Introduction
Cirrus clouds consist of ice particles and occur in the upper
troposphere and lower stratosphere at temperatures below -38 ∘C
. Their wide range of microphysical and
macrophysical properties affects the solar and terrestrial radiative budget
of the Earth's climate system. Depending on the microphysical properties
cirrus either warms or cools the layer below the clouds
. Among other factors, the ice particle shape
determines the cirrus radiative properties such as its albedo or spectral
radiative layer properties (e.g.
or ). Ice
particle shape and surface roughness may also cause biases in retrievals of
cirrus properties from satellite measurements.
To quantify the dependence of the cloud radiative forcing from cloud
properties, vertically separated observations of the cirrus microphysical and
radiative properties are needed. This can be realized by consecutive
measurements by one single aircraft or collocated observations by two
platforms. The first approach is problematic due to the (usually too large)
temporal displacement between the observations in, below, and above the
cloud. Collocated measurements using two coordinated aircraft were attempted,
for example during the Cirrus Regional Study of Tropical Anvils and Cirrus
Layers – Florida Area Cirrus Experiment (CRYSTAL-FACE) in 2002
, the Tropical Composition, Cloud and Climate Coupling
(TC4) mission in 2007 , and the Radiation–Aerosol–Cloud
Experiment in the Arctic Circle (RACEPAC) in 2014 .
However, as pointed out by and others, the exact vertical
collocation between the two aircraft with different speeds is problematic as
well. To minimize these collocation issues, towed sensor systems have been
applied in the past.
During the CARRIBA (Cloud, Aerosol, Radiation and tuRbulence in the trade
wInd regime over BArbados) project two platforms
connected by a cable to a helicopter were applied to obtain collocated
measurements of thermodynamic, turbulent, microphysical, and radiative
properties within clouds. showed that such observations
can be used to link cloud microphysical and radiative properties and estimate
the Twomey effect in shallow cumulus. However, such helicopter measurements
are limited to altitudes below 3000 m and therefore are not suited
for investigating cirrus.
introduced a new tandem measurement platform consisting of a
Learjet 35A research aircraft and an AIRcraft TOwed Sensor Shuttle (AIRTOSS),
which can operate in higher altitudes and velocities
(∼ 700 kmh-1). AIRTOSS is a sensor pod that is attached
under the right wing of the Learjet. When the Learjet reaches the measurement
area, AIRTOSS is released and towed by the aircraft via a steel wire. In 2007
a proof-of-concept campaign was conducted to evaluate the technical
feasibility, the flight safety, and the flight performance of AIRTOSS. In the
study of , AIRTOSS was equipped with a Cloud Imaging Probe
(CIP) to measure the microphysical properties of the clouds and two
navigation systems. At this time, the configuration of the tandem platform
was certified only to fly up to an altitude of 25 000 ft
(7620 m), which is below the altitude where most cirrus typically
occurs. show that turbulence as well as acceleration and
deceleration manoeuvres should be avoided to keep roll and pitch angles in a
range of ±3∘, which appears tolerable for irradiance measurements
(by definition related to a strictly horizontal receiving plane).
In this paper an advanced AIRTOSS platform including radiative and cloud
microphysical instruments is introduced, which is certified for higher
altitudes up to 41 000 ft (12 500 m). Technical details of
the redesigned AIRTOSS are presented in Sect. 2. Section 3 shows results of
collocated measurements in cirrus clouds with the Learjet 35A and AIRTOSS.
Two examples of collocated observations are discussed in Sect. 4.
Section summarizes the outcome and gives an overview of the
strengths and weaknesses of the improved AIRTOSS–Learjet tandem platform.
(a) Advanced AIRTOSS–Learjet tandem platform:
Learjet 35A with the sensor shuttle (called AIRTOSS) during a test flight.
The photograph was taken during the release of AIRTOSS. When AIRTOSS is fully
released, the distance between the Learjet and AIRTOSS is 3000 ft
(914 m). (b) Attached sensor pod under the left wing of the
Learjet with the mounted FSSP at the tip.
Different states of the AIRTOSS development process.
Panel (a) shows a perspective view of the position of the instruments
, including the Spectral Modular Airborne Radiation
measurement sysTem (SMART). Panel (b) shows the manufactured AIRTOSS
during a mission.
Masses of the different instruments and their accessories mounted
inside AIRTOSS.
Component
Mass
Explanation
(kg)
Front
CCP
9.10
Cloud Combination Probe (2–960 µm particle diameter)
Middle
Rechargeable battery
10.8
Power source for all instruments
Rear
Radiation optical inlet
0.24
4 pieces, top and bottom
Spectrometer (near infrared)
0.56
2 pieces, near infrared spectrometer (1000–2200 and 9–16 nm resolution)
Spectrometer (visible)
1.75
2 pieces, visible spectrometer (300–1000 and 3 nm resolution)
Peltier element
0.33
2 pieces
INS
0.02
Inertial navigation system
GPS sensor
0.04
Global Positioning System
Rosemount and sensors
0.60
Temperature and humidity measurements
ICH-TB
0.40
Temperature and humidity measurement electronics
Power supply BEP-5150C
0.75
Power supply (12 V, 5 V)
Computer
1.26
Data acquisition
Shutter
0.10
2 pieces for the SMART system
Shutter control
0.13
2 pieces to control the shutters
Technical development and properties of the AIRTOSS–Learjet tandem platform
The advanced AIRTOSS–Learjet tandem platform includes radiation sensors and a
sophisticated probe for cloud microphysical measurements. This set-up (see
Fig. ) was used during the AIRTOSS–Inhomogeneous Cirrus Experiment
(AIRTOSS–ICE) in spring and autumn 2013 above the North Sea and Baltic Sea.
Ten flights, five in spring (6–8 May 2013) and five in autumn
(29 August–5 September 2013), were performed during the AIRTOSS-ICE
campaign. The release of AIRTOSS was only possible under strict safety
regulations, and for this reason the measurement flights were only performed
in restricted military areas. In order to reach cirrus altitudes a full
formal aeronautical and aircraft certification had to be completed. After
this complex procedure the tandem platform consisting of the Learjet 35A and
the AIRTOSS was allowed to operate at altitudes up to 41 000 ft
(12 500 m).
The Learjet 35A research aircraft
In this study a Learjet 35A is applied (see Fig. a). It can reach a
maximum flight distance of 1700 NM (3148 km) and a maximum
altitude of 45 000 ft (13.7 km) and typically cruises at
speeds between 600 and 800 kmh-1. The aircraft is equipped with
a sensor pod mounted under the left wing (see Fig. b) and a winch
for AIRTOSS under the right wing. This additional freight limits the maximum
altitude (to ∼ 36 000 ft, 10 970 m) and endurance.
Radiative, meteorological, and microphysical instruments are mounted inside
AIRTOSS as well as on the fuselage of the aircraft.
AIRcraft TOwed Sensor Shuttle (AIRTOSS)
The original bird structure of AIRTOSS belongs to the shuttle case of the
type DO-SK6 and is manufactured by the European Aeronautic Defence
and Space Company (EADS). Primarily it is used as a flight target for
military training. The original case and the inner frame structure were
modified for implementing scientific instruments to perform measurements for
atmospheric research.
Specifications
A perspective view of the structure of AIRTOSS is shown in
Fig. a. The internal frame consists of high-strength aviation
aluminium EN AW-7075 and is separated into three sections. Structural
elements on the internal frame allow all sensors to be mounted inside
AIRTOSS, which has a length of 2.89 m and a diameter of
0.24 m. The middle section includes the eyelet, which connects the
AIRTOSS to the Learjet by a steel wire without electrical leads. A Cloud
Combination Probe (CCP) is located in the front section, and the rear part of
AIRTOSS contains mainly the radiation instruments. The original version used
the external body cover (made of glass-fibre reinforced plastic) as a
mounting point for additional payload. For the modified version, the body
cover is used only as a cover, which does not need a detailed strength
calculation and certification. It also makes it more convenient to access the
instruments and to recharge the replaceable battery after a measurement
flight.
The photograph in Fig. b was taken from an accompanying second
aircraft during a test flight for the Airworthiness Certification procedure.
Air brakes (red rectangles at the winglets) with different resistance
coefficients were mounted onto the winglets to compensate for the shape of
the asymmetric CCP and to keep the released AIRTOSS in a horizontal flight
position. More details about the air brakes and the associated flow
simulations are given in Sect. .
During transfer flights into the measurement areas, the unreleased AIRTOSS
stayed locked to the winch and was tilted such that it was closely held
underneath the wing to ensure a safe distance between sensors and ground
during the take-off and landing manoeuvres of the aircraft. The maximum
length of the steel wire between the winch and AIRTOSS is 4000 m.
During the AIRTOSS-ICE campaign the steel wire was only released to a length
of up to 914 m (3000 ft) to keep AIRTOSS under manageable
conditions within the borders of the relatively small restricted military
areas. Under these conditions and with an airspeed of 165 ms-1,
AIRTOSS stayed approximately 180 m below and 900 m behind the
Learjet. This horizontal displacement introduces a delay of about 5 s
between Learjet and the instantaneous location of AIRTOSS. During turns a
lateral displacement is also introduced. These data were rejected from the
collocated analysis presented here. The tare weight of the AIRTOSS case
without the instruments is 27.0 kg. After including the instruments
and the accessories, the total weight is 61.2 kg. To obtain the
position of the centre of gravity, a trim weight of 1.4 kg was added
to the rear section, resulting in a total weight of 62.6 kg. This is
still less than the maximum permitted total weight of 70 kg.
Table gives an overview of the masses of the included
instruments and accessories.
Energy consumption of the instruments
A rechargeable battery serves as the power source for the instruments mounted
inside AIRTOSS and is located in the centre of gravity in the middle section.
AIRTOSS reaches a continuous in-air operation time of 2 h. Safety
regulations only permit the instrumentation to be powered when AIRTOSS is
detached from the Learjet. The consequence of this constraint is that the
instruments must start to operate autonomously in an ambient temperature
between -30 and -50 ∘C. A suitable rechargeable battery for
these circumstances is the Smart VHF Modul 20S2P (24 V,
30 Ah) from SAFT batteries. To save power, several heaters of the
CCP instrument were deactivated. This was possible because the main purpose
of the heaters is to avoid icing and condensation at the optics of the
instrument, by flying through, e.g. mixed phase clouds. Only those from the
CCP Cloud Droplet Probe (CCP-CDP) instrument (see Sect. ) were
running during the measurement flights to keep the electronics under stable
temperature conditions. With these settings, all listed instruments in
Table consumed 213 W at a direct current of 28 V. The
rechargeable battery delivers 720 Wh, which leads to an operating
time of 3.5 h. However, considering that the CCP instrument turns off
below a voltage of 22.6 V in order to protect the lasers, the true
operating time of AIRTOSS is 2.5 h.
Instrumentation for microphysical cloud particle measurements
Different in situ instruments were installed on board AIRTOSS and the
Learjet sensor pod during the AIRTOSS-ICE campaign to collect information
about the microphysical properties of cirrus clouds. The CCP instrument contained in AIRTOSS is a modified version of the
instrument initially manufactured by Droplet Measurement Technologies (DMT,
Boulder, CO, USA). The position at the tip of AIRTOSS ensures that the
instrument is not influenced by proximity of aircraft structures, wings, and
fuselage, which sometimes cause issues when mounted at regular research
aircraft . To cover particles in a size range between 2 and
960 µm, the CCP contains a Cloud Imaging Probe greyscale
(CCP-CIPg) and a CCP-CDP. Shattering artefacts
are minimized by using specially designed
tips that are mounted on both instruments. Related
artefacts can be identified and excluded by using the recorded interarrival
time of each particle .
The CCP-CIPg records two-dimensional shadow images in a size range between 15
and 960 µm with a resolution of 15 µm. Computer
software, including special algorithms, is used afterwards to estimate cloud
particle parameters like maximum dimension diameter, concentration, and shape
.
In comparison to the CCP-CIPg instrument, the CCP-CDP detects particles in a
smaller particle diameter size range between 2 and 50 µm. The
instrument is based on forward light-scattering with a light collection angle
from 4∘ up to 12∘ and uses a laser diode with a wavelength of
658 nm. A sample area of 0.27±0.025 mm2 was estimated
by using a piezoelectric droplet generator laboratory set-up, similar to the
design of and . The accuracy and prior
measurements of the CCP-CDP instrument are shown in and
.
The Learjet was equipped with a Forward Scattering Spectrometer Probe (FSSP)
inside the sensor pod (Fig. b). This instrument was developed by
to measure particles in a size range between 2 and
47 µm in diameter and is a predecessor of the CCP-CDP
. Because the FSSP has neither mounted tips nor the
feasibility to exclude shattered particles by software algorithms, here it
was mainly used for testing purposes and as a cloud indicator during the
campaign. In the future it will be replaced with more advanced
instrumentation.
Spectral solar radiation measurements
To measure the up- and downward irradiance of a cirrus layer located between
the Learjet and AIRTOSS, both platforms were equipped with the Spectral
Modular Airborne Radiation measurement sysTem (SMART). For each radiation
component (upward/downward irradiance), SMART combines two Zeiss
spectrometers, each connected by fibre wires to an optical inlet mounted on
the top or at the bottom of AIRTOSS and the Learjet. The spectral range of
SMART is between 300 and 2200 nm with a resolution of 3 nm
for wavelengths below 1000 nm and 9–16 nm above
. The upward-looking irradiance sensor on
the Learjet was placed on a stabilized platform to keep it horizontally
aligned during the flights.
Due to the limited space inside AIRTOSS (see Fig. a), an active
horizontal stabilization of the radiation sensors could not be realized. For
this reason an inertial navigation system (INS) in combination with a Global
Positioning System (GPS) was used to record attitude and alignment angles.
These data were screened afterwards to identify and remove sections where
reliable measurements were not possible. A detailed analysis of the solar
radiation instruments, the measurements in cirrus, and the scientific results
of the AIRTOSS-ICE campaign are given in .
Trace gas instruments
Besides the radiation and microphysical instruments, the AIRTOSS–Learjet
tandem platform was equipped with a suite of instruments quantifying the
concentration of different trace gases.
The Fast Aircraft-Borne Licor Experiment (FABLE) was integrated on the
Learjet to detect the amount of carbon dioxide (CO2) at flight
altitude . Nitrous oxide (N2O) and carbon monoxide
(CO) were measured with the University of Mainz Airborne QCL-Spectrometer
(UMAQS; see for details).
Temperature and relative humidity measurements were made on the Learjet and
on AIRTOSS by the MOZAIC Capacitive Hygrometer (MCH), which belongs to the
Measurement of Ozone by Airbus In-Service Aircraft (MOZAIC) system. The MCH
uses a capacitative sensor and a Pt100 element to measure the relative
humidity and the temperature respectively. The accuracy is
±0.5 ∘C for the temperature measurement and ±5 % for the
detection of the relative humidity. Evaluation and measurement methods of
the MCH are described in detail in .
Water vapour measurements were taken by the Fast In-Situ Hygrometer
instrument (FISH) and the Selective Extractive Airborne Laser Diode
Hygrometer II (SEALDH-II). The FISH instrument is developed and operated by
the Forschungszentrum Jülich. It is based on Lyman-Alpha-Photometry and detects
water vapour in a range between 1 and 1000 ppmv with an uncertainty
of ±0.2 ppmv . SEALDH-II is operated by the
Physikalisch-Technischen Bundesanstalt, uses direct tunable diode laser
absorption (dTDLAS) and leads without any previous gas-based instrument
calibration to an absolute H2O concentration value. It operates in a
detection range between about 30 ppmv and roughly
40 000 ppmv with an accuracy of 0.35 % and a time resolution of
< 1 s .
Ozone (O3) measurements were performed on the Learjet by using a
UV photometry 42 M ozone analyser developed by Environment S.A.
This instrument detects the UV-absorption caused by O3 at a
wavelength of 254 nm in a measurement range between 0.9 ppb
(at 700 hPa) and 10 000 ppb with an uncertainty of 10 %
.
These instruments can be used for independent trace gas dynamics studies
(e.g. ) to better find the exact location of the
tropopause and identify tropopause folds as well as stratospheric influence
on uppermost tropospheric cirrus clouds (especially subvisual cirrus),
finding borders of air masses (e.g. the polar dome), among others.
(a) Front view of the AIRTOSS showing the asymmetrical
shape of the CCP instrument. (b) Air brake at one wing of the
AIRTOSS .
Flow simulations
With the incorporation of the CDP component of the CCP, the AIRTOSS overall
geometry has been altered in comparison with the design shown by
. Since the CDP is axially non-symmetric, the aerodynamic
properties of AIRTOSS were correspondingly modified with largely unknown
effects on alignment, attitude, and behaviour during flight.
Figure a shows a front view of AIRTOSS, which demonstrates the
asymmetry introduced by the CDP. To investigate these effects in regard to
ensuring stable flight conditions, detailed fluid flow simulations of the
AIRTOSS aerodynamics have been performed by
employing a computational fluid dynamics (CFD) methodology. We recall that, for
the formal Airworthiness Directives certification of the AIRTOSS, the
corresponding simulations resulting in the evolution of the forces and drag
coefficients were mandatory. The 3-D calculations were performed using the
AVL FIRE Thermo-Fluid Simulation Software (by AVL-List GmbH, Graz, Austria;
), employing a finite volume discretization method based on
the integral form of the general conservation law applied to polyhedral
control volumes. The turbulence model that was adopted is a four-equation,
eddy-viscosity-based turbulence model denoted by k-ε-ζ-f . Application of the “concept of elliptic
relaxation” allows for particular attention to the flow effects close to the
walls when approaching the AIRTOSS surface. In addition to the equations
governing the kinetic energy of turbulence k and its dissipation
rate ε, it solves transport equations for the quantity ζ,
representing the ratio ν2‾/k, and elliptic function f,
with ν2‾ denoting the scalar variable which behaves as the
normal-to-the-wall Reynolds stress component by approaching the solid wall.
Here, the ζ quantity represents a key parameter, as it models the
near-wall anisotropy influence on the relevant velocity scale in the
corresponding formulation for the turbulent viscosity. The
compound wall functions, blending between the integration up to the wall
with the standard equilibrium wall functions, were applied for the wall
treatment. They are especially advantageous for the high Reynolds number
flows, enabling well-defined boundary conditions irrespective of the position
of the wall-closest computational node. The numerical grid discretizing the
object surface and its surroundings consists of 12.7 million cells; this grid
represents an appropriate refinement of a coarser grid comprising 6.9 million
cells. The MINMOD bounded scheme combining the second-order accurate
schemes CDS (central differencing scheme) and LUDS (linear upwind
differencing scheme) is utilized for the discretization of the convective
transport and the conventional CDS scheme for the diffusive transport.
As a result detailed flow velocity fields were obtained as well as the
fields of turbulence quantities, drag coefficients, and aerodynamic forces.
The drag calculations were of specific concern because the connecting steel
wire only has a diameter of 2 mm. As an illustration
Fig. a shows the resulting total body pressure calculated by
the CFD simulation for flight conditions in the upper troposphere (i.e. here
25 000 ft, 7620 m) at aircraft speeds near 172 kt
(319 kmh-1). The highest total pressure regions occur in the
front of the CCP instrument and at the edges of the tail stabilizers in the
rear part of AIRTOSS. Regimes with a lower total pressure indicate flow
conditions associated with lower turbulence level in connection with the flow
acceleration. Figure b provides an example of the typical
velocity distribution around the AIRTOSS body. The deceleration zone, as
identified by in the region of the CCP measurement volume
corresponding to its front surface, can be well discerned on the left side of
the graph. The acceleration regions (red coloured areas) originating from the
streamline curvature effects follow. Figure c shows an
iso-surface of the turbulent kinetic energy with a value of
150 m2s-2 coloured by the velocity magnitude. Here the highest
speeds occur downstream of the CCP's measurement volume. As an overall result
of the CFD simulations, the horizontal tail stabilizers of the AIRTOSS body
were modified by affixing small air brakes to them in suitable positions such
that the asymmetry effects of the CDP were fully compensated (see
Fig. b). Accordingly, during level flights AIRTOSS moved
quietly in the flow, without disturbing oscillations, and the stable attitude
necessary for the radiation measurements was maintained well.
Flow simulations for flight conditions: (a) resulting total
body pressure, (b) velocity distribution around the AIRTOSS body,
(c) iso-surface of the turbulent kinetic energy with a value of
150 m2s-2 coloured by the velocity magnitude
.
MODIS high-resolution picture of the northern part of Germany, taken
at 11:00 UTC on 4 September 2013. Low stratus clouds are marked in white and
the observed cirrus is marked in blue. The yellow line indicates the flight
path of the Learjet.
Results from the cirrus measurements during AIRTOSS-ICE
On 4 September 2013, the northern part of Germany was located between a
high-pressure system with its centre above southern Germany and a low-pressure
system above Scandinavia. A related warm front in combination with cirrus
passed the measurement area above the Baltic Sea (Fig. ). The
cirrus deck was probed by the AIRTOSS–Learjet tandem platform between 09:10
and 09:40 UTC. The observations indicated that the cirrus was located at an
altitude between 8100 and 10 200 m with temperatures between -30
and -46 ∘C. Ice particle number concentrations of up to
1.4 cm-3 were found in several patches by the CCP in the upper
cloud layer (> 9000 m) where temperatures ranged below
-40 ∘C. As discussed by , these high ice
particle number concentrations only occur with vertical velocities higher
than 30 cms-1. Updraughts in warm fronts typically have vertical
speeds of less than 10 cms-1 and cannot
explain these high ice particle number concentrations. It appears that local
convective cells with stronger updraughts lifted droplets from lower cloud
layers to the cirrus altitude. As a result, homogeneous freezing in the
cirrus environment might have been initiated and would explain the high ice
particle number concentrations in the upper part of the cirrus.
Both panels show the flight path of the Learjet (red line) and the
flight path of AIRTOSS (dashed line) on 4 September 2013 with an overlay
showing colour-coded measurements of particle mean diameter (a) and
downward irradiance at 670 nm (b). The flight sections used
to calculate the mean particle diameter for specific legs are indicated
in (a).
Microphysical measurements
The flight paths of AIRTOSS and the Learjet are shown in
Fig. . The colour-coded line in Fig. a shows
the mean ice particle diameter measured by the CCP-CIPg. For each altitude a
mean particle number size distribution was calculated. The flight sections at
constant altitude that were used for the averaging are marked in
Fig. a. The legs were executed on constant altitude levels
and are longer in the lower part of the cloud to obtain appropriate counting
statistics for the optical particle instruments. Figure displays the
corresponding particle number size distributions and 2-D shadow images,
detected by the CCP for every single flight leg. The total particle number
concentration N is provided in Fig. a and shows a typical increase
with altitude from 0.26×10-2 cm-3 (8716 m) to
8.4×10-2 cm-3 (9939 m). Also, the particle size
corresponding to the maximum of the size distributions shifts with decreasing
altitude from 30 µm (9939 m) to 300 µm
(8716 m). The increase in particle diameter with decreasing altitude
is also obvious in the 2-D shadow images (Fig. b). Higher ice
particle number concentrations with small particle diameters in the upper
cloud layers and lower ice particle number concentrations with large particle
diameters in the lower cloud layers are typical for frontal, midlatitude
cirrus and result from the microphysical growth process during the formation
of the cirrus. As long as the relative humidity with respect to ice is
sufficiently high, the particles start to grow by water vapour diffusion, gain
mass, and sediment. This sedimentation process leads to a redistribution of
the ice particles inside the cirrus, with higher particle concentrations and
smaller cirrus particles at cloud top. Nevertheless, the irregular particle
shapes of the 2-D shadow images in the lower part of the cirrus indicate that
aggregation could also be a possible particle growth process. To analyse
whether diffusion or aggregation is the dominant process inside the observed cirrus,
similarly to , terminal velocities were calculated. This
is done by using the particle diameter Dp and the area ratio,
which is the area of the shadowed pixels (detected by the CCP-CIPg) divided
by the calculated particle area using the maximum dimension diameter
. As an example, a spherical (area ratio = 1) and a
horizontal-orientated column-shaped (area ratio = 0.25) ice particle with an
initialized diameter of Dp=200 µm are assumed. This
represents the measured conditions during Flight Leg 3 at an altitude of
9333 m (see Fig. ). For the spherical particle, a terminal
velocity of vt=91 cms-1 was calculated, while for
the horizontal-orientated columnar particle vt=14.5 cms-1 was estimated. With these estimated terminal fall
velocities, the particles would need 11 and 71 min, respectively,
until they reach the bottom layer of the cloud at an altitude of
8716 m. Following the discussion by , particles
with a number concentration of 5.8×10-2 cm-3 (Level 3
in Fig. ) need at least several hours before aggregation processes
occur, because the probability for collision is low. For this reason,
aggregation is unlikely, and diffusional growth seems to be the dominant
process for this particular cirrus observed during AIRTOSS-ICE.
Microphysical characteristics of the marked flight legs from
Fig. a. (a) Combined size distributions of the
CCP-CDP (red) and the CCP-CIPg (black) instrument mounted on the AIRTOSS.
With an increasing altitude, the maximum of the size distribution shifts to
smaller particle diameters. (b) Sample 2-D shadow images from every
single flight leg, recorded by the CCP-CIPg instrument. The different colours
represent the shadow intensity (grey > 35 %,
light blue > 50 %, dark blue > 65 %).
Downward spectral irradiance at 670 nm measured from the
(a) Learjet and (b) AIRTOSS as well as
the (c) number concentration (NC) measured on the AIRTOSS platform
with the CCP-CDP (2–50 µm) and the CCP-CIPg
(15–960 µm) instruments. The data were obtained at the highest
flight leg and measured on 4 September 2013, where AIRTOSS flew at an altitude
of around 9900 m. The vertical bars indicate the error of the
instruments, and the running average uses the boxcar smoothing algorithm with
10 repetitions.
Solar downward irradiance
In addition to the microphysical measurements, collocated measurements of
spectral solar radiation were performed during the cirrus event of
Sect. . Similar to Fig. a, a profile of the
spectral downward irradiance (at 670 nm wavelength) measured by SMART on
AIRTOSS and Learjet is given in Fig. b. The individual legs
were filtered for turns of both platforms which assures that only level
flight conditions were considered. Additionally, only legs flown in the same
direction and above the same locations were chosen to assure similar cloud
and surface conditions below the cirrus. In total, five legs with
simultaneous measurements on AIRTOSS and the Learjet are available with
larger vertical separation in the cirrus and less separation at cloud top and
above. The impact of the cirrus on the downward irradiance is most obvious in
the two lower legs where the radiation is attenuated by the cirrus. The
attenuation is highly variable due to the horizontal heterogeneity of the
cirrus. However, both sensors on AIRTOSS and Learjet show almost the same
pattern, illustrating the collocation of the measurements. The similarity in
the two data sets also results from the small vertical displacement of Learjet
and AIRTOSS of less than 200 m. During the higher flight legs, the
attenuation of downward irradiance by the cirrus is significantly lower. In
the third leg, only AIRTOSS measurements are slightly affected by the cirrus,
while the Learjet already observed clear-sky conditions. Above the cirrus,
the downward irradiance is almost constant over the entire legs, indicating
clear sky for both platforms.
Discussion
Two cases are selected to illustrate the potential of the collocation of
measurements achieved by the AIRTOSS–Learjet tandem platform. Due to the
different instruments that are operated on AIRTOSS and Learjet, different combined
analysis of data are possible. Beside combining in situ and radiation
measurements the simultaneous radiation measurements on both platforms
can also be analysed jointly.
Collocation of microphysical and radiative properties
Figure shows a time series of downward spectral irradiance at
670 nm wavelength measured from the Learjet (Fig. a) and AIRTOSS
(Fig. b) during a flight leg observed on 4 September 2013 between
09:35 and 09:39 UTC, when AIRTOSS was operated at an altitude of around
9900 m. In addition, Fig. c shows the detected number
concentration of the CCP-CDP and the CCP-CIPg. The cloud particle number
concentrations above zero were detected within two sections of the flight leg
and indicate that AIRTOSS penetrated two cirrus filaments at the top of
the cirrus layer. The downward irradiance has been constant for most of the
flight leg, indicating clear-sky conditions without attenuation of the
incoming solar radiation. The strongest deviation from the clear-sky
conditions was found at about 09:38:05 UTC, when the irradiance shows a rapid
decrease for both platforms.
(a) Profiles of vertical upward and downward broadband
irradiance measured on AIRTOSS and the Learjet. The bars indicate the
standard deviation of the irradiance along the individual flight legs.
(b) Solar heating rates calculated from the irradiance profile using
either a single platform or the collocated measurements. The grey area
indicates the cirrus layer as indicated by the CCP.
This coincides with higher values in the particle number concentration
measurements. The increasing number concentration indicates that AIRTOSS is
located in a thicker part of the sampled cloud and certainly the cloud top is
above AIRTOSS. As the Learjet measurements are located closer to cloud top,
the effect here is smaller compared to the AIRTOSS observations. At cloud
edges also an increase of the irradiance can occur due to three-dimensional
radiative effects . For the smaller cloud observed at the
beginning of the leg (09:35:45–09:36:40 UTC), only the downward irradiance
measured by AIRTOSS shows variation, while the downward irradiance measured
by the instruments on the Learjet remains almost constant. At this time only
AIRTOSS was located inside the cirrus while the Learjet flew above cloud top
and consequently only the downward radiation in the altitude of AIRTOSS was
reduced.
Such constellations are well suited to investigate the interaction of cloud
microphysical and radiative properties as demonstrated by
for shallow cumulus. However, the approach by for
analysing the collocated number concentration and cloud remote sensing works
only if the radiation measurements are taken well above the cloud. In the
case of the AIRTOSS–Learjet tandem this would limit the analysis to the
uppermost cirrus layer. However, when operating radiation measurements on both
platforms, the cloud optical layer properties can be derived as presented by
. Using the collocation, cloud layers well inside the
cloud can also be analysed.
Vertical profile of solar heating rates
The spectral irradiance measurements were integrated to broadband quantities
and averaged for the individual horizontal legs as indicated in
Fig. . To make measurements comparable, the change of the
solar position in between measurements of the different legs was taken into
account by normalizing the irradiance to observations from the uppermost
level. Figure a shows the corresponding vertical profiles of
upward and downward broadband irradiance measured on AIRTOSS and Learjet. The
horizontal bars indicate the standard deviation of the irradiance along the
individual flight legs.
The upward irradiance varies significantly with altitude albeit without
showing a regular pattern. This is likely caused by slight changes of the
flight track and the in-cloud situation, mainly the
presence of a low stratus cloud below the cirrus (see Fig. ).
The standard deviation of upward irradiance is higher in the upper three
legs, while the two lower legs show less variability when the sensors are
located well below cloud top. Assuming, that along the flight leg the low
stratus is homogeneous with respect to the field of view of the irradiance
optical inlet, these higher standard deviations are mainly caused by the
spatial variability of the cirrus. The cirrus is located vertically closer to
the irradiance sensor and therefore smaller horizontally inhomogeneities are
resolved by the measurements.
The profile of downward irradiance also indicates the presence of cirrus.
While above cloud top the values remain vertically constant and show only a
small standard deviation, larger variability and a decrease of the downward
irradiance are observed when the instruments enter the cloud. Upward and
downward irradiance F↓ and F↑ at two different
altitudes, z1 and z2 are used to calculate the effect of the radiation
field on the local temperature change in terms of heating rates at a certain
altitude z=1/2⋅(z1+z2). The heating rate ∂T/∂t|z
in units of Kday-1 within the layer is derived following
(Eq. 9.66):
∂T∂tz=1ϱ⋅cp⋅∂Fnet(z)∂z≈1ϱ⋅cp⋅Fnet(z2)-Fnet(z1)z2-z1≈1ϱ⋅cp⋅F↓(z2)-F↓(z1)-F↑(z2)-F↑(z1)z2-z1.
Figure b shows profiles of ∂T/∂t|z
derived in two different ways. Assuming only a single aircraft is available,
the solar heating rates can be calculated by the irradiance profile measured
by this single aircraft, either AIRTOSS (red circles) or Learjet alone (blue
circles). Having the combined collocated measurements of both AIRTOSS and
Learjet, heating rates can additionally be derived along each horizontal leg
(black circles). The heating rate profiles obtained for the investigated
cirrus significantly differ depending on the chosen method. To interpret
these differences, uncertainties of the heating rates were calculated for
both approaches. An uncertainty of 6 % in the radiometric calibrations
was assumed, which directly propagates into the calculated heating rates
(Eq. 1) as all sensors are calibrated identically. All remaining
uncertainties of the irradiance are estimated with 0.5 %. For the
single-aircraft approach the irradiances are always measured with the same
system. This reduces the impact of the remaining uncertainty to contributions
of the two net irradiance only. In the collocated approach, two independent
systems are used and all four irradiance measurements contribute to the
overall uncertainty. Additionally, the distance z2-z1 influences the
accuracy of the heating rate. Due to the geometry and the flight altitudes,
this distance differs for both approaches. Larger distances between the two
measurements provide more precise results. While z2-z1 amounts to about
200 m for the collocated approach, determined by the length of the
wire between AIRTOSS and the Learjet, z2-z1 of the single-aircraft
approach depends on the altitudes of the legs and is typically larger
(500 m at cloud bottom and 300 m at cloud top). Overall, the
uncertainty of the heating rate estimates derived from the collocated
approach are theoretically expected to be significant larger than for the
single-aircraft approach. However, although the profiles using only AIRTOSS
and only Learjet data agree with each other, the profiles show large scatter
with heating rates ranging from -13 to +33 Kday-1. These
unrealistic heating rates mainly result from changes in the upward irradiance
between two individual flight legs. As the legs are not perfectly collocated
and a low stratus layer did change its location below the cirrus during a
flight level change (∼ 2 min temporal separation), the data set is not
consistent and leads to incorrect heating rate estimates.
By contrast, the collocated data set does not suffer from changing conditions
below the cirrus as both sensors always observe the same scene at the same
time. Consequently, the heating rate profile in Fig. b does
show a smoother and more realistic pattern with values always ranging between
0 and 6 Kday-1, which are typical values for a thin cirrus.
This improvement in calculating heating rates illustrates the benefit of
collocated irradiance measurements. However, the derived heating rates still
do not represent theoretical results as provided, for example, by
and . For subvisible and optically
thin cirrus, they calculated heating rates in the range of
0.2–0.5 Kday-1. These higher values might result from the
higher optical thickness, τ=0.6, of the cirrus observed by AIRTOSS or be
caused by horizontal inhomogeneities of the observed cirrus leading to
horizontal photon transport as discussed by .
Conclusions
The advanced AIRTOSS–Learjet tandem platform was
applied during the AIRTOSS-ICE campaign to perform collocated measurements of
cirrus cloud properties. A combination of the Learjet and AIRTOSS, both
equipped with radiation and microphysical in situ instruments, allowed for
measurements of cirrus properties at different altitudes using just one
aircraft. The new certification for the AIRTOSS–Learjet tandem platform
enabled us to probe cirrus at altitudes up to 36 000 ft
(10 970 m). The campaign showed that collocated measurements with
the revised AIRTOSS–Learjet tandem platform are feasible. This is
demonstrated by combining the microphysical and radiative measurements and,
as an illustrative example, by deriving solar heating rates. Further results
are presented by in a closure study, which combines in
situ cloud and radiative measurements to quantify the impact of ice crystal
shape, effective radius, and optical thickness on cirrus radiative forcing.
A case study is presented in which AIRTOSS-ICE measurements are used to derive
vertical profiles of cloud microphysical and radiative properties. Using the
profiles of upward and downward irradiances, it is shown that solar heating
rates can be estimated with improved accuracy when collocated measurements
are applied, instead of using a single platform. Despite the expected higher
uncertainties introduced by the measurement errors from two independent
measurement systems, the collocated observations resulted in a more realistic
profile of solar heating rates as these are not affected by changes of the
radiation field below the observational altitude (e.g. inhomogeneous surface
albedo, lower cloud layers). Observations performed with a single aircraft
strongly depend on stable conditions between consecutive flight legs and,
therefore, are subject to serious uncertainties in derived profiles of solar
heating rates.
However, AIRTOSS-ICE also showed the limits of the collocated measurement
set-up. The investigated cirrus had a thickness of more than 200 m,
which is larger than the distance between Learjet and AIRTOSS during the
conducted measurement example. This did not allow for the radiative
instruments to measure concurrently with AIRTOSS below the cirrus layer and
with the Learjet above, which would have been needed to derive the cirrus
radiative layer properties . The short distance between
both platforms resulted in only small differences in the upward and downward
irradiances measured on AIRTOSS and the Learjet for this sampling example. An
increase of the vertical distance beyond 200 m is not easy to
achieve. It would require a longer steel wire and/or a slower aircraft as
well as larger areas where such flights are permitted. For clouds with a
larger vertical extent, two single aircraft could be a better choice. It
certainly depends on the scientific goals and instrumentation as to whether
or not the AIRTOSS–Learjet tandem platform is the appropriate choice.
With respect to microphysical inhomogeneities, the vertical separation of
200 m between both platforms is sufficient for cirrus studies. What
would be required additionally are microphysical in situ instruments with
overlapping measurement characteristics, or, ideally, two identical
instrument sets on both platforms. To perform microphysical measurements with
a higher temporal resolution, the implementation of holographic instruments
is also an attractive alternative. These instruments have a larger sample
volume of up to 305 cm3 , which is much higher
than the sample volume of the CCP-CDP (45 cm3 for an aircraft
velocity of 165 ms-1). Furthermore, the integration of trace gas
instruments inside AIRTOSS and the Learjet could be used, e.g. for collocated
trace gas measurements in the vicinity of the tropopause layer, the edges of
tropopause folds, and streamers. To study different atmospheric conditions or
to obtain better statistics of cirrus cloud, the operation of the
AIRTOSS–Learjet tandem platform in other regions, outside of military
restricted areas, remains a significant challenge. This could be accomplished
in less populated areas, such as the polar regions, remote areas of the
oceans, and rainforests.