ACPAtmospheric Chemistry and PhysicsACPAtmos. Chem. Phys.1680-7324Copernicus PublicationsGöttingen, Germany10.5194/acp-16-933-2016Investigation of the adiabatic assumption for estimating cloud micro- and macrophysical properties from satellite and ground observationsMerkD.merk@tropos.deDenekeH.https://orcid.org/0000-0001-8595-533XPospichalB.https://orcid.org/0000-0001-9517-8300SeifertP.https://orcid.org/0000-0002-5626-3761Leibniz Institute for Tropospheric Research, Leipzig, GermanyLeipzig Institute for Meteorology, Leipzig, GermanyD. Merk (merk@tropos.de)26January201616293395226January201524February201521December201524December2015This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://acp.copernicus.org/articles/16/933/2016/acp-16-933-2016.htmlThe full text article is available as a PDF file from https://acp.copernicus.org/articles/16/933/2016/acp-16-933-2016.pdf
Cloud properties from both ground-based as well as from geostationary passive
satellite observations have been used previously for diagnosing aerosol–cloud
interactions. In this investigation, a 2-year data set together with four
selected case studies are analyzed with the aim of evaluating the consistency
and limitations of current ground-based and satellite-retrieved cloud
property data sets. The typically applied adiabatic cloud profile is modified
using a sub-adiabatic factor to account for entrainment within the cloud.
Based on the adiabatic factor obtained from the combination of ground-based
cloud radar, ceilometer and microwave radiometer, we demonstrate that neither
the assumption of a completely adiabatic cloud nor the assumption of a
constant sub-adiabatic factor is fulfilled (mean adiabatic factor
0.63 ± 0.22). As cloud adiabaticity is required to estimate the cloud
droplet number concentration but is not available from passive satellite
observations, an independent method to estimate the adiabatic factor, and
thus the influence of mixing, would be highly desirable for global-scale
analyses. Considering the radiative effect of a cloud described by the
sub-adiabatic model, we focus on cloud optical depth and its sensitivities.
Ground-based estimates are here compared vs. cloud optical depth retrieved
from the Meteosat SEVIRI satellite instrument resulting in a bias of -4 and
a root mean square difference of 16. While a synergistic approach based on
the combination of ceilometer, cloud radar and microwave radiometer enables
an estimate of the cloud droplet concentration, it is highly sensitive to
radar calibration and to assumptions about the moments of the droplet size
distribution. Similarly, satellite-based estimates of cloud droplet
concentration are uncertain. We conclude that neither the ground-based nor
satellite-based cloud retrievals applied here allow a robust estimate of
cloud droplet concentration, which complicates its use for the study of
aerosol–cloud interactions.
Introduction
Low-level liquid clouds are found in many areas around the globe and play an
important role in the energy balance of the Earth. Their microphysical and
optical properties are strongly influenced by aerosol particles that act as
cloud condensation nuclei. Twomey (1974) first postulated the effect of an
increased aerosol number concentration in clouds on the radiative budget,
commonly referred to as the first indirect aerosol effect, as a climatically
relevant process. The quantification of such aerosol indirect effects remains
one of the main uncertainties in climate projections . If
the liquid water content as well as the geometrical depth of the cloud are
considered constant, a higher aerosol load results in an enhanced cloud
albedo. This effect is observed in particular by means of ship tracks that
form in marine stratocumulus cloud decks, e.g..
Cloud quantities that are typically used to calculate aerosol–cloud
interactions (ACI), are the cloud droplet number concentration (Nd)
and cloud geometrical depth (H). noted that a
15 % change in H can have a similar effect on cloud albedo as a
doubling of Nd. proposed to investigate a column
Nd which is the integral of Nd over H.
While remote sensing observations from ground are always local column
measurements, passive satellite observations from, e.g. SEVIRI (Spinning
Enhanced Visible and Infrared Imager) or MODIS (Moderate Resolution Imaging
Spectrometer), show a good tradeoff in terms of spatiotemporal coverage and
are therefore suitable to investigate ACI on a larger scale. Active satellite
sensors on the other hand, such as the cloud profiling radar onboard CloudSat
or the Cloud-Aerosol-Lidar with Orthogonal Polarization
(CALIOP) on-board CALIPSO Cloud-Aerosol Lidar and Infrared Pathfinder
Satellite Observation, are able to provide vertically resolved
cloud observations along their tracks and can be used to investigate aerosol
effects on cloud properties, e.g.. These lack
highly-resolved temporal coverage and have a smaller scanning swath than
passive sensors onboard polar-orbiting satellites. Despite their coarser
spatial resolution, geostationary satellite observations benefit from the
high temporal coverage of up to 5 min in conjunction with a high spatial
coverage. This can be considered as an advantage for the determination of
large-scale ACI, since the full daily cycle can be obtained and contrasted to
ground-based observations.
If entrainment in clouds leads to a deviation from a linear increasing liquid
water content, i.e. sub-adiabatic clouds, the first aerosol effect is not
easily observed . To obtain key quantities from passive
satellite observations, the sub-adiabatic cloud model is usually applied,
e.g.. Therefore obtaining
cloud adiabaticity is important for the investigation of aerosol–cloud
interactions. The combination of ground-based ceilometer and cloud radar is
able to provide reliable detection of cloud geometric borders
. Nd
from ground-based observations can be retrieved from radar–radiometer
measurements , observations including lidar measurements
, or solar radiation measurements
. To derive Nd from radar–radiometer
observations recently suggested a condensational growth
model taking the vertical velocity into account and allowing small variations
of Nd with height, while it is assumed vertically constant in most
other studies. Due to the under-constrained nature and assumptions made in
such retrieval methods, substantial differences for the microphysical
properties may occur, as pointed out by , who intercompared
several ground-based retrieval methods for one case study.
showed that the cloud optical depth is less sensitive to
the assumptions required in radar–radiometer retrieval approaches and might
be considered as an alternative key quantity.
As a consistency check, we contrast key quantities from ground-based remote
sensing using a ceilometer, a microwave radiometer and a 35-GHz cloud radar
at Leipzig, Germany (51.35∘ N, 12.43∘ E) and at
Krauthausen, Germany (50.897∘ N, 6.46∘ E) with observations
from SEVIRI onboard the geostationary satellite Meteosat Second Generation
(MSG). Those ground-based instruments are operated in the framework of
Cloudnet and ACTRIS (Aerosols, Clouds and Trace gases
Research InfraStructure Network). To our knowledge such evaluations from the
SEVIRI instrument for key parameters have been rarely carried out, e.g.
in. Thereby, we discuss the uncertainties introduced by
required assumptions when cloud microphysical properties are retrieved, and
the effect of different spatiotemporal resolution. As the sub-adiabatic
cloud model is a key concept for the retrievals discussed in this study, we
aim to quantify cloud adiabaticity using the available observations.
The paper is structured as follows. In
Sect. we introduce the sub-adiabatic
model, relevant for the satellite-based retrieval of key parameters, as well
as the retrieval methods from ground. Afterwards we describe the instruments
and data processing tools used within this study in Sect. . In
Sect. cloud adiabaticity is investigated. Subsequently
we contrast important key properties relevant for ACI from SEVIRI and
LACROS (Leipzig Aerosol and Cloud Remote Observations System) and discuss uncertainties from both perspectives
(Sect. ). Finally, a conclusion and outlook is given
in Sect. .
Cloud retrieval methods using the sub-adiabatic cloud model
In this section we present the theory of the sub-adiabatic cloud model and
retrieval strategies for ground-based instruments as well as passive
satellite observations.
For a moist rising air parcel we assume that the liquid water content
qL(z) increases linearly with height :
qL(z)=fadΓad(T,p)z.
Γad(T,p) is the adiabatic rate of increase of liquid water
content. The adiabatic factor fad can be understood as a
reduction of liquid water due to evaporation triggered by the entrainment of
drier air masses, which leads to fad<1 (sub-adiabatic).
Integrating the liquid water content with height yields the liquid water path
(QL). ACI are usually studied as changes in cloud properties and
radiative effects for a constant QL. Therefore we will express all following physical quantities as
function of given QL. Observing H in combination with
QL, and knowing Γad(T,p), fad can be
calculated as follows:
fad(QL,H)=2QLH2Γad(T,p).
The geometrical depth for adiabatic clouds is obtained by resorting to this
equation:
H(QL,fad)=2QLfadΓad.
The equivalent mean volume droplet radius (rV) in a cloud depends
on Nd and qL:
rV=3qL4πρwNd3.
In the following we assume homogeneous mixing and introduce the effective
radius re. re is defined as the third over the second
moment of the droplet size distribution and is typically
retrieved in remote sensing. re is related to the mean volume
radius introducing a factor k2 that depends on the width of the droplet
size distribution (DSD).
re=k2-13rV
Typical values for k2 are 0.67 and 0.8 for marine and continental clouds
, respectively. More details on the factor k2 for the
assumed gamma-size distribution can be found in Appendix A.
By substituting rV with re in Eq. (),
we yield re representative for the uppermost cloud layer:
re(QL,fad,Nd)=18fadΓadQL64πρwk2Nd3.
To study the microphysical response of aerosols on cloud microphysics with
remote sensing techniques, together with re the optical depth
τ is often used since both can be easily derived from, e.g. passive
satellite observations ().
τ in the sub-adiabatic model can be expressed as a function of
QL and re:
τ=9QL5ρwre.
Using this equation QL can be derived from passive satellite
observations. H can be also derived from
H=10ρwτre9fadΓad.
By substituting re from Eq. (), we yield
τ as a function of QL, Nd and fad:
τ(QL,fad,Nd)=94πk2Nd3QL56518ρw4fadΓad6.
From this equation, Nd from passive satellite observations can be
calculated as follows:
To retrieve τ and re from the given ground-based observations,
Nd is substituted in Eq. () applying a
radar–radiometer retrieval approach, e.g.see Appendix A:
Nd(QL,Z)=9k6QL22π2ρw2∫CBHCTHZ(z)dz2.
Then we find τ and re for given QL to depend on the
width of the DSD (k2, k6), fad and the integrated radar
reflectivity profile (∫Z(z)dz). It follows that τ∝(k2k6)13 and re∝(k2k6)-13. Therefore, it would be preferable to derive Nd
(zeroth moment) from the 2nd and 3rd moment (τ and QL) rather
than from the 3rd and 6th moment (QL and Z). This is the main
reason why Nd in our retrieval is very sensitive to the width of
the DSD. The other method would require observations of τ, e.g. from a
multi-frequency rotating shadowband radiometer (MFRSR).
Overview of assumptions made for the sub-adiabatic cloud model
applied to derive Nd and H in literature studies. The table lists
the values chosen for Γad, fad (calc. refers to
explicitly calculated values from additional data) and k2. The table is
sorted by publication year starting with the oldest one.
While in this study homogeneous mixing is assumed, in general two extremes of
mixing processes can be considered :
(a) homogeneous mixing, where Nd stays constant, but the droplet
radius (rV) is changed due to evaporation, (b) inhomogeneous
mixing, where the number of droplets change (dilution of whole droplets), but
the droplet radius profile is unchanged. In nature, a mixture of both
processes may occur . Without entrainment, we find
fad=1 (adiabatic clouds). The assumption of homogeneous mixing is
supported by observations from, e.g..
fad in this study is considered as representative for the full
vertical cloud depth. For such an adiabatic factor fad a range of
[0.3, 0.9] is seen as common .
Different values for k2, Γad and fad in
Eq. () have been considered in previous studies using
passive satellites (Table )
due to various reasons (e.g. different cloud regimes, continental vs.
maritime). Often even adiabatic clouds are assumed (fad=1) in the
retrieval process, e.g..
DataInstruments and retrievals
Satellite data from SEVIRI is used, which provides
12 spectral channels covering the visible, the near infrared, and the
infrared spectrum. The channels used here have a nadir resolution of
3 km× 3 km, which decreases towards the poles and
is about 4 km× 6 km over our region of interest
(central Europe). In this study we use the 5 min temporal resolution data
from the Rapid Scanning Service (RSS). The SEVIRI
radiances in the different channels are used as input for the Nowcasting
Satellite Application Facility (NWC SAF) algorithm which
provides a cloud mask, cloud top height (CTH), and cloud classification. To
obtain the cloud mask, different multispectral tests using SEVIRI channels
are applied in order to discriminate cloudy from cloud-free pixels. The cloud
top height for low, liquid clouds is obtained by using a best fit between
measured brightness temperatures in the 10.8 µm channel and
simulated values using the RTTOV radiative transfer model
applied to atmospheric profiles from the ECMWF (European
Centre for Medium-Range Weather Forecasts) numerical weather prediction (NWP)
model.
The NWC SAF cloud mask is used in order to derive cloud phase, cloud optical
depth, and effective radius with the KNMI (Royal Netherlands Meteorological
Institute) cloud physical properties (CPP) algorithm ,
developed in the context of the satellite application facility on climate
monitoring CM SAF, . Using a channel in the visible
spectrum (0.6 µm) together with an absorbing channel in the near
infrared (1.6 µm) , the CPP algorithm retrieves
τ as well as re representative for the uppermost cloud part.
As this method relies on solar reflectance channels, it is applied only
during daytime.
Also data from MODIS is used within this study. MODIS is an imaging
spectrometer onboard the satellites Terra (descending node) and Aqua
(ascending node) which probe the Earth's atmosphere from a polar orbit that
results in one daytime overpass per satellite per day over the region of
interest. MODIS measures in 36 bands in the visible, near-infrared, and
infrared spectrum, with some bands having a spatial resolution of up to
250 m. The cloud physical properties are
retrieved in a similar manner as for SEVIRI, but at 1 km spatial
resolution using the channels 0.6 µm (band 1, over land) and
2.1 µm (band 7, over land and sea). In addition, re
retrievals are available using the channels at 1.6 µm (band 6) and
3.7 µm (band 20) together with band 1. Note that band 6 on the Aqua
satellite suffers from a stripe-problem . In this study MODIS
collection 6 is used for the retrieved τ and re.
The ground-based remote sensing instruments of the Leipzig Aerosol and Cloud
Remote Observations System (LACROS) comprise a 35-GHz MIRA-35 cloud
radar, a HATPRO (Humidity And Temperature Profiler) microwave radiometer, and
a CHM15X ceilometer, which are used also for field campaigns. All instruments
are operated in a vertically pointing mode. The raw measurements are
processed with the Cloudnet algorithm package . The
output data is available at a unified temporal resolution of 30 s and
a vertical grid of 30 m. Cloudnet uses further information such as
temperature and pressure profiles from a NWP model (here: COSMO-DE). In this
study we use the attenuation-corrected radar reflectivity Z from the cloud
radar, QL obtained from the microwave radiometer, as well as the
cloud base and top height retrieved from ceilometer and cloud radar,
respectively. The vertical Doppler velocity from the cloud radar is also
utilized. Furthermore Cloudnet provides a target classification applying
a series of tests to discriminate cloud phase, drizzle or rain, and aerosols
or insects.
Time series of radar reflectivity Z(z) (in dBZ) and cloud
boundaries for the four cases listed in Table ;
(a) 27 October 2011, (b) 21 April 2013,
(c) 1 June 2012, (d) 27 September 2012. Cloud borders are
shown as detected by Cloudnet with black dots and by SEVIRI using NWC SAF in
orange dots, and MODIS in blue dots. Sample profiles of Z(z) are shown at
different times during each case.
Data selection
For this study, we use a 2-year period covering 2012 and 2013. We focus on
ideal cases to gain a better understanding of the microphysical processes
within the cloud. In order to avoid uncertainties caused by inhomogeneous
cloud scenes, such as multi-layer clouds, we consider single-layer cloud
systems which are entirely liquid and non-drizzling as ideal.
Cases used within this study sorted by date. The minimum cloud base
height (CBH) and the maximum cloud top height (CTH) of the liquid cloud layer
investigated are presented together with the temporally averaged
inhomogeneity parameter (χ) as in calculated from
τ of the ±2 surrounding SEVIRI pixels for each observation time.
Furthermore the category for each case is listed.
Cloud profiles as observed from the ground are filtered according to the
following conditions.
There is no occurrence of drizzle/rain in Cloudnet's target classification (and no
drizzle/rain in the two nearest neighbour profiles allowed.)
Values of QL are between 25 and 400 g m-2.
The lower limit is due to typical instrument uncertainty of the microwave
radiometer and the upper limit due to typical thresholds for drizzle
occurrence .
The liquid cloud layer must be situated between 300 and 4000 m above ground.
The cloud geometrical depth is between 100 and 2000 m.
There are no ice cloud layers within the first 4000 m above ground is present. Thin ice
cloud layers above are excluded from calculation of H. The microwave
radiometer is not sensitive to ice, so that QL should not be
affected.
No vertical gaps in the cloud layer are present.
Zmax<-20 dBZ within the cloud profile to avoid occurrence of drizzle .
The comparison of optical and microphysical properties between ground-based
and MODIS and SEVIRI is only applicable under daytime conditions. Thereby, we
have to consider the different spatial and temporal resolution, as well as
the different viewing zenith angle on the cloudy scene. For SEVIRI a parallax
shift occurs at higher latitudes. The satellite viewing zenith angle for
Leipzig is 58.8∘. Within this study the average cloud top height is
between 1 and 3 km (see Table ). This would
result in a horizontal displacement of max. 5 km. did
find a significant difference only for inhomogeneous clouds considering
parallax correction. Also taking into account the spatial resolution of
SEVIRI over central Europe of 4 km× 6 km, we
decided to neglect the parallax correction for our study, instead we consider
surrounding pixels. For SEVIRI a field of 3 × 3 pixels (case
studies), and 5 × 5 pixels (longer-term statistics) centred on the
ground site is used and spatially averaged.
We will furthermore present four hand-selected cases to highlight specific
problems more closely. For the four case days, we calculate the spatial
inhomogeneity parameter following , using the
3 × 3 SEVIRI pixel field, which can be interpreted also in terms of
temporal inhomogeneity (χ) if advection of clouds over a fixed location
is considered:
χ=exp(lnτ‾)τ‾.
A short overview of the case characteristics is given in
Table . The cloud boundaries are shown along with the
Z profile in Fig. . The synoptic conditions for the cases
are as follows. A high pressure system dominates the synoptic weather pattern
on 21 October 2011 (Fig. a). The temperature at the
850 hPa pressure level over Leipzig is around 5 ∘C.
Therefore the stratocumulus cloud layer that is observed between 10:30 and
13:00 Z consists entirely of water droplets. Its geometrical depth increases
in the beginning of the observation period. The weather pattern on
21 April 2013 (Fig. b) is quite similar with the high pressure
influence being stronger. The temperatures at the 850 hPa pressure
level are slightly positive. During the whole observation period at
Krauthausen a closed cloud deck is visible. The ground-obtained cloud top
height shows only small variability, while the ceilometer-derived cloud base
is more inhomogeneous during the beginning of the observation period. A thin
overlying cirrus cloud deck can be observed around 10:00 Z and between
11:00–12:00 Z. An upper-level ridge covers central Europe on 1 June 2012
(Fig. c), but the area around Leipzig is also influenced by a
surface low. Temperatures at 850 hPa lie around 10 ∘C. The
stratocumulus cloud deck with the cloud tops slightly below 2000 m between
12:00 and 14:00 Z is broken. The weather pattern for the 27 September 2012
(Fig. d) shows Leipzig directly in front of a well pronounced
trough. Temperatures at 850 hPa lie again around 10 ∘C and
the cloud types vary between stratocumulus and shallow cumulus. The cloud
base height increases throughout the day.
Cloud adiabaticity
Entrainment of dry air into the clouds leads to evaporation of cloud water
and therefore to a deviation from the adiabatic liquid water content profile.
Knowledge of fad is required to calculate key quantities for
investigating ACI from passive satellite observations. Therefore we first
study cloud adiabaticity, before conducting a intercomparison of ground-based
and satellite key properties as well as discuss sources of its uncertainties.
fad can be calculated from the ground-based observations. We will
further investigate possibilities to estimate it from passive satellite
observations.
Adiabatic factor from ground-based observations
The ground-based fad is calculated using QL from the
microwave radiometer, H as the difference of cloud top height from the
cloud radar and cloud base height from the ceilometer, and
Γad(Tcbh,pcbh) using NWP data in
Eq. ().
suggests a range of typical values of [0.3, 0.9]. We
omitted adiabatic factors with fad>1.0 since those are most
likely affected by the measurement uncertainties, since the occurrence of
“superadiabatic” cloud profiles in nature is physically implausible. Such
artefacts especially arise due to uncertainties in QL and H for
thin clouds. In contrast to the original Cloudnet code, our calculation of
fad allows for values greater than 1.0. Within Cloudnet
“superadiabatic” profiles are avoided by increasing the cloud top height if
the integrated adiabatic qL is smaller than QL measured
by the microwave radiometer.
An example time series for one case (21 April 2013) is shown in
Fig. (see the Supplement for more cases). For this
case we find values of fad between 0.2 and 0.6 before 09:00 UTC.
Measurements of Z (Fig. b) reveal that the cloud base is
more inhomogeneous during this time period than later on. After 09:00 UTC,
fad varies between 0.5 and 1.0.
Uncertainty estimation for Nd and τ by varying Z,
QL and the effective variance of the gamma distribution ν.
Relative uncertainties are given in brackets. Case 1: 21 April 2013,
11:00 UTC. QL=69 g m-2, H=311 m, fad=0.76. Retrieved values: Nd=456 cm-3 applying ν=0.1,
τ=18. Case 2: 1 June 2012, 13:30 UTC. QL=62 g m-2,
H=342 m, fad=0.55, Nd=216 cm-3, τ=13.6.
Median and standard deviation of fad (calculated from
Eq. ) for individual cases. Furthermore the median of
fad, classified into updraft (v≥0) and downdraft (v<0)
regimes, as well as the fraction of sub-adiatic cloud profiles is shown. Values of fad>1.0 are
omitted because those are likely affected by measurement uncertainties.
21 Apr 201327 Sep 201227 Oct 20111 Jun 2012Median fad0.630.620.700.44SD fad0.180.210.120.24Median fad[v≥0]0.780.640.760.44SD fad[v≥0]0.210.200.120.23Median fad[v≤0]0.610.620.660.44SD fad[v≤0]0.170.210.100.24Fraction fad<10.990.790.990.90
From Fig. a we find a mean of fad=0.63 and the
interquartile range (IQR) as [0.46, 0.81] for the entire data set covering
2012 and 2013. This corresponds well with the typical value of 0.6 given by
. Overall, there is a large spread of values covering the
full physical meaningful range from 0 to 1 (mean values for individual cases
as presented in Fig. are listed in
Table ). fad not only changes from case
to case, but also varies with time for individual days, reflecting the
natural variability of entrainment processes. The variability of
fad is larger for the inhomogeneous cases than for the
homogeneous ones (Table ), but the range of values
is similar. This shows that independent from temporal cloud homogeneity the
majority of clouds seems to be sub-adiabatic. Therefore considering a
constant fad like in previous studies
(Table ) could affect
retrievals of cloud properties.
When looking for proxies for fad, we find a tendency that
geometrically thicker clouds are less adiabatic (Fig. b).
Already found a decrease in fad with height.
It also supports the findings of , who observed the tendency
that thicker clouds are less adiabatic in the Southeast Pacific. Mainly the
thin clouds (H<400m) result in fad>1, as also
found by , and therefore the investigation of such thin
clouds remains challenging.
Time series of the adiabatic factor fad for
21 April 2013. Black dots represent fad derived using
ground-based H and QL. The gray line represents the 10 min
averaged and interpolated fad neglecting superadiabatic
values.
used observations of two cases with temporally
homogeneous stratocumulus clouds over Leipzig, Germany, and found that in
case of updrafts in clouds, the qL profile tends to be more
adiabatic. To investigate if such a behaviour also occurs for our cases we
apply the cloud radar Doppler velocity at the cloud base. The average
vertical velocity at cloud base for all samples in 2012 and 2013 is found to
be -0.1 m s-1 with the majority of points (93 %) in the range
[-1, 1] m s-1. Considering the vertical velocity as function of
fad (Fig. c) we find a large spread, which makes
it difficult to detect a distinct influence of updraft speed on cloud
adiabaticity. However, the notch around the median in the box–whisker plot
does not overlap for updraft and downdraft regimes. According to
the median can be judged to differ significantly on
the 95 % confidence interval if there is no overlay in the notches. We
further calculate the median fad for updraft and downdraft
regimes for the four selected cases, and find for three out of four cases
that clouds are slightly more adiabatic in the updraft regime (Table
). This behaviour is expected from adiabaticity and
also supported by the findings of . They report that this
effect is strongest at the cloud base and blurs when the data points are
averaged over the whole cloud profile.
(a) Histogram of fad in 2012 and 2013 at
LACROS. (b)fad as a function of observed H. Colours
indicate different QL bins. The solid lines represent the
relationship described in Eq. () for bin mean QL and
Γad=1.9×10-3 g m-4.
(c)fad separated by up- and downdraft at the cloud
base.
Adiabatic factor from satellite observations
From ground-based observations we can show that fad is highly
variable even for one location. Therefore we can also expect strong
variability for cloud regimes over different regions observed by satellite
(e.g. maritime vs. continental). To obtain ACI key quantities from passive
satellite observations, fad is required over a larger domain. The
German weather service (DWD) operates a ceilometer network
which can be used to obtain the cloud base height (CBH).
The question remains if QL and CTH from SEVIRI are accurate enough
to allow for an estimate of the adiabatic factor using Eq. (). To
address this question, we contrast QL and CTH obtained from SEVIRI
with LACROS.
We investigate liquid clouds in a 2-year period covering 2012 and 2013.
Since the estimate of fad from passive satellite observations is
expected to be applied over a larger domain, it should be independent from
ground-based information. Therefore the sampling is now done in terms of
satellite observed quantities. An area of 5 × 5 pixels (total of
25 pixels) centred at the location of LACROS is considered for each
available SEVIRI observation. For this pixel field we obtain average,
standard deviation of CTH and the liquid cloud fraction. The liquid fraction
is determined by the cloud type classification for each pixel from CPP. We
require 90 % of the pixel field (23 out of 25 pixels) to be classified as
pure liquid clouds. As additional constraint, the standard deviation of CTH
for the 25 pixels has to be smaller than 400 m. For LACROS we use the
observation averaged using a window of 10 min around the SEVIRI observation
time. No requirements regarding the cloud phase are made for LACROS.
We first look at the CTH, which can be compared at daytime and night-time. The
ground-based instruments give the actual geometrical CTH while from passive
satellites a radiative CTH is obtained. Ignoring this physical difference we
can see that the SEVIRI CTH is positively biased
(Fig. a). reports a very
similar overestimation (320 m) with a large standard deviation of 1030 m for
low, opaque clouds. Considering the central pixel of the field does not
change the result significantly, showing that the cloud fields are rather
homogeneous and should therefore be suitable for such a comparison. The
observed bias is not explained by the limited vertical step size of 200 m in
the SEVIRI CTH product. A likely explanation of this bias is found in the
representation of inversions. Splitting the sample by model inversions did
not provide significantly better results, but the actual inversions might not
be well represented by the model. Such a case can be seen for
27 October 2011. There, the CTH is roughly 1000 m lower than for the other
three cases presented here, but the retrieved satellite CTH lies at 2000 m.
Considering the closest radiosounding of Lindenberg (Germany), we find two
inversion layers on top of each other between 900 and 3000 m, which results
in ambiguities in finding the correct cloud height. Differences may also
result from semitransparent cirrus cloud layers (21 April 2013), or broken
cloud conditions (1 June and 27 September 2012).
Histogram of differences between SEVIRI and LACROS derived cloud
properties for 2012 and 2013: (a) cloud top height (CTH),
(b) cloud optical depth (τ), (c) liquid water path
(QL). Median of 5 × 5 SEVIRI pixels centred at LACROS
(dark gray), closest pixel to LACROS (light gray). Zero difference is marked
by a dashed red line.
For the comparison of QL we impose the condition that the values
are between 20 and 400 g m-2. The comparison can only be applied
during daytime. Both requirements reduce the number of samples by 56 %
compared to the CTH sample. The difference of QL has a distribution
with a distinct peak close to zero (Fig. c).
There is a small negative bias of -21 g m-2, which is within the
uncertainty range of the ground-based measurements, not even considering the
uncertainty of the satellite-based estimate. Similar to the CTH comparison we
see that the distribution of the central pixel is not significantly different
from the field average, although the spread is larger. The distribution and
the standard deviation are consistent with the observations in the validation
study of for the Cloudnet stations of Chilbolton and
Palaiseau. Similar to their study we see a slight negative skewness, which
stems from larger QL values seen from the ground-based microwave
radiometer. also reported that accuracy is reduced for
higher QL values. Further possible explanations for differences in
QL observed from ground and SEVIRI can be found in remaining cloud
inhomogeneities and sampling differences. Generally, unfavourable viewing
angles that occur especially in winter conditions can lead to large
uncertainties in the satellite retrieval. In our sample the majority of the
cases occur in summer months (April to September, 80 %). Looking at
specific case days, we find the mean difference of QL for two
homogeneous cases between SEVIRI and the ground-based microwave radiometer in
reasonable agreement (8 g m-2 (10 %) for 21 April 2013,
25 g m-2 (32 %) for 27 October 2011), while there are larger
differences for two inhomogeneous cases (50 g m-2 (87 %) for
1 June 2012 and 33 g m-2 (80 %) for 27 September 2012).
A similar study by found a standard deviation of
369 m between satellite-based adiabatic CBH and ceilometer CBH. They applied
CTH and QL from AVHRR (Advanced Very High Resolution Radiometer)
and assumed adiabatic clouds to compare the spatially and temporally averaged
satellite product. The same comparison between SEVIRI and radiosonde
observations resulted in a standard deviation of ±290 m
. They suggest that this method can be applied for
convective clouds in their early growth stage, which are located near the
condensation level. Their sample is focused on relatively thin water clouds
(∼ 250 m), which are more likely close to adiabaticity according to our
Fig. b. As we will discuss in the following the adiabatic
factor for such thin clouds is very sensitive to errors in H, so that an
instantaneous retrieval of fad is not feasible.
Uncertainty estimate of fad
To investigate the uncertainties that influence the calculation of
fad, we consider an adiabatic cloud (fad=1) with
QL=100 g m-2 and H=324m and Γad=1.9× 10-3 g m-4. The QL retrieval
uncertainty (microwave radiometer instrument error + retrieval error) is
approximately 25 g m-2 and the vertical resolution of the ceilometer
and the cloud radar results in at least ±60 m uncertainty of H.
Accounting for the maximum uncertainty (QL=125 g m-2, and
Hobsground=264 m) or (QL=75 g m-2
and Hobsground=384 m), the resulting fad
would be 1.89 or 0.54, respectively. This shows that with the current
uncertainty limits of the ground-based observations fad is still
prone to large uncertainties especially for geometrically thin clouds.
If we consider the root mean square differences (RMSD) of the comparison of
ground and satellite-based values with ΔQL=67 g m-2
and ΔCTH = 1174 m, we can clearly see that especially the
observed bias in CTH can result in large uncertainties of an instantaneous
estimate of fad especially for thin clouds. For the adiabatic
cloud considered above, this RMSDs result in a relative uncertainty for the
adiabatic factor of 727 %, neglecting uncertainties at the CBH. Even
considering a cloud that is twice as thick, the relative uncertainty is still
362 %. This shows that subsampling the SEVIRI observations to homogeneous,
liquid clouds does still show differences when compared to a ground-based
reference that are too large to estimate fad with sufficient
reliability, mainly due to uncertainties in the CTH product. With this
approach using QL and H we cannot determine the adiabaticity of
clouds with a reasonable accuracy. Therefore we will have a look at the
microphysical quantities.
Microphysical key quantities relevant for ACI
H and Nd are used as the main parameters in many investigations
of ACI as both cloud properties have a direct effect on cloud albedo. Due to
the required assumptions about the DSD, a retrieval of Nd from a
radar–radiometer approach remains highly uncertain.
follows an alternative approach to retrieve τ instead of Nd
and demonstrated it to be less sensitive to the assumption of the width of
the DSD.
In the following, we will cross-check key quantities H and τ from
ground and satellite. We will also discuss the effect of uncertainties in our
observations for the sub-adiabatic cloud model on Nd, τ and
H.
Cloud geometrical depth H intercomparison from space and ground
Contrasting SEVIRI H (Eq. , using fad from
ground-based observations) with the LACROS H, we are able to investigate
the same quantity obtained with two independent physical retrieval
approaches.
The correlation coefficient is 0.89 for 21 April 2013, 0.70 for
27 October 2011, 0.38 for 1 June 2012, and 0.45 for 27 September 2012 and
increases by 10, 39, 118, and 71 % for 30 min temporal averaging,
respectively (see Table ). The improvement of
correlation is not surprising when comparing averaged data,
e.g.. However, a longer averaging
period removes the original variability of the data. The correlation for
temporally averaged data is within the range of values that were obtained by
, and .
found correlations of 0.71 between SEVIRI and Cloudnet
for a homogeneous stratocumulus cloud layer. found
correlations of 0.62 between in situ and MODIS retrieved H, and could show
a better agreement of H when fad is explicitly calculated and
considered. found correlations of 0.54 (0.7 for
H<400m with cloud fraction > 90 %) comparing
radiosonde-derived H to respective MODIS observations. In their study
reported that satellite values were higher compared to
the ground-based ones. The reason for this can potentially be explained by
a bias of MODIS-retrieved re but also in the choice of
fad in the retrieval of H.
Correlation coefficient of H from LACROS and from SEVIRI (3x3
pixel spatial average) for different temporal averaging periods applied to
both data sets.
Date5 min average10 min average30 min average21 Apr 20130.890.960.9827 Oct 20110.700.720.9727 Sep 20120.450.610.771 Jun 20120.380.530.83
Relationship between QL and τ for the four case days
(Table ). Blue crosses represent the LACROS
observations for the case day, black dots the SEVIRI observations. The solid
blue line represents the relationship between τ and QL for the
median fad and Nd of the LACROS observations.
Uncertainty estimates of τ as a function of QL is given in
terms of temporal variability using the IQR of the time series (dashed), and
as 50 % relative uncertainty in Nd and fad (dotted).
Furthermore the histograms of ground-based and SEVIRI observations are shown
on each axis in the same colours as stated before.
Cloud optical depth τ intercomparison from space and ground
The intercomparison of SEVIRI with LACROS retrieved τ results in
differences of 2.3 (8 %) for 21 April 2013, 3.6 (21 %) for
27 October 2011, 9.3 (76 %) for 1 June 2012 and 8.0 (61 %) for
27 September 2012. The higher resolution of the ground-based observations
leads to larger variability also for the homogeneous cases. The median
conditions result in a good fit to the satellite (τ, QL)-pairs
(Fig. ) for the homogeneous case on 21 April 2013.
For this case the satellite pairs are also within the ground-based temporal
IQR. The situation is similar even for the inhomogeneous case on 1 June 2012.
The situation turns out to be more complicated when looking at the
inhomogeneous case on 27 September 2012. Overall satellite τ and
QL show lower values, which result likely due to broken-cloud
effects in the SEVIRI retrieval. For broken clouds within the SEVIRI pixel
the satellite receives a combined signal from the clouds but also from the
surface. Such moving, broken cloud fields result in a smoother temporal
pattern from the satellite perspective. From the time–height Z
cross section on 27 September 2012 between 11:00 and 15:00 UTC a larger
number of cloud gaps can be seen, which could explain why the subpixel
surface contamination plays a larger role than on 1 June 2012. The Cloudnet
observations on 27 September 2012 show rapid changes of QL with
peaks around 400 g m-2 and cloud free periods. The observed larger
deviations of SEVIRI found on 27 October 2011 are likely due to low values
(< 5 µm) of effective radius in the KNMI–CPP retrieval. These
are likely a result of the unfavourable viewing conditions with a large solar
zenith angle (> 60∘) under relative azimuth angles close to
180∘ around noon for this case, for which
pointed out the low precision of the retrieval. These values are filtered out
following , but the remaining points might also be
affected by the same issue.
To highlight the importance of considering the actual fad for the
retrieval process, we calculated τ (Eq. ) from
the ground-based observations following the radar–radiometer approach with
fad=1 and with the ground-obtained fad. Afterwards we
compare it to the satellite-retrieved values. Applying fad=1 the
mean difference in optical depth is increased from 2.3 to 8.5 on
21 April 2013, and is also higher for the other cases (see
Table ).
The distribution of differences between SEVIRI and ground-based τ for
the 2012 and 2013 sample of low-level, homogeneous, liquid clouds is
presented in Fig. b. As for QL
there is a distinct peak around zero with negligible bias, but a considerable
standard deviation of 16. This shows that on average the agreement between
satellite and ground-based τ is reasonable, considering the number of
uncertainties in the retrieval as well as uncertainties due to parallax,
collocation, and spatial resolution. Those uncertainties will be discussed in
more detail in the following sections.
Mean difference of τ between SEVIRI and LACROS for each case,
when fad as obtained from the ground-based observations is
applied and fad is considered constantly 1.0.
DateτSEVIRI-τLACROS‾fad=fadLACROSτSEVIRI-τLACROS‾(fad=1)21 Apr 20132.38.527 Oct 20113.66.627 Sep 20127.910.91 Jun 20129.312.8Ground-based uncertainties
The radar–radiometer retrieval depends upon the observations of QL,
H, and Z(z). Also the choice of the mixing model is able to change the
retrieved quantities, but comes to the conclusion that this
effect is small. Nd depends further on k6, which only depends on
the width of the DSD (see Eq. in Appendix A).
We take two typical cloud profiles from our observations. For those cloud
profiles we evaluate the sensitivity of the retrieved Nd to the
uncertainties of the input parameters based on . In
Table we list the sensitivities to
each input parameter when the other parameters are kept constant.
For Z(z) we follow and assume an uncertainty range of
±2 dBZ, which would represent a calibration bias constant with height.
Drizzle does have a strong influence on Z,
e.g.. Errors of 30–60 %
have to be anticipated for qL profile retrievals. Those retrieval
approaches are based on very similar principles as the radar–radiometer
retrieval method . In our study we filtered out drizzling
profiles as well as possible. For the four case days re observed
from satellites near cloud top lies clearly below the value of 14 µm
which was suggested by as the threshold for
drizzle/rain forming clouds. The maximum of Z(z) in each profile also did
not exceed -20 dBZ, which is commonly taken as a drizzle threshold
. We cannot totally rule out the possibility
that few larger droplets were present, to which Z is very sensitive. For
the uncertainty of H, we assume ±60 m. For QL we assume a
typical uncertainty of ±25 g m-2 given microwave radiometer
observations. The width of the DSD for continental clouds exhibits a large
spread of values in literature as can be seen in . If we
consider the maximum range of observations, the effective variance ν of
the gamma size distribution could take values between 0.043 up to 0.2
(k2=0.87 and k2=0.48, respectively). For the standard retrieval we
assume ν=0.1 (k2=0.72).
Nd is most sensitive to the assumption about the width of the DSD,
especially to changes in the range of smaller values of the effective
variance. This can be understood as Nd∝k6 and k6 is a
monotonically decreasing function of the effective variance. For higher
values of ν the other uncertainty contributions are equally or even more
important. Since the real DSD is usually unknown, it is difficult to estimate
the actual uncertainty when assuming ν=0.1. From our cases we find that
the uncertainty in QL might be more important than the uncertainty
in radar reflectivity. Both can result in more than 50 % relative
uncertainty for the retrieval of Nd.
As can be seen from Eq. (), the optical depth τ
is sensitive to the same input parameters as Nd, but also depends
on fad. Therein the combined uncertainty of QL and H
is reflected. From Table we find
that τ is most sensitive to uncertainties in QL, especially
for observed low values of QL. In contrast to Nd, it is
not as sensitive to the assumption about the width of the DSD. While for
Nd the uncertainty in the low-range of ν is above 100 %, it
is below 20 % for τ. Since the natural variability of DSDs is large
and difficult to constrain without in situ observations, τ turns out to
be a more stable quantity for contrasting to other observation, as already
suggested by .
In Fig. we present the uncertainty of τ as a
function of QL, based on the median observations from the
ground-based time series. We use a representative average of Nd
over the whole time period and investigate the effect of its temporal
variability on the retrieved τ.
Recognizing the difficulty in retrieving Nd from the 3rd and 6th
moments, used a climatological mean value for Nd
in order to retrieve re. They reported an average Nd of
212±107 cm-3 at the Southern Great Plains site for continental
clouds, which is similar to the median value found for our example cases in
Fig. . We see that assuming a 50 % uncertainty for
both, Nd and τ, results in an increasing uncertainty of τ
with QL, with the uncertainty due to ΔNd being
slightly larger, although Δfad cannot be neglected.
Satellite uncertaintiesUncertainties of Nd and H
Since Nd is obtained with the sub-adiabatic model using
Eq. (), it depends on the uncertainties of τ and
re, but also on fad, k2 and Γad.
reported a 150 cm-3 error for optically thick
clouds (τ>20) resulting from a 10 % error in τ. The absolute
error of Nd increases with increasing τ assuming a constant
error in re. Nd is also very uncertain for values of
re<8µm. found that cases with
re<5µm are rare compared to typical value of
10 µm for liquid clouds. argue that those
should not be considered due to the large uncertainty.
If the individual errors are assumed to be normally distributed, the relative
errors of Nd and H are given by the following:
ΔNdNd2=Δk2k22+ΔΓad2Γad2+Δfad2fad2+Δτ2τ2+5Δre2re2
and
ΔHH2=ΔΓad2Γad2+Δfad2fad2+Δτ2τ2+Δre2re2.
Uncertainties of τ and re stem from the assumption of
plane-parallel vertical-uniform cloud layers, partially covered cloud pixels
, 3-D effects , and large solar zenith
angles . Uncertainties in re further arise
from its vertical profile. The use of different channels results in
discrepancies in re. MODIS uses a channel centred at
2.1 µm, while SEVIRI uses 1.6 µm for the standard
retrieval. From MODIS, additional re retrievals from channels at
1.6 and 3.7 µm are available. Theoretically, the 3.7-µm
channel should represent re closer to the cloud top for adiabatic
clouds, while the 2.1- and 1.6-µm channels receive the main signal
from deeper layers within the cloud. Cloud observations do not always show an
increase of re from channel 1.6 µm over 2.1 to
3.7 µm as is expected for plane-parallel, adiabatic clouds
. In this study we estimate the uncertainties
in passive satellite τ and re with 10 % following
(SEVIRI) and following (MODIS),
although uncertainties are probably larger for unfavourable conditions (large
solar zenith angles, broken clouds).
For Δfad we assume a relative error of 35 % considering a
constant fad=0.6 and its variability (0.22) as obtained from
2-year LACROS observations. For comparison, assumed an
uncertainty in fad of 0.3. This resulted in a numerically
evaluated error of around 26 % considering typical values of re
and τ.
estimated the uncertainty of k2 to be negligible
(around 3 %) for Nd<100 cm-3, following
. used a variability of k2=0.8±0.1 in a global study, which results in a relative uncertainty of 12.5 %.
found a similar mean value for 33 cases of
stratocumulus and cumulus clouds with an even smaller variability, even
slightly lower than the variability in . Therefore 12.5 %
might be seen as an upper uncertainty limit for k2.
By considering the whole seasonal variability of cloud base temperature,
obtained an error of 24 % for
Γad(T,p). In our study Γad has a smaller
contribution to those uncertainties due to the fact that we are using model
data to gain more reliable information about cloud base temperature and
pressure instead of considering a constant value of Γad as
in, e.g. . If we compare Γad calculated from
satellite cloud top temperature and pressure with the one calculated from
cloud base values observed from ground we find an uncertainty of 15 %
considering the four case days. As we see deviations in the cloud top height,
we believe that this uncertainty can be mainly attributed to incorrect
satellite estimates of cloud top temperature and pressure.
state for satellite retrievals of Nd (and also
Had) that fad and Γad are the most
important uncertainty factors. Considering our uncertainty estimates, the
largest contribution to the uncertainty of Nd is given by the
relative uncertainty of re (25 %), followed by fad
(18 %), k2 (12.5 %), Γad (7.5 %) and τ
(5 %). Considering the error propagation of H, assuming the same errors
as for Nd, we find the largest uncertainty due to fad
with 17.5 %, followed by Γad (7.5 %) and τ (5 %)
and re (5 %).
The importance of re for the retrieval of Nd from passive
satellite imagers has already been pointed out by previous studies. Those
were mainly based on observations from MODIS
and report a high bias
of MODIS re, especially for broken clouds .
also state that the choice of the other parameters in
the retrieval (namely k2, Γad) is able to compensate for
this effect so that still a good agreement between MODIS retrieved and
in situ values could be achieved. As mentioned before, for our study we
focused on the intercomparison of τ instead of Nd, since the
ground-based retrieval of τ is less sensitive to the required
assumptions.
Effect of spatial resolution by comparing MODIS and SEVIRI
observations for two timesteps: (a) inhomogeneous case, 1 June 2012
at 12:25 UTC, (b) homogeneous case, 21 April 2013 at 11:50 UTC.
SEVIRI values are shown in black, MODIS values in blue and ground-based ones
in red. The closest pixel (central) to LACROS is shown as a dark square.
Field averages from the sensors original resolution are given as dots. For
MODIS also the average to SEVIRI resolution is presented (MODIS geos, light
blue square). Also the standard deviation is shown together with the averages
in terms of error bars.
Uncertainties due to spatial resolution
To investigate the effect of spatial resolution, we use collocated MODIS and
SEVIRI observations. We use the products of MODIS at 1 km spatial resolution.
We re-project all MODIS pixels to the 3 × 3 SEVIRI pixels so that
both instruments cover the same area. We then average the MODIS 1 km
resolution data to SEVIRI's spatial resolution (4 km × 6 km). In a
further step we average a 3 × 3 pixel field from SEVIRI and the
MODIS pixels at original resolution and calculate their standard deviation.
In this way we tried to use MODIS to account for SEVIRI's subpixel
variability, while neglecting deviations due to the differences of both
instruments and retrievals. In Fig. the results for
(a) the inhomogeneous case at 1 June 2012 and (b) the homogeneous case at
21 April 2013 are shown. For the inhomogeneous case we can clearly see the
large spread of MODIS τ values, which is reduced to a similar range as
for SEVIRI τ when averaged to the same spatial resolution. The spread of
τ is found larger than for re. For the homogeneous case the
spread is smaller. Differences between MODIS and SEVIRI after averaging are
in a similar range for both cases. When comparing averaged data, MODIS and
SEVIRI show similar results for both cases. However, the differences,
especially in terms of re can be of the same magnitude than those
to ground-retrieved values. There is considerable difference when taking
either the closest pixel to the ground-based location or the spatially
averaged value, while the closest pixel does not necessarily result in a
better agreement with the ground-based value (Fig. ).
Therefore we can conclude that especially for inhomogeneous cases, the
sub-pixel variability introduces an important additional uncertainty factor.
Summary and conclusions
In this work, we aimed to evaluate the consistency and limitations of our
ground-based and satellite cloud retrieval products which are typically used
to quantify aerosol–cloud interactions (ACI). We used a 2-year data set with
four selected case studies.
Cloud properties have been used previously for diagnosing ACI and
specifically the first indirect aerosol effect from both ground-based
supersites, e.g. as well as geostationary passive
satellite observations, e.g.. The sub-adiabatic cloud
model as a conceptional tool is commonly applied and modified using an
adiabatic factor fad to account for entrainment within the cloud.
Based on cloud geometric depths obtained from the combination of ground-based
cloud radar and ceilometer, and liquid water path from a microwave
radiometer, we demonstrated that for a 2-year data set, neither the
assumption of an adiabatic cloud nor the assumption of a temporally constant
fad is fulfilled (mean fad=0.63±0.22).
As fad is required to estimate key quantities for ACI studies,
but cannot be obtained from passive satellite observations within a
sufficient uncertainty range, an independent method to estimate
fad, and thus the influence of mixing, would be highly desirable
for global-scale analyses. We were able to support previous findings which
reported that thinner clouds are closer to adiabaticity as
well are clouds that show upward motion at the cloud base
.
To investigate ACIs from passive satellites the cloud droplet number
concentration Nd is widely used as a key parameter. An
intercomparison with ground-retrieved values is complicated as it turns out
that its retrieval from a ground-based radar–radiometer approach is very
sensitive to assumptions about the width of the DSD and the radar
calibration. The Nd retrieval from radar is poorly posed because of
the moment disparity and the potential instability of the ratio in
Eq. () as pointed out by .
Retrieved values of Nd can change by more than 135 % just due to
wrong assumptions made for the width of the DSD. From passive satellite we
find the main sensitivity to uncertainties in the effective radius. We
conclude that neither the ground-based nor satellite-based cloud retrieved
properties used here allow to obtain a robust instantaneous estimate of
Nd, which complicates their use for the study of ACIs.
We demonstrated that cloud optical depth τ from ground-based
radar–radiometer retrievals is less sensitive to the assumptions about the
DSD and is therefore better suited to investigate ACIs, consistent with the
conclusions of . It is most sensitive to uncertainties in
the liquid water path (changes of up to 50 % for an uncertainty of
25 g m-2 are possible).
Given an independent retrieval of τ, e.g. from MFRSR retrievals
, and information such as radar Doppler velocity
, should give further options for validation. Applying
such additional observations in an optimal estimation scheme might give the
opportunity to better constrain the retrieved Nd. Also the
application of cloud radar scanning capabilities together with radiance
zenith measurements might improve the retrieval . For
validation of those Nd retrievals accompanying in situ measurements
are required.
Instantaneous comparisons of τ between space and ground may result in
large differences, especially for broken cloud conditions and unfavourable
viewing conditions. Applying spatial and temporal averaging and subsampling
to rather homogeneous, liquid clouds leads to a reasonable agreement in
τ for a majority of observations during a 2-year period at LACROS,
especially considering the large number of assumptions and uncertainties.
Besides the retrieval uncertainties, differences in spatial resolution
affect the comparison not only between space and ground observations, but
also between space-based instruments of different resolution and viewing
angles (i.e. SEVIRI, MODIS). We highlighted, that especially for
inhomogeneous cases, sub-pixel variability is an important uncertainty
factor, but that averaging does not necessarily result in a better agreement
to ground-based observations than taking the closest pixel to the location.
To generalize such results more collocated MODIS, SEVIRI and ground-based
observations need to be examined.
Given the network of Cloudnet/ACTRIS in central Europe this offers the
opportunity to investigate the climatology of fad and investigate
its regional, seasonal or synoptical dependency in further studies.
With the upcoming Meteosat Third Generation (MTG) satellite
a higher spatial resolution of cloud products will be
available and should therefore mitigate issues due to spatial resolution for
the geostationary perspective. Also the sounder capabilities of MTG should
give new opportunities, e.g. to overcome problems of cloud geometrical depth
retrievals from passive satellites by using additional information from the
oxygen A-band following the method as outlined by, e.g. ,
and therefore might give
the possibility to obtain fad over a larger domain.
To obtain the factors k2 and k6 in the sub-adiabatic cloud model a
gamma size distribution is assumed in the form of :
η(r)=Arβexp-Λr=η0Γ(1-2νν)reν1-2ννrre(1-3ν)νexp-rreν
with
β=1-3ννΛ=1reνA=η0Λβ+1Γ(β+1).
Hereby the effective radius re, its effective variance ν, and
the total number density of droplets η0 are used. re is
defined as the third over the second moment of the DSD and
can be linked to the mean volume radius (rv) with the following
relationship:
re3=k2-1rv3.
From the gamma size distributions its nth moments can be derived by
:
Mη,n=A∫rn+βexp-Λrdr=AΓβ+n+1Λ(β+n+1).
The factor k2 is then only a function of the width of the DSD:
k2=M2(η)3M3(η)2=(1-2ν)(1-ν).
Z as proportional to the sixth moment of the DSD can be expressed as a
function of Nd, qL and factors that depend on the width
of the DSD (k6) :
Z=92π2ρw2k6qL2Nd.
Similar to k2, the factor k6 is defined:
k6=M6(η)M3(η)2=(ν+1)(2ν+1)(3ν+1)(1-2ν)(1-ν).
Integrating over H, we can solve the equation for QL:
QL=92π2ρw2-12∫1k6(ν(z))Nd(z)Z(z)dz.
In the homogeneous mixing model, Nd(z) and ν(z) are assumed
constant with height. considers a column-averaged
Nd by weighting with the square-root of Z(z):
∫Nd(z)dz=∫Nd(z)Z(z)dz∫Z(z)dz=Nd‾.
Using the latter relationship, we yield a retrieval method for the
column-averaged Nd‾:
Nd‾(QL,Z,k6)=9k6QL22π2ρw2∫Z(z)dz2.
Equation () can be substituted into
Eqs. () and () to
eliminate Nd and to obtain a ground-based estimate of τ and
re.
The Supplement related to this article is available online at doi:10.5194/acp-16-933-2016-supplement.
Acknowledgements
The first author's work was funded by the Leipzig Graduate School on
Radiation (LGS-CAR). We would like to thank the Cloudnet project (European
Union Contract EVK2-2000-00611) for providing the ground-based cloud
products, and the EUMETSAT SAFS for providing the SEVIRI cloud products, as
well as the NASA's Earth–Sun System Division for providing MODIS cloud
products. We further acknowledge colleagues participating in the HOPE campaign
of the HD(CP)2 project in Jülich. We also thank our colleagues
Anja Hünerbein, Andreas Macke, Fabian Senf, Johannes Quaas, and three
anonymous reviewers and the editor for their helpful suggestions and
comments. Edited by: G. Feingold
ReferencesAckerman, A. S., Toon, O. B., Taylor, J. P., Johnson, D. W., Hobbs, P. V.,
and Ferek, R. J.: Effects of Aerosols on Cloud Albedo: Evaluation of Twomey's
Parameterization of Cloud Susceptibility Using Measurements of Ship Tracks,
J. Atmos. Sci., 57, 2684–2695,
10.1175/1520-0469(2000)057<2684:EOAOCA>2.0.CO;2,
2000.Ahmad, I., Mielonen, T., Grosvenor, D., Portin, H., Arola, A., Mikkonen, S.,
Kühn, T., Leskinen, A., Juotsensaari, J., Komppula, M., Lehtinen, K.,
Laaksonen, A., and Romakkaniemi, S.: Long-term measurements of cloud droplet
concentrations and aerosol-cloud interactions in continental boundary layer
clouds, Tellus B, 65, 20138, 10.3402/tellusb.v65i0.20138,
2013.Albrecht, B. A., Fairall, C. W., Thomson, D. W., White, A. B., Snider, J. B.,
and Schubert, W. H.: Surface-based remote sensing of the observed and the
Adiabatic liquid water content of stratocumulus clouds, Geophys. Res. Lett.,
17, 89–92, 10.1029/GL017i001p00089, 1990.Baker, M. B., Blyth, A. M., Carruthers, D. J., Choularton, T. W., Fullarton,
G., Gay, M. J., Latham, J., Mill, C. S., Smith, M. H., Stromberg, I. M.,
Caughey, S. J., and Conway, B. J.: Field studies of the effect of entrainment
upon the structure of clouds at Great Dun Fell, Q. J.
Roy. Meteor. Soc., 108, 899–916, 10.1002/qj.49710845810,
1982.
Battan, L. J.: Radar observation of the atmosphere, University of Chicago
Press, 1973.Bennartz, R.: Global assessment of marine boundary layer cloud droplet number
concentration from satellite, J. Geophys. Res.-Atmos.,
112, D02201, 10.1029/2006JD007547,
2007.Boers, R., Russchenberg, H., Erkelens, J., Venema, V., van Lammeren, A.,
Apituley, A., and Jongen, S.: Ground-Based Remote Sensing of Stratocumulus
Properties during CLARA, 1996, J. Appl. Meteorol., 39, 169–181,
10.1175/1520-0450(2000)039<0169:GBRSOS>2.0.CO;2,
2000.Boers, R., Acarreta, J. R., and Gras, J. L.: Satellite monitoring of the
first
indirect aerosol effect: Retrieval of the droplet concentration of water
clouds, J. Geophys. Res.-Atmos., 111, D22208,
10.1029/2005JD006838, 2006.
Boucher, O., Randall, D., Artaxo, P., Bretherton, C., Feingold, G., Forster,
P., Kerminen, V.-M., Kondo, Y., Liao, H., Lohmann, U., Rasch, P., Satheesh,
S. K., Sherwood, S., Stevens, B., and Zhang, X. Y.: Clouds and Aerosols, in:
Climate Change 2013: The Physical Science Basis. Contribution of Working
Group I to the Fifth Assessment Report of the Intergovernmental Panel on
Climate Change, edited by: Stocker, T. F., Qin, D., Plattner, G.-K., Tignor,
M., Allen, S. K., Boschung, J., Nauels, A., Xia, Y., Bex, V., and Midgley, P.
M., Cambridge University Press, Cambridge, United Kingdom and New York, NY,
USA, 2013.Brandau, C., Russchenberg, H., and Knap, W.: Evaluation of ground-based
remotely sensed liquid water cloud properties using shortwave radiation
measurements, Atmos. Res., 96, 366–377,
10.1016/j.atmosres.2010.01.009, 2010.Brenguier, J.-L., Pawlowska, H., Schüller, L., Preusker, R., Fischer, J.,
and Fouquart, Y.: Radiative Properties of Boundary Layer Clouds: Droplet
Effective Radius versus Number Concentration, J. Atmos.
Sci., 57, 803–821,
10.1175/1520-0469(2000)057<0803:RPOBLC>2.0.CO;2,
2000.Brenguier, J.-L., Burnet, F., and Geoffroy, O.: Cloud optical thickness and
liquid water path – does the k coefficient vary with droplet
concentration?, Atmos. Chem. Phys., 11, 9771–9786,
10.5194/acp-11-9771-2011, 2011.Bréon, F.-M., Tanré, D., and Generoso, S.: Aerosol Effect on Cloud
Droplet Size Monitored from Satellite, Science, 295, 834–838,
10.1126/science.1066434,
2002.Cahalan, R. F., Ridgway, W., Wiscombe, W. J., Bell, T. L., and Snider, J. B.:
The Albedo of Fractal Stratocumulus Clouds, J. Atmos.
Sci., 51, 2434–2455,
10.1175/1520-0469(1994)051<2434:TAOFSC>2.0.CO;2,
1994.Christensen, M. W. and Stephens, G. L.: Microphysical and macrophysical
responses of marine stratocumulus polluted by underlying ships: Evidence of
cloud deepening, J. Geophys. Res.-Atmos., 116, D03201,
10.1029/2010JD014638, 2011.Deneke, H., Knap, W., and Simmer, C.: Multiresolution analysis of the
temporal variance and correlation of transmittance and reflectance of an
atmospheric column, J. Geophys. Res., 114, D17206,
10.1029/2008JD011680, 2009.
Derrien, M.: Algorithm theoretical basis document for cloud products
(cma-pge01 v3.2, ct-pge02 v2.2, ctth-pge03 v2.2), Tech. rep., SAFNWC, 2012.
Derrien, M., Gléau, H., Daloze, J.-F., and Haeffelin, M.: Validation of
SAFNWC/MSG cloud products with one year of SEVIRI data, in: 2005 EUMETSAT
Meteorological Satellite Conference, pp. 95–103, 2005.Dong, X., Ackerman, T. P., Clothiaux, E. E., Pilewskie, P., and Han, Y.:
Microphysical and radiative properties of boundary layer stratiform clouds
deduced from ground-based measurements, J. Geophys. Res.-Atmos., 102, 23829–23843, 10.1029/97JD02119, 1997.Dong, X., Mace, G. G., Minnis, P., Smith, W. L., Poellot, M., Marchand,
R. T., and Rapp, A. D.: Comparison of Stratus Cloud Properties Deduced from
Surface, GOES, and Aircraft Data during the March 2000 ARM Cloud IOP, J.
Atmos. Sci., 59, 3265–3284,
10.1175/1520-0469(2002)059<3265:COSCPD>2.0.CO;2,
2002.Feingold, G., Eberhard, W. L., Veron, D. E., and Previdi, M.: First
measurements of the Twomey indirect effect using ground-based remote sensors,
Geophys. Res. Lett., 30, 1287, 10.1029/2002GL016633, 2003.Fielding, M. D., Chiu, J. C., Hogan, R. J., and Feingold, G.: A novel
ensemble method for retrieving properties of warm cloud in 3-D using
ground-based scanning radar and zenith radiances, J. Geophys. Res.-Atmos.,
119, 10912–10930, 10.1002/2014JD021742, 2014.Fischer, J., Cordes, W., Schmitz-Peiffer, A., Renger, W., and Mörl, P.:
Detection of Cloud-Top Height from Backscattered Radiances within the Oxygen
A Band. Part 2: Measurements, J. Appl. Meteorol., 30, 1260–1267,
10.1175/1520-0450(1991)030<1260:DOCTHF>2.0.CO;2,
1991.Flentje, H., Heese, B., Reichardt, J., and Thomas, W.: Aerosol profiling
using the ceilometer network of the German Meteorological Service, Atmos.
Meas. Tech. Discuss., 3, 3643–3673, 10.5194/amtd-3-3643-2010, 2010.Fox, N. I. and Illingworth, A. J.: The Retrieval of Stratocumulus Cloud
Properties by Ground-Based Cloud Radar, J. Appl. Meteorol., 36,
485–492, 10.1175/1520-0450(1997)036<0485:TROSCP>2.0.CO;2,
1997.Frisch, A. S., Fairall, C. W., and Snider, J. B.: Measurement of Stratus
Cloud
and Drizzle Parameters in ASTEX with a Ka-Band Doppler Radar and a Microwave
Radiometer, J. Atmos. Sci., 52, 2788–2799,
10.1175/1520-0469(1995)052<2788:MOSCAD>2.0.CO;2,
1995.Frisch, S., Shupe, M., Djalalova, I., Feingold, G., and Poellot, M.: The
Retrieval of Stratus Cloud Droplet Effective Radius with Cloud Radars,
J. Atmos. Ocean. Tech., 19, 835–842,
10.1175/1520-0426(2002)019<0835:TROSCD>2.0.CO;2,
2002.George, R. C. and Wood, R.: Subseasonal variability of low cloud radiative
properties over the southeast Pacific Ocean, Atmos. Chem. Phys., 10,
4047–4063, 10.5194/acp-10-4047-2010, 2010.Greuell, W. and Roebeling, R. A.: Toward a Standard Procedure for Validation
of Satellite-Derived Cloud Liquid Water Path: A Study with SEVIRI Data, J.
Appl. Meteorol. Clim., 48, 1575–1590,
10.1175/2009JAMC2112.1, 2009.Han, Q., Rossow, W. B., and Lacis, A. A.: Near-Global Survey of Effective
Droplet Radii in Liquid Water Clouds Using ISCCP Data, J. Climate, 7,
465–497, 10.1175/1520-0442(1994)007<0465:NGSOED>2.0.CO;2,
1994.Han, Q., Rossow, W. B., Chou, J., and Welch, R. M.: Global variation of
column
droplet concentration in low-level clouds, Geophys. Res. Lett., 25,
1419–1422, 10.1029/98GL01095, 1998.
Hansen, J. E. and Travis, L. D.: Light scattering in planetary atmospheres,
Space Sci. Rev., 16, 527–610, 1974.
Illingworth, A. J., Hogan, R. J., O'Connor, E. J., Bouniol, D., Brooks, M.
E., Delanoë, J., Donovan, D. P., Eastment, J. D., Gaussiat, N., Goddard,
J. W. F., Haeffelin, M., Klein Baltink, H., Krasnov,O. A., Pelon, J., Piriou,
J.-M., Protat, A., Russchenberg, H. W. J., Seifert, A., Tompkins, A. M., Van
Zadelhoff, G.-J., Vinit, F., Willén, U., Wilson, D. R., and Wrench, C.
L.: Cloudnet:
Continuous evaluation of cloud profiles in seven operational models using
ground-based observations, B. Am. Meteorol. Soc.,
88, 883–898, 2007.Janssen, R. H. H., Ganzeveld, L. N., Kabat, P., Kulmala, M., Nieminen, T.,
and Roebeling, R. A.: Estimating seasonal variations in cloud droplet number
concentration over the boreal forest from satellite observations, Atmos.
Chem. Phys., 11, 7701–7713, 10.5194/acp-11-7701-2011, 2011.Kim, B.-G., Miller, M. A., Schwartz, S. E., Liu, Y., and Min, Q.: The role of
adiabaticity in the aerosol first indirect effect, J. Geophys.
Res.-Atmos., 113, D05210, 10.1029/2007JD008961, 2008.King, N. J., Bower, K. N., Crosier, J., and Crawford, I.: Evaluating MODIS
cloud retrievals with in situ observations from VOCALS-REx, Atmos. Chem.
Phys., 13, 191–209, 10.5194/acp-13-191-2013, 2013.
Krzywinski, M. and Altman, N.: Points of Significance: Visualizing samples
with box plots, Nat. Meth., 11, 119–120, 2014.Lehmann, K., Siebert, H., and Shaw, R. A.: Homogeneous and Inhomogeneous
Mixing in Cumulus Clouds: Dependence on Local Turbulence Structure, J. Atmos.
Sci., 66, 3641–3659, 10.1175/2009JAS3012.1, 2009.Loeb, N. G. and Coakley, J. A.: Inference of Marine Stratus Cloud Optical
Depths from Satellite Measurements: Does 1D Theory Apply?, J.
Climate, 11, 215–233, 10.1175/1520-0442(1998)011<0215:IOMSCO>2.0.CO;2,
1998.Löhnert, U., Crewell, S., Simmer, C., and Macke, A.: Profiling Cloud
Liquid
Water by Combining Active and Passive Microwave Measurements with Cloud Model
Statistics, J. Atmos. Ocean. Tech., 18, 1354–1366,
10.1175/1520-0426(2001)018<1354:PCLWBC>2.0.CO;2,
2001.Löhnert, U., Feingold, G., Uttal, T., Frisch, A. S., and Shupe, M. D.:
Analysis of two independent methods for retrieving liquid water profiles in
spring and summer Arctic boundary clouds, J. Geophys. Res.-Atmos., 108, 4219, 10.1029/2002JD002861, 2003.Mace, G. G. and Sassen, K.: A constrained algorithm for retrieval of
stratocumulus cloud properties using solar radiation, microwave radiometer,
and millimeter cloud radar data, J. Geophys. Res.-Atmos., 105, 29099–29108, 10.1029/2000JD900403, 2000.Marshak, A., Platnick, S., Várnai, T., Wen, G., and Cahalan, R. F.:
Impact of three-dimensional radiative effects on satellite retrievals of
cloud
droplet sizes, J. Geophys. Res.-Atmos., 111, d09207,
10.1029/2005JD006686, 2006.Martin, G. M., Johnson, D. W., and Spice, A.: The Measurement and
Parameterization of Effective Radius of Droplets in Warm Stratocumulus
Clouds, J. Atmos. Sci., 51, 1823–1842,
10.1175/1520-0469(1994)051<1823:TMAPOE>2.0.CO;2,
1994.Martucci, G. and O'Dowd, C. D.: Ground-based retrieval of continental and
marine warm cloud microphysics, Atmos. Meas. Tech., 4, 2749–2765,
10.5194/amt-4-2749-2011, 2011.Martucci, G., Milroy, C., and O'Dowd, C. D.: Detection of Cloud-Base Height
Using Jenoptik CHM15K and Vaisala CL31 Ceilometers, J. Atmos. Ocean. Tech., 27, 305–318, 10.1175/2009JTECHA1326.1, 2010.McComiskey, A. and Feingold, G.: The scale problem in quantifying aerosol
indirect effects, Atmos. Chem. Phys., 12, 1031–1049,
10.5194/acp-12-1031-2012, 2012.Meerkötter, R. and Bugliaro, L.: Diurnal evolution of cloud base heights
in convective cloud fields from MSG/SEVIRI data, Atmos. Chem. Phys., 9,
1767–1778, 10.5194/acp-9-1767-2009, 2009.Meerkötter, R. and Zinner, T.: Satellite remote sensing of cloud base
height for convective cloud fields: A case study, Geophys. Res.
Lett., 34, L17805, 10.1029/2007GL030347, 2007.Miles, N. L., Verlinde, J., and Clothiaux, E. E.: Cloud Droplet Size
Distributions in Low-Level Stratiform Clouds, J. Atmos.
Sci., 57, 295–311,
10.1175/1520-0469(2000)057<0295:CDSDIL>2.0.CO;2,
2000.Miller, M. A., Jensen, M. P., and Clothiaux, E. E.: Diurnal Cloud and
Thermodynamic Variations in the Stratocumulus Transition Regime: A Case Study
Using In Situ and Remote Sensors, J. Atmos. Sci., 55,
2294–2310, 10.1175/1520-0469(1998)055<2294:DCATVI>2.0.CO;2,
1998.Min, Q. and Duan, M.: Simultaneously retrieving cloud optical depth and
effective radius for optically thin clouds, J. Geophys. Res.-Atmos., 110, d21201, 10.1029/2005JD006136, 2005.Min, Q., Joseph, E., Lin, Y., Min, L., Yin, B., Daum, P. H., Kleinman, L. I.,
Wang, J., and Lee, Y.-N.: Comparison of MODIS cloud microphysical properties
with in-situ measurements over the Southeast Pacific, Atmos. Chem. Phys., 12,
11261–11273, 10.5194/acp-12-11261-2012, 2012.Nakajima, T. and King, M. D.: Determination of the Optical Thickness and
Effective Particle Radius of Clouds from Reflected Solar Radiation
Measurements. Part I: Theory, J. Atmos. Sci., 47,
1878–1893, 10.1175/1520-0469(1990)047<1878:DOTOTA>2.0.CO;2,
1990.Painemal, D. and Zuidema, P.: Microphysical variability in southeast Pacific
Stratocumulus clouds: synoptic conditions and radiative response, Atmos.
Chem. Phys., 10, 6255–6269, 10.5194/acp-10-6255-2010, 2010.Painemal, D. and Zuidema, P.: Assessment of MODIS cloud effective radius and
optical thickness retrievals over the Southeast Pacific with VOCALS-REx in
situ measurements, J. Geophys. Res.-Atmos., 116, D24206,
10.1029/2011JD016155, 2011.Painemal, D. and Zuidema, P.: The first aerosol indirect effect quantified
through airborne remote sensing during VOCALS-REx, Atmos. Chem. Phys., 13,
917–931, 10.5194/acp-13-917-2013, 2013.Pawlowska, H., Brenguier, J., and Burnet, F.: Microphysical properties of
stratocumulus clouds, Atmos. Res., 55, 15–33,
10.1016/S0169-8095(00)00054-5,
2000.Pawlowska, H., Grabowski, W. W., and Brenguier, J.-L.: Observations of the
width of cloud droplet spectra in stratocumulus, Geophys. Res.
Lett., 33, l19810, 10.1029/2006GL026841, 2006.Petty, G. W. and Huang, W.: The Modified Gamma Size Distribution Applied to
Inhomogeneous and Nonspherical Particles: Key Relationships and Conversions,
J. Atmos. Sci., 68, 1460–1473,
10.1175/2011JAS3645.1, 2011.Platnick, S.: Vertical photon transport in cloud remote sensing problems,
J. Geophys. Res.-Atmos., 105, 22919–22935,
10.1029/2000JD900333, 2000.Platnick, S. and Valero, F. P. J.: A Validation of a Satellite Cloud
Retrieval
during ASTEX, J. Atmos. Sci., 52, 2985–3001,
10.1175/1520-0469(1995)052<2985:AVOASC>2.0.CO;2,
1995.
Platnick, S., King, M. D., Ackerman, S. A., Menzel, W. P., Baum, B. A.,
Riédi, J. C., and Frey, R. A.: The MODIS cloud products: Algorithms and
examples from Terra, IEEE T. Geosci. Remote, 41,
459–473, 2003.Quaas, J., Boucher, O., and Lohmann, U.: Constraining the total aerosol
indirect effect in the LMDZ and ECHAM4 GCMs using MODIS satellite data,
Atmos. Chem. Phys., 6, 947–955, 10.5194/acp-6-947-2006, 2006.Quaas, J., Boucher, O., Bellouin, N., and Kinne, S.: Satellite-based estimate
of the direct and indirect aerosol climate forcing, J. Geophys.
Res.-Atmos., 113, D05204, 10.1029/2007JD008962, 2008.Rémillard, J., Kollias, P., and Szyrmer, W.: Radar-radiometer retrievals
of cloud number concentration and dispersion parameter in nondrizzling marine
stratocumulus, Atmos. Meas. Tech., 6, 1817–1828,
10.5194/amt-6-1817-2013, 2013.Roebeling, R. A., Feijt, A. J., and Stammes, P.: Cloud property retrievals
for
climate monitoring: Implications of differences between Spinning Enhanced
Visible and Infrared Imager (SEVIRI) on METEOSAT-8 and Advanced Very High
Resolution Radiometer (AVHRR) on NOAA-17, J. Geophys. Res.-Atmos., 111, D20210, 10.1029/2005JD006990, 2006.Roebeling, R. A., Placidi, S., Donovan, D., Russchenberg, H., and Feijt, A.:
Validation of liquid cloud property retrievals from SEVIRI using ground-based
observations, Geophys. Res. Lett., 35, L05814, 10.1029/2007GL032115, 2008a.Roebeling, R. A., Deneke, H. M., and Feijt, A. J.: Validation of Cloud Liquid
Water Path Retrievals from SEVIRI Using One Year of CloudNET Observations,
J. Appl. Meteorol. Clim., 47, 206–222,
10.1175/2007JAMC1661.1, 2008b.Rosenfeld, D., Wang, H., and Rasch, P. J.: The roles of cloud drop effective
radius and LWP in determining rain properties in marine stratocumulus,
Geophys. Res. Lett., 39, L13801, 10.1029/2012GL052028, 2012.Saunders, R., Matricardi, M., and Brunel, P.: An improved fast radiative
transfer model for assimilation of satellite radiance observations, Q.
J. Roy. Meteor. Soc., 125, 1407–1425,
10.1002/qj.1999.49712555615, 1999.Schmetz, J., Pili, P., Tjemkes, S., Just, D., Kerkmann, J., Rota, S., and
Ratier, A.: An introduction to Meteosat Second Generation (MSG), B. Am. Meteorol. Soc.,
83, 977–992, 10.1175/1520-0477(2002)083<0977:AITMSG>2.3.CO;2,
2002.Schmidt, J., Ansmann, A., Bühl, J., Baars, H., Wandinger, U., Müller,
D., and Malinka, A. V.: Dual-FOV Raman and Doppler lidar studies of
aerosol-cloud interactions: Simultaneous profiling of aerosols, warm-cloud
properties, and vertical wind, J. Geophys. Res.-Atmos.,
119, 5512–5527, 10.1002/2013JD020424, 2014.Schueller, L., Brenguier, J.-L., and Pawlowska, H.: Retrieval of
microphysical,
geometrical, and radiative properties of marine stratocumulus from remote
sensing, J. Geophys. Res., 108, 8631, 10.1029/2002JD002680, 2003.Schüller, L., Bennartz, R., Fischer, J., and Brenguier, J.-L.: An
Algorithm
for the Retrieval of Droplet Number Concentration and Geometrical Thickness
of Stratiform Marine Boundary Layer Clouds Applied to MODIS Radiometric
Observations, J. Appl. Meteorol., 44, 28–38,
10.1175/JAM-2185.1, 2005.Schulz, J., Albert, P., Behr, H.-D., Caprion, D., Deneke, H., Dewitte, S.,
Dürr, B., Fuchs, P., Gratzki, A., Hechler, P., Hollmann, R., Johnston,
S., Karlsson, K.-G., Manninen, T., Müller, R., Reuter, M., Riihelä,
A., Roebeling, R., Selbach, N., Tetzlaff, A., Thomas, W., Werscheck, M.,
Wolters, E., and Zelenka, A.: Operational climate monitoring from space: the
EUMETSAT Satellite Application Facility on Climate Monitoring (CM-SAF),
Atmos. Chem. Phys., 9, 1687–1709, 10.5194/acp-9-1687-2009, 2009.Shupe, M. D.: A ground-based multisensor cloud phase classifier, Geophys.
Res. Lett., 34, L22809, 10.1029/2007GL031008, 2007.Stephens, G. L., Vane, D. G., Boain, R. J., Mace, G. G., Sassen, K., Wang,
Z., Illingworth, A. J., O'Connor, E. J., Rossow, W. B., Durden, S. L.,
Miller, S. D., Austin, R. T., Benedetti, A., Mitrescu, C., and CloudSat
Science Team, T.: THE CLOUDSAT MISSION AND THE A-TRAIN, B. Am.
Meteorol. Soc., 83, 1771–1790, 10.1175/BAMS-83-12-1771, 2002.Stuhlmann, R., Rodriguez, A., Tjemkes, S., Grandell, J., Arriaga, A.,
Bézy, J.-L., Aminou, D., and Bensi, P.: Plans for EUMETSAT's Third
Generation
Meteosat geostationary satellite programme, Adv. Space Res., 36,
975–981, 10.1016/j.asr.2005.03.091, 2005.Szczodrak, M., Austin, P. H., and Krummel, P. B.: Variability of Optical
Depth
and Effective Radius in Marine Stratocumulus Clouds, J.
Atmos. Sci., 58, 2912–2926,
10.1175/1520-0469(2001)058<2912:VOODAE>2.0.CO;2,
2001.Turner, D. D., Vogelmann, A. M., Johnson, K., Miller, M., Austin, R. T.,
Barnard, J. C., Flynn, C., Long, C., McFarlane, S. A., Cady-Pereira, K.,
Clough, S. A., Chiu, J. C., Khaiyer, M. M., Liljegren, J., Lin, B., Minnis,
P., Marshak, A., Matrosov, S. Y., Min, Q., O'Hirok, W., Wang, Z., and
Wiscombe, W.: Thin Liquid Water Clouds: Their Importance and Our Challenge,
B. Am. Meteorol. Soc., 88, 177–190,
10.1175/BAMS-88-2-177, 2007.Twomey, S.: Pollution and the planetary albedo, Atmos. Environ.,
8, 1251–1256, 10.1016/0004-6981(74)90004-3,
1974.
Wang, L., Qu, J. J., Xiong, X., Hao, X., Xie, Y., and Che, N.: A New Method
for Retrieving Band 6 of Aqua MODIS, IEEE T. Geosci. Remote Sens., 3, 267,
10.1109/LGRS.2006.869966, 2006.Warner, J.: The Water Content of Cumuliform Cloud, Tellus, 7, 449–457,
10.1111/j.2153-3490.1955.tb01183.x, 1955.Winker, D. M., Vaughan, M. A., Omar, A., Hu, Y., Powell, K. A., Liu, Z.,
Hunt,
W. H., and Young, S. A.: Overview of the CALIPSO Mission and CALIOP Data
Processing Algorithms, J. Atmos. Ocean. Tech., 26,
2310–2323, 10.1175/2009JTECHA1281.1, 2009.
Wood, R.: Relationships between optical depth, liquid water path, droplet
concentration, and effective radius in adiabatic layer cloud, University of
Washington, 3, 2006.Yang, Y., Marshak, A., Mao, J., Lyapustin, A., and Herman, J.: A method of
retrieving cloud top height and cloud geometrical thickness with oxygen A and
B bands for the Deep Space Climate Observatory (DSCOVR) mission: Radiative
transfer simulations, J. Quant. Spectrosc. Ra., 122, 141–149,
10.1016/j.jqsrt.2012.09.017, 2013.Zeng, S., Riedi, J., Trepte, C. R., Winker, D. M., and Hu, Y.-X.: Study of
global cloud droplet number concentration with A-Train satellites, Atmos.
Chem. Phys., 14, 7125–7134, 10.5194/acp-14-7125-2014, 2014.Zinner, T. and Mayer, B.: Remote sensing of stratocumulus clouds:
Uncertainties and biases due to inhomogeneity, J. Geophys. Res., 111,
D14209, 10.1029/2005JD006955, 2006.