statistical error
below 10 %.
The most remarkable feature of the water vapor observations is the
ubiquitous presence of deep intrusions of dry upper tropospheric,
lower stratospheric air into the mid and lower troposphere. Thereby,
a value of q100 ppmv (0.0622 g/kg) is used as an approximate
threshold value for separating tropospheric and stratospheric air
(see discussion in Section 4 about the origin of the air masses).
These intrusions have different vertical extent and width. In Figs. 4
and 5, the dry intrusions tilt westward in most of the cases: near
15°W, 40°W, and 80°W on the westbound flights and at around
85°W, 40°W, and 20°W during the eastbound flights. However, as
the aircraft heading always had southerly or northerly components,
the tilting direction is actually also north-southward. The only
exception is a very narrow filament which tilts into the opposite
direction close to 10°W in Fig. 4 (see discussion in the next
section). Most of the intrusions are related to low tropopause
altitudes whereas in none of the cases the ECMWF analyses give
such deep 2 PVU tropopause as the observed stratospheric q-values
in the tropopause folds would suggest.
Figure 4
Figure 5
Table 1
7
In terms of the aerosol distribution, the entire troposphere is
interspersed with optically thin filaments of background particles
(R1064nm<2) which are known to be relatively small: r  0.1 µm
[Jäger et al., 1995]. Elevated particle backscatter ratios (R1064nm25) observed in the vicinity of the dry intrusions are correlated with
enhanced depolarization and small color ratios (not shown). This
indicates that the corresponding particles are solid, roughly one µm
in radius and are embedded in the smaller background aerosol
particles. The particles inside the intrusions clearly originate in the
tropopause region and exhibit stacked aerosol layers as a signature
of stable stratification and little vertical mixing. As they are
transported into the mid troposphere with the intrusions the majority
of the multi-layered aerosol filaments roughly follow the axes of the
low water vapor intrusions, however they also extend beyond the
dry domains. Emanating from an initially quasi-horizontally
alignment the downward transport leaves the inner structure of the
air-mass mostly unaltered. Similar coherent aerosol patterns have
been observed in the vicinity of stratospheric intrusions during
earlier campaigns over the Alps e.g. [Hoinka et al., 2003] already.
The origin of the particles will be discussed in section 4.
In the following sections, selected structures of the observations
will be discussed in more detail. A fine-scale comparison of the
DIAL water vapor data with model analysis will be given. It is
beyond the scope of this paper to present an in-depth interpretation
of all the details – this will be subject of following papers. Here,
basic findings along the transfer flights will be summarized and the
potential of this data set to answer outstanding questions will be
discussed.
3.2.1. A narrow filament of a mesocyclone, 13 May 2002
During most of the flight on 13 May 2002 from Germany via
Scotland toward Iceland the DLR Falcon flew in and above thick
cirrus clouds extending vertically down to about 6 km altitude (Fig.
4). These compact cloud layers were associated with the warm
sector of the depression system west of the UK (see the stratiform
cloud band extending from UK toward Norway in the NOAA
AVHRR imagery at 13 May 2002 1639 UTC and 14 May 2002
0218 UTC in http://www.sat.dundee.ac.uk). In the cloud layer the
backscatter ratio increased up to R100 and although the laser beam
was strongly attenuated and partly blocked, a few cloud gaps
exceeding the horizontal averaging scale (cf. Table 1) allowed the
detection of mid-tropospheric clouds. There, the water vapor
distribution with typical mid-tropospheric values of q1 g/kg 
1607 ppmv could be determined between the upper and middle
tropospheric clouds (see Fig. 4). In the western part of this flight
(from 10°W to 20°W) the water vapor concentration dropped
rapidly due to the post-frontal clearing. At low altitudes broken
stratocumulus clouds near the top of the PBL dominate the aerosol
backscatter whereas at greater altitudes the backscatter ratio is
rather low (R1...2) except in a few cirrus clouds detected below
the dynamical tropopause. The most surprising feature of this low
water vapor region is the eastward tilting narrow filament at 16°W
extending down to about 3 km altitude.
Figure 6 shows a close-up of this roughly 300 km long water vapor
filament with a vertical depth of about 1500 m. The water vapor
mixing ratio ranges from less than 20 mg/kg (30 ppmv) at the top to
about 70 g/kg in the deepest parts of the filament. The mesoscale
Figure 6
8
model simulation reproduces the shape and position of the observed
filament with high accuracy (Fig. 6). For this simulation a nested
domain with x=12 km had to be used in order to simulate the
dynamics properly. However, the very low water vapor mixing ratio
inside the filament is not reproduced: at 4 km altitude the modeled
water vapor mixing ratio is about 5 times larger than the DIAL
measurements while the discrepancy is only a factor of two in the
upper region.
Remarkably, the underlying dynamical processes generating the
narrow filament are simulated very well by the mesoscale model.
According to the superimposed wind field, the filament is located in
a region of strong vertical and horizontal shear separating two
different air masses: wet air from the warm sector of the depression
above the British Isles slides up- and northward over colder
descending air masses flowing south along the rear of the cut-off
low over the Norwegian Sea. Figure 7 depicts the simulated
horizontal water vapor distribution at two stages of the evolution of
the polar low between Iceland and Norway. The time scale of the
formation and meridional elongation of the mesoscale cyclone is
about 3 days whereas the narrow filament develops in just a couple
of hours (Figs. 7 and 8). Due to the ambient horizontal deformation
the dry filament becomes horizontally stretched and elongated. Ten
days backward trajectories run with the code LAGRANTO [Wernli
and Davis, 1997] and based on ECMWF analyses reveal the
stratospheric origin of the air inside the filament (Fig. 8). The dry
and high-PV air parcels associated with the filament originate from
the high Arctic and descend rapidly to their arrival points at 350
hPa in the last two days before the observation. Thus, it is the
combined horizontal deformation and the spiral cyclonic descent of
the dry stratospheric air parcels that result in the narrow filament
measured by the water vapor DIAL.
The lower panel of Fig. 6 details the aerosol distribution associated
with the intrusion. Apparently, the vertical extent of the aerosol
filaments seems to exceed the signature in water vapor, however
this is probably due to the lower horizontal resolution of the water
vapor data of  0.3-0.5 km compared to  0.03 km for the aerosol
observations. The aerosol filaments extend down to about 3 km
altitude and they are connected to high cirrus clouds and an aerosol
layer in 8 km height close to the ECMWF dynamical tropopause.
The particle composition cannot be fixed with this data while pure
liquid phase can be excluded. According to T-Matrix calculations
[Mishchenko, 1991], a backscatter ratio R=2.30.2, color ratio
CR1.80.1 and particle depolarization ratio p,10640.150.02 are
compliant with nearly spherical (ellipsoid or cylindrical) ice
particles of about 1.5-2 µm radius, however this is questionable
regarding the altitude-/temperature independence. As discussed
below, also ammoniated sulfates in the mixed or solid phase cannot
be excluded.
The large water vapor and aerosol backscatter gradients at the
filament boundary imply a mixing time scale substantially slower
than the mesoscale generation process of the narrow filament of
about 24 hours. The mixing rate at the filament’s boundary and the
question which fraction of the intruded dry air remains in the
troposphere long enough to be mixed microscopically and not be
rejected to the upper troposphere is beyond the scope of this
overview and will be discussed in a later publication.
Figures 7,8
Können Eisteilchen existieren?
T,p NOCH CHECKEN!
9
3.2.2. Low water vapor tongue above Greenland, 14 May 2002
During the flight from Iceland via Greenland to Goose Bay
(Canada) on 14 May 2002 from the Falcon intercepted the southern
edge of the high pressure ridge above Greenland (Fig. 2).
Correspondingly, the observed aerosol field is dominated by the
backscatter from a stratocumulus layer in about 1 km height and by
a few cirrusstratus towards the end of the flight when a trough over
Canada was approached (in accordance with the descending
ECMWF dynamical tropopause, see Fig. 4). A common feature
occurs in the middle of the flight in both aerosol and water vapor:
an intrusion-like pattern of dry air, a tongue of low water vapor
mixing ratios of q0.1...0.3 g/kg near the southern tip of Greenland.
The structure bends down to 3 km altitude and tilts westward with
decreasing height. At its boundaries the mixing ratio increases to
values q>0.5 g/kg. Along the whole flight segment the ECMWF
tropopause is higher than 10 km and its profile does not indicate a
deep intrusion responsible for the dry air in the mid and lower
troposphere (Fig. 4). We guess, that the observed structure is not
caused by vertical descent as occurring in folds but by quasihorizontal (isentropic) transport.
This hypothesis is supported by the simulated water vapor
distribution (Fig. 9). The mesoscale simulations show advection of
moist subtropical air along the western side of Greenland towards
north at the 308 K isentrope. Due to the anticyclonic flow the moist
air returns on the eastern side of the island forming a horseshoe-like
pattern above Greenland. Selected ECMWF backward trajectories
arriving along the flight track of the Falcon also reveal that the
moist air on the western and eastern side of Greenland originates
from mid-latitudes (Fig. 10). However, their individual history is
different: whereas the air parcel arriving on the eastern side traveled
the long way around the island while descending and warming, the
western counterparts spiral up and cool. The up and down of the
parcels is very slow; the simulated vertical velocities are of the
order of 1 cm/s (see Fig. 9). Parcels arriving at the dry end points of
the selected trajectories also descend but their paths are essentially
confined in the anticyclone close to the southern tip of Greenland.
Thus, the mesoscale simulations as well as the ECMWF analyses
reveal an extremely extended mixing region of moist subtropical
and dry Arctic air on the edge of the anticyclone. Keeping the
quasi-persistence of the Greenland high in mind [Scorer, 1988] this
region turns out to be a favored place for stratosphere-troposphere
exchange by quasi-horizontal mixing in an anti-cyclonically
dominated flow regime.
Figure 9, 10
3.2.3. Deep Tropopause fold, 15 May 2002
The last two legs of the westbound ferry from Goose Bay to
Montreal and continuing to Oklahoma on 15 May 2002, 0540-0650
and 0840-1105 UTC crossed an intensifying cyclone centered over
Montreal (Fig. 11 and cf. Fig. 4). The flight legs intercepted the left
hand exit region (blue segment in Fig. 11) and the entrance region
(red segment in Fig. 11) of the curved jet stream. Along the flight
tracks the dynamical tropopause decreases from 10 km to 6.5 km
and wells up to 12 km towards the anticyclone above the mid west
of the United States. The substantial 2 km tropopause drop near
80°W in Fig. 4 indicates the existence of a tropopause fold
associated with the jet stream. In contrast, there is only a weak
Figure 11 ad 12
10
indication of a tropopause fold near the jet exit region. The water
vapor concentration is low down to about 8 km altitude (Fig. 11).
The sharp vertical gradient at this altitude marks the location of the
tropopause at the exit region. Below the tropopause ascending
motions associated with the warm sector of the frontal system
squeeze a fold-like structure vertically and sustain the large vertical
gradient in the water vapor mixing ratio (cf. the simulated vertical
wind with a maximum of +6 cm/s and the water vapor field in Fig.
12). The thin strongly backscattering layer due to clouds at this
location supports this finding (see Fig. 4).
At the anticyclonic side of the jet stream the observed water vapor
distribution reveals a much more complex structure. There, the dry
stratospheric air intrudes down to about 2 km altitude in a main
tropopause fold with a length of about 800 km (Fig. 11). The width
of the fold at 8 km altitude is about 300 km (cf. also Fig. 12). The
minimum mixing ratio measured inside the main fold amounts to
q0.04 g/kg extending down to 4 km altitude and increases to about
q0.1 g/kg in 2 km altitude. Additionally, finger-like structures to
the southwest appear as vertical oscillations in the water vapor field
with nearly 2 km amplitude. Although these dry fingers do not
extend vertically as deep as the main fold their area covers a
significant portion of the middle and upper troposphere in the
entrance region of the jet stream. The water mixing ratio in this
region is less than 0.5 g/kg but never falls below 0.1 g/kg.
The aerosol distribution in the region of the tropopause fold is
dominated by weakly backscattering (R1064mn2-3) striae. Similar to
the narrow filament case on 13 May 2002 these aerosol filaments
are included in the downward motion in the tropopause fold,
actually low backscatter values penetrate wedge-like down to the
ground. In Figure 11 the 0.3 g/kg = 483 ppmv water vapor mixing
ratio contour marks an approximate boundary between tropospheric
and former air masses. Evidently, the backscattering filaments
follow the main tropopause directly whereas elevated values of R
near or outside the main tropopause fold mark mixing between
different air masses. This is consistent with the fact that tropopause
folds extend along the air mass boundaries. The hypothesis of the
particles being composed of ice, grown by air-mass humidification,
can hardly explain the occurrence of the aerosol streak near 44°N in
5-8 km altitude where q<80 ppmv (50 mg/kg) and the DIAL
measurement would be highly sensitive even to small scale humid
patches (see discussion in Section 4.2).
The simulated water vapor field agrees well with the observed
structures, even for the small amplitude oscillations of the dry
fingers in the transition zone from high to low q-values (Fig. 11) at
the anticyclonic side of the jetstream. As in the narrow filament
case, however, the numerical model simulations predict a factor 2
larger water vapor mixing ratio inside the tropopause fold and near
the tropopause. This is most likely an effect of the lower spatial
resolution of the mesoscale model compared to the DIAL data. The
simulated vertical wind in the tropopause fold reaches values of –6
cm/s directly beneath the core of the jet stream (Fig. 12b).
Furthermore, the model indicates several patches of cut-off low
water vapor concentration near 2 km altitude. These patches are
separated from the main intrusion and will later be mixed with
lower tropospheric air. Although we have not calculated ECMWF
backward trajectories for this particular case the observations and
the mesoscale model simulations imply convincingly irreversible
11
injection and mixing of dry stratospheric air masses into the lower
troposphere.
3.2.4. Streamer, 17 June 2002
Pronounced gradients of middle to upper tropospheric water vapor
mixing ratio also occur on the first two eastbound flights in midJune 2002 (Fig. 5). Similar to the event described in the former
section a tropopause fold was measured near 85°W on 16 June 2002
and again, weakly backscattering aerosol filaments exist in and near
the tropopause fold. On the way to the Azores the DIAL observed
strong zonal water vapor gradients, e.g. near 40°W on 17 June 2002
(Fig. 5). These gradients are obviously related to the differential
advection of moist subtropical and dry polar air masses, (cf. Fig.
2b). However, the most surprising and unique structure occurs on
the third flight leg west of 20°W on 17 June 2002 between 1800 and
1900 UTC: a horizontal 300 km long dry layer between 4 and 6 km
altitude that opens into a tropopause fold at about 19°W (Fig. 13).
Structures like this have not been found during previous campaigns
over central Europe. Inspecting Figs. 2b and 3 reveals that the DLR
Falcon crossed the southernmost tip of a streamer between the
Azores and northwest Spain. In the tropopause maps, the evolution
of the streamer is represented by the south-eastward propagation of
potentially colder air and the associated undulation of jet stream
near 20°W (Fig. 2b).
The streamer propagates from west to east along the flight track
between 1200 UTC to 1800 UTC (Fig. 14). As expected, the
ECMWF analyses show ascending motion on the front side and
descending flow on the rear side of the streamer. Whereas the
ascending moist air increases the water vapor gradient in the
vicinity of the streamer, the descending motion on the rear side
results in a more diffusive structure. The propagation of the
stratospheric air inside the streamer dries the mid-tropospheric air in
a layer between 5 and 6.5 km altitude. In this way the streamer
leaves an extremely dry layer behind that remains at the release
position for several hours (see Fig. 13). The location of the dry
layer, the humid region above and the onset of increasing water
vapor near 14°W is precisely reproduced by the ECMWF analyses.
However, the water vapor mixing ratios observed by the DIAL, i.e.
q30 – 80 ppmv (20-50 mg/kg ) in the streamer and minima below
30 ppmv (20 mg/kg) in the driest layer, are again significantly
lower than the ECMWF values.
At the tip of the streamer a structure with enhanced aerosol
backscatter ratio R1.2 is observed in the tropopause fold just
where the ECMWF 2 PVU tropopause dips down to minimum
altitude (Fig. 5). We can only hypothesize about the origin of this
aerosol filament but it is particularly striking in that it seems to
visualize vortical motion associated with the residual circulation
near the jet. This means, the ascending motion near the northerly jet
on the front and descending motion near the southerly jet on the
rear side of the streamer.
On both sides of the streamer, cirrus clouds were observed while
the streamer itself was cloud free, see the infrared and visible
NOAA AVHRR imagery at 1423 UTC in Fig. 15. Below an altitude
of 2 km the maritime boundary layer is marked by enhanced aerosol
backscatter ratios and intensive entrainment near its top (note also
Figure 13
Figure 14
12
the weak stratification in the analyses in Fig. 14). Missing
wavelength dependence of the backscatter coefficient (not shown
here) suggests the presence of large sea-salt particles.
The last segment of the eastbound ferry flights took place in quiet
high pressure conditions over southern and central Europe: the
water vapor mixing ratio decreases exponentially with height and
no significant horizontal gradient have been observed.
4. Discussion
The previous section gave a survey on DIAL water vapor and
aerosol/cloud observations over the northern Atlantic in May/June
2002. Synoptic scale and mesoscale meteorological analyses are
utilized to explain the tremendously complex structures of dry
intrusions. Various cases of fold-like structures in the upper
troposphere and lower stratosphere could be identified and
attributed to different dynamical origin. Streamers and tropopause
folds as discussed in Sections 3.2.3 and 3.2.4 are associated with the
subtropical and polar jet stream and constitute favorable sites of
cross-troposphere exchange [Holton, 1995]. The narrow filament of
the mesocyclone observed over the Norwegian Sea (Section 3.2.1)
might be regarded as a smaller scale version of the folds occurring
in synoptic-scale systems. The analysis of this particular event
reveals that the descent of stratospheric air goes along with
significant vortex stretching. The resulting elongation and
fragmentation of water vapor filaments and their fine-scale structure
are responsible for the consequent irreversible exchange of air
between the stratosphere and troposphere [e.g. Appenzeller and
Davies, 1992; Appenzeller et al., 1996]. An unexpected location for
the mixing between dry and moist air masses in the upper
troposphere represents the edge of the anticyclone over Greenland
(section 3.2.2). However, this case and the streamer case need
further investigation to clarify the origin of the dry air in the upper
troposphere and to study their mixing behavior. In this section we
discuss selected topics as the gross features of stratospheretroposphere exchange, the origin and composition of the aerosol
particles, and the implication of the observation for space lidar
instruments.
4.1. Stratosphere-Troposphere Exchange
Both sets of trans-Atlantic flights reveal diverse low water vapor
structures which penetrate sledge-like down to the lower
troposphere. From the two-dimensional snapshots of observed
water vapor and the superimposed ECMWF dynamical tropopause
(Figs. 4 and 5) it is not possible to deduce the origin of the sampled
air masses. As a first guess we used a threshold value of q100
ppmv to separate tropospheric and stratospheric air masses. In
Figures 9, 12, and 14 the green lines mark simulated water vapor
mixing ratios of q=0.05 (80), 0.10 (160), and 0.15 (240) g/kg
(ppmv), respectively. According to Ovarlez et al. (1999), the water
vapor mixing ratio at the dynamical tropopause has a large spread:
it varies between 20 and 120 ppmv for a value of 2 PVU and for 3.5
PVU the spread is reduced to about 20 to 80 ppmv. The three cases
(Greenland anti-cyclone, deep tropopause fold, and streamer) show
that values q0.05 g/kg (80 ppmv) are situated close to or above the
dynamical 2 PVU tropopause. However, values q0.10 g/kg (160
13
ppmv) can be either stratospheric (deep tropopause fold, Fig. 12a
and near the tip of the streamer event, Fig. 14) or tropospheric.
Meteorological simulations and analyses demonstrate clearly the
stratospheric origin of the dry intrusions for the deep tropopause
fold case and for the streamer event over the Atlantic.
Backward trajectory calculations have been applied to reveal the
origin of the dry air masses for the two other cases. For the narrow
filament (Section 3.2.1) the ECMWF back trajectories could
unambiguously show the stratospheric origin of the air (Fig. 8). For
the Greenland case the transport took place over a longer time
period. Here, we have to keep in mind that if an air parcels cools
below the saturation temperature, loss of water vapor occurs due to
condensation and sedimentation of ice particles. These parcels
remain saturated at the lower temperature. If these parcels later
warm, the water vapor mixing ratio remains unchanged and,
therefore, reflects essentially the previously encountered low
temperature. Therefore, the dry structure could be also result from
cold temperatures the air parcels encountered during their life cycle
(see Fig. 10c). However, this topic and whether and to what extent
the transport is irreversible will be investigated in a future
publication. Furthermore, we plan to analyze the water vapor field
in a similar manner as Ovarlez et al. [1999] and correlate the
observed water vapor mixing ratios to the simulated PV field.
Another interesting point is the vertical extent of the exchange
regions. The deep tropopause fold and the narrow filament range
from 8 km altitude down to 2 km. In contrast the dry regions of the
streamer and the anticyclone only span the altitude range from 5 to
6 km and 5 to 8 km, respectively. The dynamical origin of these
cases was explained and discussed in Section 3. In general, the
geographical distribution of our observations agrees with the results
of the ECMWF tropopause fold climatology by Sprenger et al.
[2003], where medium and deep stratospheric intrusions most
frequently appear north of 40°N.
The correlation of the dry intrusions with the analyzed PV is of
particular relevance for the assimilation and parameterization of
synoptic-scale water vapor fields in NWP models. Owing to the
limited coverage of operationally available water vapor
observations, the H2O-fields may be derived with aid of the
calculated PV distribution. Thereby, the H2O-PV anti-correlation at
upper tropospheric/lower stratospheric levels is utilized, which
holds under many common synoptic conditions and in most of the
observed cases, too. However, occasionally deviations occur, which
necessarily would imply errors on the NWP initialization. Firstly,
ECMWF PV-analyses do not resolve details of the low-H2O
structures and generally underestimate their strength and the
associated gradients. Secondly, considerable displacements occur
between analyzed PV-anomalies and their real location. Thirdly, the
rapid highly non-linear temporal development of the structures may
cause time-step and resolution induced deviations from the real
atmospheric state.
We are aware that our discussion of STE is only based on a few
snapshot-like observations. They do not qualify for being
representative. However, they represent the strong observational
variability of the different dynamical structures responsible for the
cross-tropopause exchange. Most former experimental studies of
STE rely on single altitude observations [e.g. Ovarlez et al., 1999;
14
Beuermann et al., 2002]. Furthermore, reverse domain filling
trajectories running on isentropic surfaces and depending on
arbitrary time lags have difficulties to reproduce the highly variable
water vapor structure. It is the vertical coverage of the water vapor
data and the meteorological analyses applied over a wide range of
atmospheric scales that will make this study valuable for different
purposes, e.g. constraining climatologies of STE.
4.2. Particle Observations
Figures 4 and 5 survey essentials of the observed tropospheric
particle distribution along the northern Atlantic storm track. Three
domains, clouds, aerosols and nearly particle free background
atmosphere can be clearly distinguished in the two-dimensional R
transects. The typical zonal scale of synoptic systems (few 1000
km) reflects in the intermittent occurrence of cirrus clouds
associated with fronts, e.g. 92°W, 72°W, 52°W and east of
20°W on the westbound and 50°W, 35°W, 21°W and 14°W on
the eastbound ferry flight. In between there are large areas ( 60%)
with tropospheric background conditions, i.e. backscatter ratios
R1064nm<2. Finally, extended coherent aerosol structures with R2 as
those described in section 3, occur mostly in and near synopticscale disturbances as tropopause folds and streamers.
There are several possibilities for the origin and composition of the
particle filaments inside and around the stratospheric intrusions.
Particles surrounded by water vapor mixing ratios q<100 ppmv
originate in the lower stratosphere while those in the adjacent
moister regions presumably descended from the upper troposphere.
Occasionally the backscatter ratios even suggest that the aerosol
filaments are connected with lower tropospheric aerosol which
however can be excluded by the low q-values. The particle
composition is much less clear. Besides tracer transport from
synoptic-scale upwelling and deep convection, the main source of
aerosol particles in the UT/LS is aviation, mainly between 9 and 12
km [Schumann et al., 1996; Schröder and Ström, 1997]. While the
size of fresh aircraft exhaust particles is much too small to be
observed at the used wavelengths of the DIAL instrument [Petzold
and Schröder, 1998, Petzold et al., 1999], aged (mostly ice-)
particles grown from the gaseous precursors may reach lidardetectable sizes and form sub-visible cirrus clouds [Kärcher and
Solomon, 1999]. Sulfate aerosols, partially ammoniated aqueous
sulfuric acid particles, also among the most abundant tropospheric
species may occur either solid, liquid or in the mixed phase
depending on relative humidity and (due to the large
deliquescence/efflorescence hysteresis) its Lagrangian history
[Colberg et al., 2002]. Crustal/soil, biogenic or forest fire particles
reach the UT/LS region only occasionally and would thus be much
more limited in extension than observed.
Particle growth induced by entrainment of humid air into the
stratospheric intrusions is a possible mechanism to generate
enhanced backscatter ratios as supposed in a discussion of a similar
structure connected to a tropopause fold over central Europe by
Hoinka et al. [2003]. However in case of ice, this should result in a
closer correlation of enhanced backscatter ratios and low water
vapor than observed. In case of sulfate aerosol, the large hysteresis
may account for low correlation. Regarding the limited amount of
injected humidity however, it seems questionable that
ice/hygroscopic particles in the intrusions could conserve as
15
uniformly distributed sizes and phases under rapid vertical
displacement and mixing as observed.
4.3. Implications for Space Lidars (WALES)
An envisaged benefit of this data set will be its application as
atmospheric input for end-to-end simulations of the WAter vapour
Lidar Experiment in Space (WALES) instrument which has been
selected for phase A of an earth explorer mission by the European
Space Agency (ESA) in 2000 [ESA, 2001; Gerard et al., 2004].
This satellite borne instrument which is envisaged to be launched
after 2010 is to be a multi-wavelength water vapor/aerosol DIAL
for operational water vapor sounding with quasi global coverage as
motivated in section 1. As a major difference in the design, the
dynamic range will be substantially improved by simultaneously
measuring at three differently strong absorption lines. Compared to
hitherto available and climatological data sets the high resolution
and coverage of the presented H2O-distributions are a substantial
step forward regarding a realistic simulation of the WALES
instrument performance. Firstly, H2O gradients, particularly near
stratospheric intrusions or dry/humid layers are not resolved with
this resolution in previous observations. The water vapor
distribution exhibits more complexity than generally expected,
particularly the influence of large gradients and the large range of
H2O mixing ratios can be analyzed and considered in optimizing the
envisaged space borne instrument. Secondly, detailed distributions
of clouds and day/night background radiation enable more realistic
simulation of data coverage and –quality, respectively, particularly
with respect to profiling through cloud gaps. Thirdly, extinction
coefficients of aerosol and thin clouds allow to estimate the particle
induced beam attenuation under various synoptic conditions. In
addition, misalignment constraints between on- and off-line
soundings and a Rayleigh-Doppler error correction scheme can be
established, depolarization effects on WALES data retrieval
estimated, vertical raw data resolution and laser power fluctuation
requirements in inhomogeneous backscatter regime be analyzed.
Thus it is highly recommended that operational acquisition and
assimilation of water vapor data is prepared and simulated by this
and other comparable data sets.
5. Summary and Conclusions
Water vapor and aerosol/cloud distributions have been measured
with an airborne DIAL in mid May and mid June 2002 during two
transfers across the northern and middle Atlantic Ocean. The 8
flights went from Germany via Iceland, Greenland, Goose Bay to
Oklahoma (US) and back via Maine, the Azores, and Spain. During
a total of about 25 flight hours with 7 stopovers summing up to
about 8 hours of ascent or descent periods, about 16 hours of DIAL
data have been collected. During about 7 hours dense clouds at
different levels blocked the laser beam.
The observations show, that the water vapor and particle
distributions over the northern hemisphere mid-latitudes are very
inhomogeneous and exhibit large vertical and horizontal gradients.
Most of these gradients are associated with active weather systems
as rapidly developing cyclones. Therefore, it is not surprising that
the dominant and ubiquitous structure present in both transfers
observed by the lidar is the tropopause fold connected either with
the subtropical or the polar jet stream. Surprisingly, the water vapor
16
field on the anti-cyclonic side of the tropopause fold is not reflected
in the PV field. The quasi-eastward propagation of cyclones and
streamers enhances water vapor gradients on the front side of the
systems due to the predominant ascending motions. On the rear
side, however, the dried mid-tropospheric air shows a large variety
of structures: finger-like dry intrusions on the anti-cyclonic side of
the fold and horizontal layers left behind by a propagating streamer.
The particle backscatter ratio is not generally correlated with the
water vapor mixing ratio although its signature frequently follows
the axes of the stratospheric intrusions. While their origin clearly is
in the upper troposphere/lower stratosphere region, their
composition can not yet be determined – owing to their moderate
depolarization they are most likely composed of ice or ammoniated
sulfate.
Furthermore, there are also regions of enhanced gradients that are
related to weaker dynamical structures. A new aspect of our
observations is the presence of a large-scale dry intrusion into the
troposphere on the edge of the quasi-stationary high pressure
system over Greenland. Due to the long-lasting advection of moist,
sub-tropical air masses and their mixing with dry, polar air masses
this region might be a favorite place for stratosphere-troposphere
exchange by quasi-horizontal (isentropic) transport on long time
scales.
Meteorological analyses (ECMWF and mesoscale modeling) and
backward trajectories have been used to trace the dynamical origin
of the observations. For the narrow filament (Section 3.2.1), for the
deep tropopause fold (Section 3.2.3) and for the streamer (Section
3.2.4) we have shown that the intruded air originates from the
stratospheric reservoir. The origin of the dry air in the anticyclone
Section 3.2.2) is not as clear, here additional investigations have to
be done. Furthermore, we have not studied the further fate of the air
masses to answer the question conclusively if the stratospheric air
has been mixed irreversibly in the troposphere.
The analyses reproduce the observed structures very well. However,
a detailed comparison left some questions open: does the
quantitative disagreement of the water vapor distribution inside the
small to mesoscale intruded structures only result from the coarser
spatial resolution of the dynamical processes? As previously stated,
the fold structures are represented very well. Thus, the water vapor
initialization procedure must be investigated more thoroughly in the
future. Furthermore, we plan to analyze the water vapor data in a
similar way as done in a previous study [Ovarlez et al., 1999] for insitu ozone and water vapor.
Acknowledgements. We acknowledge the ECMWF for supplying
high quality synoptic data. This work was funded by the ESA under the
contract number 16180/02/NL/FF and the United States National Science
Foundation (NSF) under the contract no. ?????????
References
Appenzeller, C. and H.C. Davies, Structure of stratospheric intrusions into
the troposphere, Nature, 358, 570-572, 1992.
Appenzeller, C., H.C. Davies, and W. A. Norton, Fragmentation of
stratospheric intrusions, J. Geophys. Res., 101, D1, 1435-1456, 1996.
Beuermann, J., Konopka, P., Brunner, D., Bujok, O., Günther, O.,
McKenna, D.S., Lelieveld, J., Müller, R., and Schiller, C., Highresolution measurements and simulation od stratospheric and
17
tropospheric intrusions in the vicinity of the polar jet stream, Geophys.
Res. Lett., 29, doi:10.29/2001GL014162, 2002.
Browell, E., S. Ismail and W.B. Grant, Differential absorption lidar
(DIAL) measurements from air and space, Appl. Phys., B 67, 399-410,
1998.
Bruneau, D, P. Quaglia, C. Flamant, M. Meissonnier and J. Pelon,
Airborne lidar LEANDRE II for water vapour profiling in the
troposphere. I. System description, Appl. Opt, 40, 3450-3461, 2001.
Colberg, C. A., B. P. Luo, H. Wernli, T. Koop, and T. Peter, A novel
model to predict the physical state of atmospheric H2SO4/NH3/H2O
aerosol particles, Atmos. Chem. Phys. Discuss., 2, 2449-2487, 2002.
Dudhia, J., A non-hydrostatic version of the Penn State—NCAR
Mesoscale Model: Validation tests and simulation of an Atlantic
cyclone and cold front. Mon. Weather Rev., 121, 1493--1513, 1993.
Dudhia, J., D. Gill, Y.-R. Guo, K. Manning, and W. Wang, PSU/NCAR
Mesoscale Modeling System Tutorial Class Notes and User’s Guide:
MM5 Modeling System Version 3, PSU/NCAR, 300pp., 2001. Also at
http://box.mmm.ncar.edu/mm5.
Ehret, G., A.Fix, V.Weiß, G.Poberaj, Th. Baumert, Diode-laser-seeded
optical parametric oscillator for airborne water vapor DIAL application
in the upper troposphere and lower stratosphere, Appl. Phys., B 67, 427431 (1998)
Ehret, G., K.P. Hoinka, J. Stein, A. Fix, C. Kiemle, and G. Poberaj, LowStratospheric Water Vapor Measured by an Airborne DIAL, Journal of
Geophysical Research, 104, D24, S. 31,351-31,359, 1999
Elliott, W.P. and D.J. Gaffen, On the utility of radiosonde humidity
archives for climate studies, Bull. Am. Met. Soc., 72, 1507-1520, 1991.
ESA :Report for Assessment —The WALES Candidate Earth Explorer
Core Mission.ESA SP-1257(2), 91 pp, 2001.
Gérard, É., Tan, D. G. H., Garand, L., Wulfmeyer, V., Ehret, G., Di
Girolamo, P. 2004: Major Advances Foreseen in Humidity Profiling
from the Water Vapour Lidar Experiment in Space (WALES). Bulletin
of the American Meteorological Society: Vol. 85, No. 2, pp. 237–251.
Hoinka, K.P., E. Richard, G. Poberaj, R. Busen, J.-L. Caccia, A. Fix and
H. Mannstein, Analysis of a potential vorticity streamer crossing the
Alps during MAP-IOP 15 on 6 November 1999., Q. J. R. Met. Soc.,
129, 609-632, 2003.
Holton, J. R., P. E. Haynes, M. E. McIntire, A. R. Douglas, R.R. Rood,
and L. Pfister, Stratosphere-Troposphere Exchange, Rev. Geophys., 33,
403-439, 1995
Jäger, H., T. Deshler and D. J. Hoffmann, Midlatitude lidar backscatter
conversions based on balloonborne aerosol measurements, Geophys.
Res. Lett., 22, 1729-1732, 1995.
Kärcher, B. and S. Solomon, On the composition and optical extinction of
particles in the tropopause region, J. Geophys. Res., 104, 27441-27459,
1999.
Kiehl, J.T. and K.E. Trenberth, Earth’s annual global mean energy budget,
Bull. Am. Met. Soc., 78, 197-208, 1997.
Klett, J.D., Lidar inversion with variable backscatter/extinction ratios,
Appl. Optics, 24, 1638-1643, 1985.
Kiemle, C., G. Ehret, A. Giez, K. J. Davis, D. H. Lenschow, and S. P.
Oncley, Estimation of boundary-layer humidity fluxes and statistics
from airborne differential absorption lidar (DIAL), BOREAS special
issue of the J. Geophys. Res., 102, 29189-29203, 1997.
Kursinski, E. R., G. A. Hajj, W. I. Bertiger, S. S. Leroy, T. K. Meehan, L.
J. Romans, J. T. Schofield, D. J. McCleese, W. G. Melbourne, C. L.
Thornton, T. P. Yunk, J. R. Eyre, and R. N. Nagatani, Initial results of
radio occultation observations of Earth’s atmosphere using the Global
Positioning System, Science, 271, 1107-1110, 1996.
Leiterer, U., H. Dier and T. Naebert, Improvements in radiosonde humidity
profiles using RS80/RS90 radiosondes of Vaisala, Contrib. Atmos.
Phys., 70, 319-336, 1997.
Manabe, S. and R. Weatherald, Thermal equilibrium of the atmosphere
with a given distribution of atmospheric humidity, J. Atmos. Sci., 24,
241-259, 1967.
Marseille, G. J. and Bouttier, F., Climatologies of sensitive areas for shortterm forecast error over Europe. A EUMETNET-EUCOS study.
ECMWF Technical Memorandum No. 334, ECMWF, Reading 2001.
Mishchenko, M. I., Light scattering by randomly oriented axially
symmetric particles, J. Opt. Soc. Am. A. Opt. Image Sci., 8, 871-882,
1991.
Möller, F., On the influence of changes in the CO2 concentration in air on
the climate, J. Geophys. Res., 68, 3877-3886, 1963.
18
Moreau, E., P. Lopez, P. Bauer, A. M. Tompkins, M. Janiskova, and F.
Chevallier, Variational retrieval of temperature and humidity profiles
using rain rates versus microwave brightness temperatures, ECMWF
Technical Memorandum No. 412, 2003 (http://www.ecmwf.int).
Morgan, M.C. and J.W. Nielsen-Gammon, Using tropopause maps to
diagnose midlatitude weather systems, Mon. Wea. Rev, 126, 2555-2579,
1998.
Newell, R., V. Thouret, J.Y.N. Cho, P. Stoller, A. Marenco and H. G.
Smit, Ubiquity of quasi-horizontal layers in the troposphere, Nature,
398, 316-319, 1999.
Ovarlez, J. P.F.J. van Velthoven and H. Schlager, Water vapor
measurements from the troposphere to the lowermost stratosphere:
Some signatures of tropopshere to stratosphere exchanges,
J.Geophys.Res. 104, 14, 16,973-16,978, 1999.
Petzold, A. and F. P. Schröder, Jet engine exhaust aerosol characterization,
Aerosol Sci. & Technol. 28, 62-76, 1998.
Petzold, A., A. Döpelheuer, C.A. Brock and F.P. Schröder, In situ
observations and model calculations of black carbon emission by
aircraft at cruise altitude, J. Geophys. Res., 104, 22171-22181, 1999.
Poberaj, G., A. Fix, A. Assion, M. Wirth, C. Kiemle, and G. Ehret, AllSolid-State Airborne DIAL for Water Vapour Measurements in the
Tropopause Region: System Description and Assessment of Accuracy,
Appl. Phys. B 75, 165-172, 2002.
Schumann, U., J. Ström, R. Busen, R. Baumann, K. Gierens, M.
Krautstrunk, F. P. Schröder and J. Stingl, In situ observations of
particles in jet aircraft exhausts and contrainls for different sulfurcontaining fuels, J. Geophys. Res., 101, 6853-6869, 1996.
Schröder, F. and J. Ström, Aircraft measurements of sub micrometer
aerosol particles (> 7 nm) in the midlatitude free troposphere and
tropopause region, Atmos. Res., 44, 333-356, 1997.
Scorer, R.S:, Sunny Greenland, Q. J. R. Met. Soc., 114, 3-29, 1988.
Shine, K.P. and A. Sinha, Sensitivity of the earth’s climate to height
dependent changes in the water vapour mixing ratio, Nature, 354, 382384, 1991.
Sprenger, M., M. Croci Maspoli, and H. Wernli, Tropopause folds and
cross-tropopause exchange: A global investigation based upon
ECMWF analyses for the time period March 2000 to February 2001, J.
Geophys. Res., 108, D12, 8518, doi:10.1029/2002JD002587, 2003.
Volkert, H., Keil, C., Kiemle, C., Poberaj, G., Chaboureau, J.P., etal.:
Gravity Waves over the Eastern Alps: A Synopsis of the 25 October
1999 Event (IOP 10) Combining In-Situ and Remote Sensing
Measurements with a High-Resolution Simulation. Q. J. Roy. Met. Soc.,
129, 777-797, 2003.
Wernli, H., and H. C. Davis, A Lagrangian-based analysis of extratropical
cyclones. I: The method and some applications, Q. J. R. Meteorol. Soc.,
123, 467-489, 1997.
World Climate Research Programme (WCRP, 2000): Assessment of Upper
Tropospheric and Stratospheric Water vapour, SPARC Report no 2, WCRP
no 113, MWO/TD No-1043; also at: http://www.aerc.jussieu.fr/~sparc.
WMO, Statement of guidance regarding how well satellite capabilities meet
WMO requirements in several application areas, WMO/TD 992, 2000.
Zängl, G., Stratified flow over a mountain with a gap: Linear theory and
numerical simulations, Q. J. R. Meteorol. Soc., 128, 927-949, 2002.
Mailing addresses: author names should be grouped by address and
listed alphabetically.
(Received June 6, 2000; revised July 27, 2000;
accepted March 31, 2001.)
1
Now at, Also at, Formerly at, On leave at, and On leave from are
acceptable. Do not include mailing information.
AGU Copyright:
Copyright 2001 by the American Geophysical Union.
Paper number 2001JA012345.
0148-0227/01/2001JA012345$09.00
Wo sollen die zitiert
werden?
oder raus?
19
Public Domain Copyright:
This paper is not subject to U.S. copyright. Published in 2001 by the
American Geophysical Union.
Paper number 2001JA012345.
Crown Copyright:
Published in 2001 by the American Geophysical Union.
Paper number 2001JA012345.
**Provide running head (45 character max for short title):
FLENTJE ETAL.: LONG RANGE H2O DIAL MEASUREMENTS
20
Fig. 1: Flight tracks of the ferry flights from Europe to Oklahoma
(US) on 13 – 15 May 2002 (westbound) and return on 16 – 18 June
2002 (eastbound) with start and stop times of data acquisition. Gaps
are due to ascent and descents from and to refuel -stopovers.
Fig. 1: Flight tracks of the ferry flights from Europe to Oklahoma (US) on 13 – 15 May 2002 (westbound) and
return on 16 – 18 June 2002 (eastbound) with start and stop times of data acquisition. Gaps are due to ascents and
descents from and to refuel -stopovers.
21
Fig. 2a: Potential temperature (K, left) and horizontal wind speed
(m/s, right) on the dynamical tropopause (defined as 2 PVU surface;
1PVU = 10-6 m2s-1Kkg-1) at the marked days valid at 0000 UTC.
ECMWF T511L60 operational analyses. The Longitude given in °E
corresponds to °W+360°.
Fig. 2a: Potential temperature (K, left) and horizontal wind speed (m/s, right) on the dynamical tropopause
(defined as 2 PVU surface; 1PVU = 10-6 m2s-1Kkg-1) at the marked days valid at 0000 UTC. ECMWF T511L60
operational analyses. The Longitude given in °E corresponds to °W+360°.
22
Fig. 2b: Potential temperature (K, left) and horizontal wind speed
(m/s, right) on the dynamical tropopause (defined as 2 PVU surface;
1PVU = 10-6 m2s-1Kkg-1) at the marked days valid at 0000 UTC, as
in Fig. 2a.
Fig. 2b: Potential temperature (K, left) and horizontal wind speed (m/s, right) on the dynamical tropopause
(defined as 2 PVU surface; 1PVU = 10-6 m2s-1Kkg-1) at the marked days valid at 0000 UTC, as in Fig. 2a.
23
Fig. 3: Geopotential height (m, solid black lines) and relative
vorticity (10-5 s-1, color shaded) at 500 hPa at the assigned dates.
The DLR Falcon flight legs are marked by blue and green lines.
Plots are based on MM5 simulations with 36 km horizontal
resolution (see text for details).
Fig. 3: Geopotential height (m, solid black lines) and relative vorticity (10 -5 s-1, color shaded) at 500 hPa at the
assigned dates. The DLR Falcon flight legs are marked by blue and green lines. Plots are based on MM5
simulations with 36 km horizontal resolution (see text for details).
green & blue lines stronger ???
14 MAY 2002
13 MAY 2002
15 MAY 2002
Fig. 4: Water vapor mixing ratio (mg/kg; upper panel)) and aerosol
backscatter ratio at 1064 nm (middle panel) along the westbound
transfer flights from Germany to Oklahoma (USA) on 13 – 15 May
2002. The black lines mark the height of the dynamical tropopause
(2 PVU surface) based on operational T511L60 ECMWF analyses.
Their valid times are 14 May 00 UTC for the segments east of
60°W and 15 May 06 UTC for the segments west of 60°W. Lower
Panel: corresponding DLR Falcon flight legs. Note, that the vertical
scale is strongly exaggerated due to the tremendous horizontal scale
compression; typical “steep” intrusions actually have aspect ratios
(vertical to horizontal scale) of about 1/50.
Fig. 4: Water vapor mixing ratio (mg/kg; upper panel)) and aerosol backscatter ratio at 1064 nm (middle panel)
along the westbound transfer flights from Germany to Oklahoma (USA) on 13 – 15 May 2002. The black lines
mark the height of the dynamical tropopause (2 PVU surface) based on operational T511L60 ECMWF analyses.
Their valid times are 14 May 00 UTC for the segments east of 60°W and 15 May 06 UTC for the segments west of
60°W. Lower Panel: corresponding DLR Falcon flight legs. Note, that the vertical scale is strongly exaggerated due
to the tremendous horizontal scale compression; typical “steep” intrusions actually have aspect ratios (vertical to
horizontal scale) of about 1/50.
24
25
Fig. 5: Water vapor mixing ratio (mg/kg; upper panel)) and aerosol
backscatter ratio at 1064 nm (middle panel) along the eastbound
transfer flights from Oklahoma (USA) to Germany on 16 – 18 June
17 JUN 02
2002. The black lines mark the height of the dynamical tropopause
(2 PVU surface) based on operational T511L60 ECMWF analyses.
Their valid times are 16 June 18 UTC for the segments east of
60°W and 17 June 18 UTC for the segments west of 60°W. Lower
Panel: corresponding DLR Falcon flight legs.
18 JUN 02
Fig. 5: Water vapor mixing ratio (mg/kg; upper panel)) and aerosol backscatter ratio at 1064 nm (middle panel)
along the eastbound transfer flights from Oklahoma (USA) to Germany on 16 – 18 June 2002. The black lines
mark the height of the dynamical tropopause (2 PVU surface) based on operational T511L60 ECMWF analyses.
Their valid times are 16 June 18 UTC for the segments east of 60°W and 17 June 18 UTC for the segments west of
60°W. Lower Panel: corresponding DLR Falcon flight legs.
26
Achsenb
eschriftu
ng
g/kg
R1064
Fig 6: Upper panel: Water vapor measured by DIAL with
superimposed contours of simulated water vapor mixing ratio [g/kg]
by the MM5 mesoscale model along leg 1 from Germany to Iceland
on 13 May 1800 UTC. Lower panel: Backscatter ratio at 1064 nm
along same flight leg with superimposed contours of wind speed
and ECMWF tropopause (2 PVU surface). Dashed lines mark wind
directed out of, solid lines wind into the plane.
Fig 6: Upper panel: Water vapor measured by DIAL with superimposed contours of simulated water vapor mixing
ratio [g/kg] by the MM5 mesoscale model along leg 1 from Germany to Iceland on 13 May 1800 UTC. Lower
panel: Backscatter ratio at 1064 nm along same flight leg with superimposed contours of wind speed and ECMWF
tropopause (2 PVU surface). Dashed lines mark wind directed out of, solid lines wind into the plane.
Justagestreifen rausnehmen
-> Andreas
27
(a)
(b)
Fig 7: Simulated water vapor mixing ratio (g/kg, color shaded),
horizontal wind (m/s; wind flags: small barbs 5 m/s, long barbs 10
m/s) and potential vorticity (3.5, 2.5, and 1.5 PVU, thick, middle
and thin black lines) at 5 km altitude. (a) valid time 13 May 2002
1200 UTC (+12 h since MM5 initialization), (b) valid time 13 May
2002 2200 UTC (+22 h). The flight path of the DLR Falcon is
marked in green, the segment shown in Fig. 6 in red.
Fig 7:. Simulated water vapor mixing ratio (g/kg, color shaded),
horizontal wind (m/s; wind flags: small barbs 5 m/s, long barbs 10
m/s) and potential vorticity (3.5, 2.5, and 1.5 PVU, thick, middle
and thin black lines) at 5 km altitude. (a) valid time 13 May 2002
1200 UTC (+12 h since MM5 initialization), (b) valid time 13 May
2002 2200 UTC (+22 h). The flight path of the DLR Falcon is
marked in green, the segment shown in Fig. 6 in red.
kann man das ändern?
Küstenlinien und PV Konturen
unterscheidbar?
28
(a)
(b)
Fig 8: Ten days ECMWF backward trajectories arriving at 350 hPa
in a region between Scotland and Iceland on 14 May 2002 0000
UTC. (a) water vapor mixing ratio (mg/kg) and (b) potential
vorticity (PVU). Air parcel trajectories arriving close to the DLR
Falcon flight path are color shaded in the hemispheric as well as in
the close-up panels.
Fig 8: Ten days ECMWF backward trajectories arriving at 350 hPa in a region between Scotland and Iceland on 14
May 2002 0000 UTC. (a) water vapor mixing ratio (mg/kg) and (b) potential vorticity (PVU). Air parcel
trajectories arriving close to the DLR Falcon flight path are color shaded in the hemispheric as well as in the closeup panels.
29
Fig 9: (a) Simulated water vapor mixing ratio (g/kg; shaded),
horizontal wind speed (m/s; small barbs 5 m/s, long barbs 10 m/s)
and vertical wind (cm/s, red: positive, blue negative, increment 1
cm/s) at 308 K surface valid on 14 May 2002 0100 UTC (+13 h
since MM5 initialization). (b) vertical section along the red line in
(a). Additional fields are the section normal wind speed (m/s; gray
solid lines wind into the plane, gray dashed line wind out of the
plane), potential temperature (K, black solid lines) and the
tropopause marked by the gray 2-2.5 PVU ribbon. Water vapor
mixing ratios of 0.05, 0.10, and 0.15 g/kg are highlighted in green.
Fig 9: (a) Simulated water vapor mixing ratio (g/kg; shaded), horizontal wind speed (m/s; small barbs 5 m/s, long
barbs 10 m/s) and vertical wind (cm/s, red: positive, blue negative, increment 1 cm/s) at 308 K surface valid on 14
May 2002 0100 UTC (+13 h since MM5 initialization). (b) vertical section along the red line in (a). Additional
fields are the section normal wind speed (m/s; gray solid lines wind into the plane, gray dashed line wind out of the
plane), potential temperature (K, black solid lines) and the tropopause marked by the gray 2-2.5 PVU ribbon.
Water vapor mixing ratios of 0.05, 0.10, and 0.15 g/kg are highlighted in green.
30
(a)
(b)
(c)
Fig. 10: ECMWF backward trajectories arriving at 500 hPa along
the DLR Falcon flight track on 14 May 2002 0000 UTC. (a) water
vapor mixing ratio (mg/kg), (b) pressure (hPa), and (c) temperature
(K). Every full day is marked by a big every half day by small black
dot.
Fig. 10:. ECMWF backward trajectories arriving at 500 hPa along the DLR Falcon flight track on 14 May 2002
0000 UTC. (a) water vapor mixing ratio (mg/kg), (b) pressure (hPa), and (c) temperature (K). Every full day is
marked by a big every half day by small black dot.
31
(a)
g/kg
(b)
R1064
southwest
(c)
northeast
(d)
Fig 11: (a) Water vapor mixing ratio (g/kg, color shaded) measured
by the DIAL on 15 May 2002 0540–0650 UTC (blue segment) and
from 0840-1105 UTC (red segment). Superimposed simulated
water vapor mixing ratio (g/kg, thin black contour lines) valid at
0800 UTC (+20 h). (b) Aerosol backscatter (color shaded) and
simulated potential temperature (K, thin black solid lines every 2 K,
thick lines every 10 K) valid at the same times as in (a); the white
dots mark the locations where the DIAL measured a water vapor
mixing ratio of 0.3 g/kg. (c,d) Geopotential height (m, black lines)
and horizontal wind speed (m/s, color shaded and blue contour
lines) at 500 hPa (c) and 300 hPa (d).
Fig 11: (a) Water vapor mixing ratio (g/kg, color shaded) measured by the DIAL on 15 May 2002 0540–0650 UTC
(blue segment) and from 0840-1105 UTC (red segment). Superimposed simulated water vapor mixing ratio (g/kg,
thin black contour lines) valid at 0800 UTC (+20 h). (b) Aerosol backscatter (color shaded) and simulated potential
temperature (K, thin black solid lines every 2 K, thick lines every 10 K) valid at the same times as in (a); the white
dots mark the locations where the DIAL measured a water vapor mixing ratio of 0.3 g/kg. (c,d) Geopotential height
(m, black lines) and horizontal wind speed (m/s, color shaded and blue contour lines) at 500 hPa (c) and 300 hPa
(d).
32
(a)
(b)
Fig 12: Water vapor mixing ratio (g/kg, color shaded and green
isolines at 0.05, 0.10, and 0.15 g/kg), potential temperature (K,
black solid lines), section normal wind speed (m/s, gray solid lines
wind into, gray dashed lines wind out of the plane) and PV ribbon
2-2.5PVU indicating the dynamical tropopause. Vertical sections of
the mesoscale model simulations along the blue (a, valid at 15 May
2002 0600 UTC (+18 h since MM5 initialization)) and red (b, valid
at 15 May 2002 0800 UTC (+20 h)) flight tracks depicted in Fig. 9.
Fig 12:. Water vapor mixing ratio (g/kg, color shaded and green isolines at 0.05, 0.10, and 0.15 g/kg), potential
temperature (K, black solid lines), section normal wind speed (m/s, gray solid lines wind into, gray dashed lines
wind out of the plane) and PV ribbon 2-2.5PVU indicating the dynamical tropopause. Vertical sections of the
mesoscale model simulations along the blue (a, valid at 15 May 2002 0600 UTC (+18 h since MM5 initialization))
and red (b, valid at 15 May 2002 0800 UTC (+20 h)) flight tracks depicted in Fig. 9.
33
Fig. 13: Water vapor mixing ratio across the streamer between the
Azores and Spain on 17 June 2002 from 1530 – 1740 UTC.
Fig. 13: Water vapor mixing ratio across the streamer between the Azores and Spain on 17 June 2002 from 1530 –
1740 UTC.
34
(a)
(b)
Fig. 14: Vertical sections of water vapor (g/kg, shaded) along the
third leg of the eastbound ferry flights on 17 June 2002 1200 UTC
(a) and 1800 UTC (b). Other fields as in Fig. 12.
Fig. 14: Vertical sections of water vapor (g/kg, shaded) along the third leg of the eastbound ferry flights on 17 June
2002 1200 UTC (a) and 1800 UTC (b). Other fields as in Fig. 12.
35
Fig. 15: NOAA AVHRR satellite imagery on 17 June 2002 at 1423
UTC. (a) Infrared channel 5 and (b) visible channel 2. Images from
http://www.sat.dundee.ac.uk.
Fig. 15: NOAA AVHRR satellite imagery on 17 June 2002 at 1423 UTC. (a) Infrared channel 5 and (b) visible
channel 2. Images from http://www.sat.dundee.ac.uk.
36
Table 1. Water vapour DIAL data coverage and resolutiona
Flight
Time
Horizontal resolution [km]
Munich  Keflavik
13.05. 19:48-22-24
2.2 (H2O), 0.1-1 (aerosol) aa
Keflavik  Goose Bay
14.05. 00:36-02:52
2.3
Goose Bay  Montreal
15.05. 05:37–06:44
4.5
Montreal  Oklahoma
15.05. 08:40-10:15
2.3
Oklahoma  Bangor
16.06. 15:25-17:30
13.7
Bangor  Azores
17.06. 14:30-15:00
13.5
Azores  Santiogo
17.06. 17:52-18.45
13.6
Santiago  Munich
18.06. 06:54-07:54
4.7
a
The resolution is chosen such that the statistical error remains below 10%
aa
Vertical resolution of aerosol/cloud data is always 30 m, horizontal resolution is few 100 m
Vertical resolution top/bottom [km]
0.5 / 0.3 (H2O), 0.03 (aerosol) aa
0.5 / 0.4
1 / 0.4
0.5 / 0.2
2/1
1/1
1 / 0.5
0.5 / 0.5