AMTAtmospheric Measurement TechniquesAMTAtmos. Meas. Tech.1867-8548Copernicus PublicationsGöttingen, Germany10.5194/amt-10-3697-2017Comparison of the GOSAT TANSO-FTS TIR CH4 volume mixing
ratio vertical profiles with those measured by ACE-FTS, ESA MIPAS,
IMK-IAA MIPAS, and 16 NDACC stationsOlsenKevin S.ksolsen@atmosp.physics.utoronto.cahttps://orcid.org/0000-0002-2173-9889StrongKimberlyhttps://orcid.org/0000-0001-9947-1053WalkerKaley A.https://orcid.org/0000-0003-3420-9454BooneChris D.RaspolliniPierahttps://orcid.org/0000-0002-5408-1809PlieningerJohanneshttps://orcid.org/0000-0001-9096-5962BaderWhitneyhttps://orcid.org/0000-0003-0766-8460ConwayStephanieGrutterMichelhttps://orcid.org/0000-0001-9800-5878HanniganJames W.https://orcid.org/0000-0002-4269-1677HaseFrankJonesNicholasde MazièreMartineNotholtJustusSchneiderMatthiashttps://orcid.org/0000-0001-8452-0035SmaleDanSussmannRalfSaitohNaokoDepartment of Physics, University of Toronto,
Toronto, Ontario, CanadaDepartment of Chemistry, University of Waterloo, Waterloo, Ontario, CanadaIstituto di Fisica Applicata “N. Carrara” (IFAC) del Consiglio Nazionale delle Ricerche (CNR), Florence, ItalyInstitut für Meteorologie und Klimaforschung, Karlsruhe Institute of Technology, Karlsruhe, GermanyInstitute of Astrophysics and Geophysics, University of Liège, Liège, BelgiumCentro de Ciencias de la Atmósfera, Universidad Nacional Autónoma de México, Mexico City, MexicoAtmospheric Chemistry Division, National Center for Atmospheric Research, Boulder, CO, USACentre for Atmospheric Chemistry, University of Wollongong, Wollongong, AustraliaBelgisch Instituut voor Ruimte-Aëronomie-Institut d'Aéronomie Spatiale de Belgique (IASB-BIRA), Brussels, BelgiumInstitute for Environmental Physics, University of Bremen, Bremen, GermanyNational Institute of Water and Atmospheric Research Ltd (NIWA), Lauder, New ZealandCenter for Environmental Remote Sensing, Chiba University, Chiba, JapanKevin S. Olsen (ksolsen@atmosp.physics.utoronto.ca)9October20171010369737188January201727March201727July201721August2017This work is licensed under the Creative Commons Attribution 3.0 Unported License. To view a copy of this licence, visit https://creativecommons.org/licenses/by/3.0/This article is available from https://amt.copernicus.org/articles/10/3697/2017/amt-10-3697-2017.htmlThe full text article is available as a PDF file from https://amt.copernicus.org/articles/10/3697/2017/amt-10-3697-2017.pdf
The primary instrument on the Greenhouse gases Observing SATellite
(GOSAT) is the Thermal And Near infrared Sensor for carbon
Observations (TANSO) Fourier transform spectrometer (FTS).
TANSO-FTS uses three short-wave infrared (SWIR) bands to retrieve
total columns of CO2 and CH4 along its optical
line of sight and one thermal infrared (TIR) channel to retrieve
vertical profiles of CO2 and CH4 volume mixing
ratios (VMRs) in the troposphere. We examine version 1 of the
TANSO-FTS TIR CH4 product by comparing co-located
CH4 VMR vertical profiles from two other remote-sensing
FTS systems: the Canadian Space Agency's Atmospheric Chemistry
Experiment FTS (ACE-FTS) on SCISAT (version 3.5) and the European
Space Agency's Michelson Interferometer for Passive Atmospheric
Sounding (MIPAS) on Envisat (ESA ML2PP version 6 and IMK-IAA
reduced-resolution version V5R_CH4_224/225), as well as 16 ground
stations with the Network for the Detection of Atmospheric
Composition Change (NDACC). This work follows an initial
inter-comparison study over the Arctic, which incorporated a
ground-based FTS at the Polar Environment Atmospheric Research
Laboratory (PEARL) at Eureka, Canada, and focuses on tropospheric
and lower-stratospheric measurements made at middle and tropical
latitudes between 2009 and 2013 (mid-2012 for MIPAS).
For comparison, vertical profiles from all instruments
are interpolated onto a common pressure grid, and smoothing is applied
to ACE-FTS, MIPAS, and NDACC vertical profiles. Smoothing is needed
to account for differences between the vertical resolution of each
instrument and differences in the dependence on a priori profiles. The
smoothing operators use the TANSO-FTS a priori and averaging kernels
in all cases. We present zonally averaged mean CH4
differences between each instrument and TANSO-FTS with and without
smoothing, and we examine their information content, their sensitive altitude
range, their correlation, their a priori dependence, and the variability within
each data set. Partial columns are calculated from the VMR vertical
profiles, and their correlations are examined. We find that the
TANSO-FTS vertical profiles agree with the ACE-FTS and both MIPAS
retrievals' vertical profiles within 4 % (±∼40ppbv)
below 15 km
when smoothing is applied to the profiles from instruments with finer
vertical resolution but that the relative differences can increase
to on the order of 25 % when no smoothing is applied. Computed
partial columns are tightly correlated for each pair of data sets.
We investigate whether the difference between TANSO-FTS and
other CH4 VMR data products varies with latitude. Our study
reveals a small dependence of around 0.1 % per 10 degrees
latitude, with smaller differences over the tropics and greater
differences towards the poles.
Introduction
The Greenhouse gases Observing SATellite (GOSAT) was developed by
Japan's Ministry of the Environment (MOE), National Institute for
Environmental Studies (NIES), and the Japan Aerospace Exploration
Agency (JAXA), and it was launched in 2009 with an inclination of
98∘. The objectives of the GOSAT mission
include monitoring the global distribution of greenhouse gases,
estimating carbon dioxide (CO2) source and sink locations
and strengths, and verifying the reduction of greenhouse gas
emissions as mandated by the Kyoto Protocol. GOSAT carries two
instruments: the Thermal And Near infrared Sensor for carbon
Observations (TANSO) Fourier transform spectrometer (FTS) and the
TANSO Cloud and Aerosol Imager (TANSO-CAI). In this work we compare
TANSO-FTS measurements with those made by similar instruments in
order to validate its quality. Any biases in the data product need
to be well understood for it to be used by other researchers, and
their discovery may lead to improvements of future versions.
TANSO-CAI is a radiometer with four spectral bands that is able to
measure the cloud fraction in the field of view of TANSO-FTS
. TANSO-FTS is a nadir-viewing
double-pendulum FTS, whose
technical details are described in Sect. .
TANSO-FTS makes observations of infrared radiation emitted from the
Earth's atmosphere in four bands. Three bands are in the short-wave
infrared region and are used to measure total columns of CO2
and methane (CH4). The fourth channel is in the thermal
infrared (TIR) to provide GOSAT with sensitivity to the vertical
structure of CO2 and CH4.
This work follows , who compared Atmospheric Chemistry
Experiment (ACE) FTS version 3.5 (v3.5) and
TANSO-FTS TIR version 1 (v1) vertical profiles with those measured by a
ground-based FTS at the Polar Environment Atmospheric Research
Laboratory (PEARL) at 80∘ N in Eureka, Canada .
We employ a similar methodology, extend that study globally, and
include multiple ground-based FTSs that are part of the Network for
the Detection of Atmospheric Composition Change (NDACC;
). observed that, after
smoothing the ACE-FTS profiles using the TANSO-FTS averaging kernels
and a priori profiles, the difference is close to 0 above
15 km but that there is a bias at lower altitudes, where
TANSO-FTS retrieves more CH4, with a mean excess of
20 ppbv in the troposphere.
The data analyzed by are limited to a single location
characterized by cooler temperatures and lower humidity than lower
latitudes, and limited latitudinal transport.
Our objective is to investigate whether the results of
are local or hold at all latitudes and to provide additional
global validation of the TANSO-FTS
v1 CH4 data product.
FTS instruments used in the CH4 VMR vertical profile
comparisons presented herein.
a For NDACC instruments, the best achievable spectral resolution is listed here.
Operationally achieved spectral resolutions for NDACC instruments may be coarser.
b NDACC instruments use optical filters that reduce the effective spectral range when
making measurements. c MIPAS' spectral resolution is divided into four narrower bands.d The Altzomoni site came online in late 2012. e The Maïdo, Réunion site came online in early 2013.
In this manuscript, we examine the TIR data product from TANSO-FTS,
specifically, CH4 volume mixing ratio (VMR) vertical
profiles, by determining when TANSO-FTS TIR retrievals of CH4
were made in coincidence with those of other satellite-borne and
ground-based FTS instruments. Comparisons of satellite instruments
are made with the ACE-FTS on
SCISAT, described in Sect. , and the Michelson
Interferometer for Passive Atmospheric Sounding (MIPAS) on the
Environmental Satellite (Envisat), described in Sect. .
The NDACC InfraRed Working Group (IRWG) has a network of ground-based
FTSs; we used 16 that retrieve vertical profiles of CH4 VMR to
compare with the TANSO-FTS TIR data. The NDACC data are described
in Sect. . A summary of the instruments used in this
study is given in Table .
The question we are asking in this validation study is not, what
is the magnitude of the difference between retrieved CH4
vertical profiles from TANSO-FTS and other instruments, but: given
the vertical resolution, information content, and a priori dependence of
TANSO-FTS, would CH4 vertical profile retrievals derived from
another co-located instrument's measurements agree with those for
TANSO-FTS? To answer this question, a smoothing operator is applied to
the vertical profiles of the instruments with finer vertical resolution
(and therefore finer structure in the vertical profiles). This smoothing
operator, described by and presented in
Sect. , uses the a priori profiles and averaging
kernels from TANSO-FTS. In this study, results with and without smoothing
are presented (Sect. ).
For each comparison pair, the averaging kernels, information content,
and variability of the retrievals are examined in
Sects. and . The instrument with
finer vertical resolution is smoothed using the averaging kernels of
the instrument with coarser vertical resolution (TANSO-FTS in all
cases presented here) in order to account for the structure intrinsic
to a finer-resolution instrument. For each coincident pair, the absolute and
relative differences of the smoothed and unsmoothed VMR vertical
profiles are found, and their means, correlation coefficients, R2,
and numbers of coincident pairs are computed at each pressure level.
For each vertical profile in a coincident pair, an overlapping vertical
extent is selected using the sensitivity, or response, of the TANSO-FTS
retrieval (area of the averaging kernel matrix),
partial columns are computed over this range, and their correlations
are examined. Finally, this altitude range is used to estimate the
mean VMR difference taken over the vertical range for each coincident
pair of profiles. This data set shows any biases related to
latitude, or any other parameters of the TANSO-FTS retrieval, such as
incidence angle or surface type (land or water).
Section describes the methods and criteria for
determining coincident measurements between TANSO-FTS and each
instrument. Section provides a detailed description
of the comparison methodology. Comparison results for each
instrument are presented in Sect. . The satellite
instruments are zonally averaged, and each NDACC site is shown.
Partial column calculation
methodology is presented in Sect. , and correlation
results are shown in Sect. . A discussion follows in
Sect. , focusing on our investigation of biases within
the TANSO-FTS retrievals related to latitude and other parameters.
Data setsTANSO-FTS
TANSO-FTS makes measurements of radiance in four bands;
the TIR band is between 700 and 1800 cm-1 and
is used to retrieve vertical
profiles of CH4 VMRs. TANSO-FTS has a spectral resolution
of 0.2 cm-1 and operates in a nadir- or near-nadir-viewing geometry . To improve coverage, its field of
view sweeps longitudinally, and TANSO-FTS makes several measurements
along each cross track: five measurements prior to August 2010 and
three since then . This leads to TANSO-FTS having the
highest density of measurements and greatest spatial coverage among
the instruments considered herein.
Retrievals of v1 CH4 follow the nonlinear maximum a
posteriori method used for
v1 CO2 presented in . They are
performed on a fixed pressure grid, and the pressure levels are
adjusted based on the averaging kernels for the retrieval.
In the v1 retrieval algorithm, water vapour, nitrous oxide,
ozone concentrations, temperature, surface temperature, and surface
emissivity were retrieved simultaneously with CH4
concentration from V161.160 L1B spectra.
A priori data are based on simulated data from the
NIES transport model (TM; ), and the retrievals
use the HITRAN 2008 line list with several updates
up to 2011 .
An initial comparison of TANSO-FTS v1 to a single NDACC
station, Eureka, and to ACE-FTS measurements made in the Arctic
within a quadrangle surrounding PEARL (60–90∘ N and
120–40∘ W) has been recently made .
The v1 CH4 product was also compared globally with the
version 6 CH4 data product from the Atmospheric Infrared
Sounder (AIRS) on Aqua .
ACE-FTS
ACE-FTS was launched into low Earth orbit in 2003 on board the
Canadian Space Agency's (CSA's) SCISAT. The
scientific objectives of ACE are to study ozone distribution
in the stratosphere, the relationship between atmospheric chemistry
and climate change, the effects of biomass burning on the troposphere,
and the effects of aerosols on the global energy budget .
ACE-FTS is a high-resolution, double-pendulum FTS with a spectral
resolution of 0.02 cm-1 that covers a broad spectral
range between 750 and 4400 cm-1. It operates in solar
occultation mode, making a series of measurements for tangent altitudes
down to 5 km (or cloud tops) at local sunrise and sunset
along its orbital path . Its Level 2 data products are
vertical profiles of temperature, pressure, and the VMRs of 36 trace
gases, as well as isotopologues of
major species, reported on an altitude grid at the measurement
tangent altitudes or interpolated onto a 1 km grid.
Retrievals of the version 2.2 (v2.2) data product are described in
, and updates regarding the latest release, version 3.5
(v3.5), are described in . V3.5 retrievals, with the
data quality flags (v1.1) described in , are used herein.
When performing trace gas retrievals, tangent altitudes for each
observation and vertical profiles of temperature and pressure are
also retrieved using spectral fitting (not simultaneously).
Comparisons with TANSO-FTS are
made on a pressure grid using the retrieved pressure values at the
ACE-FTS measurement heights. A priori temperature and pressure
for ACE-FTS are derived from the NRLMSISE-00 model (MSIS; ) and from
meteorological data provided by the Canadian Meteorological Centre
with their Global Environmental Multiscale (GEM) model .
Fitted spectra are computed using the HITRAN
2004 spectral line list with modifications
described in .
Validation of v2.2 CH4 VMR vertical profiles is presented in
and was performed using several ground-based FTSs
that are part of NDACC, as well as one at Poker Flat. For that
comparison, partial columns were computed from the ACE-FTS
CH4 profiles, and the correlation between partial columns
computed from ground-based FTSs and from ACE-FTS was investigated.
Validation was also done against the balloon-borne SPIRALE
(Spectroscopie Infra-Rouge d'Absorption par Lasers Embarqués),
the Halogen Occultation Experiment (HALOE) on the Upper
Atmosphere Research Satellite, and MIPAS.
determined that the ACE-FTS v2.2 CH4 data are accurate
to within 10 % in the upper troposphere and lower stratosphere
and to within 25 % at high altitudes. More recently,
compared CH4 from the Canadian Middle Atmosphere Model (CMAM)
with measurements from ACE-FTS, the Sub-Millimeter Radiometer (SMR)
on Odin, and the Microwave Limb Sounder (MLS) on Aura, and they found
agreement with ACE-FTS within 30 %. Updates to the ACE-FTS
validation effort using v3.0 data and a description of the
differences between v2.2 and v3.0 are presented in .
found a slight reduction in CH4 VMR in the
v3.0 data near 23 km and a larger reduction of around 10%
between 35 and 40 km.
MIPAS
MIPAS is a limb-sounding FTS that was placed in polar
low Earth orbit in 2002 on board the European Space
Agency's (ESA's) Envisat. MIPAS aimed to provide global observations,
during both night and day, of changes in the
spatial and temporal distributions of long- and short-lived species,
temperature, cloud parameters, and radiance. The instrument was
intended to have a maximum spectral resolution of
0.025 cm-1, but the slide system for the
interferometer mirrors encountered a problem in 2004, and
observations used in this study were made with a reduced effective
spectral resolution of 0.0625 cm-1 but with finer
vertical sampling. Further complications arose in
2012, and ESA lost communication with Envisat, ending the mission.
The spectral range of MIPAS is 685–2410 cm-1, allowing
the retrieval of multiple trace gases. MIPAS spectra are processed
independently by four research groups . In
this paper, we consider two: the ESA operational analysis and the
Karlsruhe Institute of Technology Institute of Meteorology and
Climate Research (IMK) and the Instituto de Astrofísica de
Andalucía (IAA) analysis, both described in the following
subsections.
ESA MIPAS
We use MIPAS Level 2 Prototype Processor version 6
(ML2PP v6) of the ESA operational analysis. Early versions of the
ESA MIPAS gas retrievals are described in
(full-resolution Instrument Processing Facility version 4.61; IPF
v4.61), and the ML2PP v6 upgrades and reduced-resolution adaptations
are described in . Retrievals are made using
a global fitting scheme followed by a posteriori Tikhonov
regularization with self-adapting constraints .
The ML2PP v6 data provide retrieved VMR vertical profiles of
10 atmospheric gases between approximately 6 and 70 km.
Temperature and pressure are retrieved from the spectra at each
tangent point of a limb scan, and a corresponding altitude grid is
built from the lowest engineering tangent altitude using the equation
of hydrostatic equilibrium. Initial guesses for vertical profiles
of a target trace gas, temperature, and interfering species are
the weighted average of the results from the previous scan, an
appropriate merging of IG2 (initial guess 2) climatological profiles
and, if available, data from the European Centre for Medium-range
Weather Forecasts (ECMWF). Spectra are computed using a specialized
line list derived from HITRAN 1996 .
The IPF v4.61 CH4 data product has been validated by
against four balloon instruments; including
SPIRALE; three aircraft instruments; six ground-based FTSs (all
are considered herein), and HALOE. They found good agreement
with a 5 % positive bias in the lower stratosphere and upper
troposphere. ML2PP v6 CH4 was compared with BONBON
air sampling measurements by . The
reduced-resolution CH4 measurements (2005–2012) agree
with in situ data within 5–10 %. CH4 (and N2O)
from ESA MIPAS has been assimilated by the BASCOE code, and the
assimilated products have been compared with MLS and ACE-FTS
. The analysis has proven the high quality of
the MIPAS data, but it has also identified the presence of some
outliers, especially in the tropical lower stratosphere, and some
discontinuities due to issues in the measurements.
IMK-IAA MIPAS
The IMK-IAA MIPAS retrieval algorithm has been developed to include
and account for deviations from local thermal equilibrium. The
data presented here are IMK-IAA reduced-resolution version
V5R_CH4_224/225. The early retrieval algorithms are described
by , and the updates made to the current version
are described by . Temperature and tangent
altitude are retrieved from the spectra, and pressure is computed
from the equation of hydrostatic equilibrium. V5R_CH4_224/225
uses the HITRAN 2008 line list . Temperature
a priori profiles are determined from ECMWF analyses and MIPAS
engineering information. The IMK-IAA retrieval uses Tikhonov
first-order regularization in combination with an all-zero
CH4 a priori profile, which serves to smooth the
profiles.
Validation of the IMK-IAA MIPAS V5R_CH4_222/223 data has been
presented in . They compare data against ACE-FTS,
HALOE, the MkIV balloon FTS, the Solar Occultation For Ice
Experiment (SOFIE) on the Aeronomy of Ice in the Mesosphere (AIM)
satellite, the SCanning Imaging Absorption spectroMeter for
Atmospheric CHartographY (SCIAMACHY) on Envisat, and a cryogenic
whole-air sampler (collects gas bottle samples during aircraft flights).
They found an agreement within 3 % in the upper stratosphere with
other satellite instruments, but in the lower stratosphere (below
25 km) a high bias was found in the MIPAS retrievals of
up to 14 %. The V5R_CH4_224/225 has more recently been
validated by , using ACE-FTS, HALOE, and
SCIAMACHY. They found MIPAS CH4 retrievals to be larger
by around 0.1 ppmv below 25 km, or around 5 %.
NDACC
NDACC is a global network of a variety of instruments that provides
measurements of tropospheric and stratospheric gases that are directly
self-comparable . The network consists of over 70 stations
sparsely distributed at all latitudes. Information about NDACC
is available
at www.ndacc.org. In this work, we only consider a small subset of
NDACC stations that feature high-resolution FTSs and provide a
CH4 VMR vertical profile data product via the NDACC database.
demonstrated the good quality of
CH4 profiles that can be retrieved from the NDACC FTS measurements.
The stations are listed in Table , along with their
locations, spectral range and resolution, and references.
The stations do not use identical instruments, spectroscopic lines, or
retrieval methods. All but one station use a version of a Bruker 120/5 M
or HR and have predominantly adopted, or upgraded to, the Bruker 125HR.
Some stations have more than one instrument, and the type of instrument
has changed over time at many of the stations. Toronto,
43.6∘ N, uses a Bomem DA8.
Retrievals are generally performed using either PROFFIT
or SFIT4 following harmonized
retrieval settings recommended by the NDACC IRWG
. Data used herein are from the NDACC
database. A summary of retrieval settings is
provided by . Lauder and Arrival Heights, at
45.0 and 77.8∘ S, respectively, use a retrieval strategy that
adheres to that defined in , with a relaxed Tikhonov
regularization constraint at Arrival Heights due to the characteristic
atmospheric dynamics over Antarctica. Jungfraujoch, at 46.6∘ N,
uses SFIT2. It has been established within the NDACC IRWG that the
regularization strength of the CH4 retrieval strategy should be
optimized so that the number of degrees of freedom for signal
(DOFS) is limited to approximately 2 .
Data set variability
To provide context for the VMR differences found when comparing each
instrument to TANSO-FTS, shown in Sect. , we have
examined the variability of retrievals made for each instrument.
We are interested in determining whether the mean differences found when
comparing TANSO-FTS to another instrument are comparable to the
differences found when comparing pairs of retrievals for a single
instrument. Each pair of observations compared in this study is
made at different times and locations and subject to instrument noise and
analysis errors. Examining the variability within each data set
provides an indication of the magnitude of these effects. Because the
observation geometries and rates of spectral acquisition are different
for each instrument, our internal comparisons differ for each instrument.
For example, TANSO-FTS and MIPAS have a much higher data density than
ACE-FTS, which only makes two sets of observations per orbit.
Following , we are aware that TANSO-FTS CH4
retrievals are dependent on the a priori used, especially at
high altitudes. TANSO-FTS vertical profiles tend to be similar to
their a priori and, therefore, to each other. To provide context for
our validation results, we computed the magnitude of the mean
differences between the TANSO-FTS retrievals and their a priori.
This is indicative of the instrument sensitivity discussed in
Sect. and shows by how much the retrievals deviate
from the a priori. We examined 3000 randomly selected TANSO-FTS
measurements by interpolating the a priori and retrieved profiles to
the pressure grid used in our comparisons (Sect. ) and
then computed the difference between the retrieval and the a priori
at each pressure level, and their mean and standard deviation.
Figure shows the mean ±1 standard
deviation of the difference between the TANSO-FTS CH4
retrievals and their corresponding a priori profiles. The peak value
is 30 ppbv near 10 km (∼ 1.5 %)
with a standard deviation of the same magnitude.
Results for investigating the variability within each
CH4 VMR profile data set. Shown are the following
comparisons: TANSO-FTS retrievals compared to their a priori
(green), pairs of sequential ACE-FTS retrievals (red), ESA
MIPAS retrievals compared to IMK-IAA MIPAS retrievals made for the
same limb observations (blue), and pairs of NDACC retrievals
made on the same day (orange). All retrieved profiles used
are coincident with TANSO-FTS.
Dashed lines are 1 standard deviation.
To examine the variability of the ACE-FTS CH4 data
product, we compared each retrieved profile from an ACE-FTS
sunset/sunrise (occultation direction) to that from the next
orbit, taking care to avoid
a comparison between sunset and sunrise occultations (which are
in different hemispheres), or when an acquisition was not recorded
during a subsequent orbit.
Considering all sunset occultations in 2011, there were 1402
retrieved vertical profiles,and 820 sequential pairs. These pairs
are separated by 97 min and have a mean spatial separation of
1180±20km, depending on the latitude of the
measurement. For each pair, we computed the VMR difference on
the ACE-FTS 1 km tangent altitude grid and then
found the mean and standard deviation, which are shown in
Fig. . Within the ACE-FTS data, the largest
systematic variability (-4ppbv) occurs around
30 km, with extreme outliers being observed at the lowest
tangent altitudes. The mean magnitude of the ACE-FTS variability
is 2 ppbv (0.1 %) at all altitudes and
9 ppbv below 15 km (0.4 %).
To examine the variability of the MIPAS data sets, we compared the
vertical profiles retrieved by IMK-IAA and ESA that were made from
the same MIPAS limb observations and within our coincident
data set. This provides an indication of the impact of different
retrieval algorithms on retrieved profiles.
For each pair of retrieved vertical profiles from a
single set of MIPAS spectra, we interpolated the ESA retrieval to
the IMK-IAA 1 km grid and computed their difference
(IMK-IAA-ESA), and then found the mean and standard
deviation. Figure shows the mean ±1 standard
deviation for this comparison. The two retrievals show good agreement
above 30 km (not shown), while the IMK-IAA data have a
positive bias relative
to the ESA data product of around 0.15 ppmv between 20 and
30 km. This bias is consistent with the validation
results presented in . The ESA and IMK-IAA
comparison exhibits the largest variability, with a mean magnitude
(mean of absolute values) of 50 ppbv (2 %) for the
altitude range considered (9–34 km). Since the two products
use the same spectra, it is possible that part of the internal
instrument variability is hidden in this approach.
To investigate the variability of the NDACC data, we compared pairs
of observations made at an NDACC site on the same day. We considered
only NDACC CH4 VMR vertical profiles that were in coincidence
with TANSO-FTS. For each pair of NDACC measurements, we computed
the CH4 VMR differences on the standard NDACC retrieval grid
(earlier profile minus later profile; if there are multiple coincidences
in a day, differences are found relative to the earliest). The mean
and standard deviation of these differences are also shown in
Fig. . When examining several measurements from the
same day, the NDACC differences show a systematic mean increase
in tropospheric CH4 with time during a single day. This
variability is small, however, with a mean of -4ppbv
below 30 km and a peak at 12 km of
-6ppbv (0.3 %).
Our variability investigation found that
the ACE-FTS data exhibit the smallest variability between
measurements, that MIPAS exhibits the largest, and that NDACC and
TANSO-FTS are of similar magnitudes. The magnitude of the internal
variability of the data sets is between ±2ppbv (e.g., for
NDACC and ACE-FTS in the upper troposphere) and ±3ppbv,
or around 2 % (e.g., for TANSO-FTS and the lower limits of ACE-FTS).
Coincidences
Due the coverage and data collection rates of each
instrument, different coincidence criteria were used.
ACE-FTS has an inclination of 74∘ and operates in
solar occultation mode, recording only two occultations per
orbit, predominantly at high latitudes; the NDACC sites are
stationary; MIPAS makes frequent observations at all latitudes;
and the spatial distribution of TANSO-FTS observations is
enhanced by its cross-track observation mode. In the
case of ACE-FTS and NDACC stations, the objective of the
coincidence criteria was to maximize the number of measurements
used. Conversely, in the case of MIPAS, the objective was to reduce
the number of potential coincident measurements. For ACE-FTS
and NDACC, we sought measurements made within 12 h
and within 500 km of each TANSO-FTS measurement (spatial
separation calculated using the Vincenty method ).
For the MIPAS data sets, we sought measurements made within
3 h and 300 km. When searching for
MIPAS–TANSO-FTS coincidences within 12 h and
500 km, we find approximately 180 000 coincidences per
month.
The criteria used in this study are comparable to previous
CH4 studies. For example, used
criteria of 24 h and 1000 km when comparing
ACE-FTS CH4 to ground sites, and 6 h and
300 km when comparing ACE-FTS to MIPAS.
used criteria of 3 h and 300 km when comparing
MIPAS CH4 to ground- and satellite-based spectrometers.
used criteria of 9 h and 800 km
when comparing MIPAS CH4 to ACE-FTS, and 24 h
and 1000 km when comparing MIPAS to HALOE.
TANSO-FTS CH4 VMR vertical profiles tend not to be
sensitive above the upper troposphere (see Sect. ),
while ACE-FTS and MIPAS retrievals have a limited vertical
extent in the troposphere. To ensure that measurements made
by each instrument overlap, a restriction was placed on
ACE-FTS and MIPAS measurements: that their retrieved vertical
profiles extend to low enough altitudes, after applying data quality
criteria. For ACE-FTS, this requirement was 10 km. For
MIPAS, this requirement was relaxed to less than
12 km. IMK-IAA MIPAS CH4 VMR vertical profile
retrievals do not extend as low as those made by ESA, to the extent
that having the same restriction on altitude range results in only
a quarter as many coincidences as the ESA data product. Relaxing
the constraint to only 12 km maintains the assurance that
retrieved VMRs will overlap with the TANSO-FTS altitude range,
though there are only 60 % as many IMK-IAA coincidences as ESA coincidences.
TANSO-FTS makes nadir observations in a grid pattern by sweeping
its line of sight across its ground-track. This results in a
high density of vertical profiles, such that – for a single
observation made by ACE-FTS, MIPAS, or NDACC – there are an
average of 11 coincident TANSO-FTS measurements. The subsequent
measurement made by MIPAS or an NDACC station will be
coincident with a similar number of TANSO-FTS measurements,
and most of those will also be coincident with the previous
MIPAS or NDACC measurement. A common way to deal with multiple
coincidences is to take the mean of the VMR vertical profiles from
each instrument and to compute the difference of the means
e.g.,. When comparing MIPAS to TANSO-FTS,
however, this results in some measurements contributing
to the analysis more times than others, biasing the
computed VMR difference profiles. Furthermore, this leads to
using a mean TANSO-FTS VMR vertical profile that is strongly
smoothed, while a coincident ACE-FTS (or NDACC, depending on
the station's rate of acquisition at the time) VMR vertical
profile is not.
To reduce biases caused by over-counting, when comparing TANSO-FTS
to MIPAS, and by smoothing, when comparing TANSO-FTS to ACE-FTS,
we reduced the number of coincident measurements by seeking
a set of one-to-one coincidences for unique measurements in the
sparser data set (which is always ACE-FTS, MIPAS, or NDACC).
For each measurement that is being compared to TANSO-FTS, we
find the TANSO-FTS measurement with the minimum of the sum of
ratios of distance in space and time to the coincidence criteria,
giving equal weight to both parameters as
min(dx/xcrit+dt/tcrit), where dx and dt are the
distance and time between a given measurement and a TANSO-FTS
coincidence, and xcrit and tcrit are the coincidence
criteria. This method is similar to using a standard score to compare
the spatial and temporal separation, but the sample size of the
set of TANSO-FTS measurements coincident with another measurement
is on the order of only 10. Furthermore, the mean and standard
deviations of dx and dt reflect the time and distance between
each consecutive TANSO-FTS measurement, rather than the time and
spatial separation between each TANSO-FTS measurement and those
from MIPAS, ACE-FTS, or NDACC.
Number of coincident CH4 VMR vertical profile
measurements that were found between TANSO-FTS retrievals
and those from ESA MIPAS, IMK-IAA MIPAS, ACE-FTS, and NDACC stations.
The three columns show the total number of coincidences found, the
number of unique TANSO-FTS measurements within those coincidences,
and the size of the reduced one-to-one coincidences used.
* The Izaña NDACC coincidence data set
is the only one in which TANSO-FTS measurements are more sparse.
For consistency, Izaña was not treated as a special case.
Table shows the total number of coincidences
found between TANSO-FTS and each validation target instrument, as
well as the subsets of unique TANSO-FTS measurements and the
one-to-one coincidences used in this paper (equivalent to the number
of unique measurements made by each target instrument).
Figure shows an example of the global distribution
of coincident measurements. Shown are the first 200 one-to-one
coincidences after 1 January 2012. For the ESA and IMK-IAA MIPAS
data products, this number of coincidences is found in around 2
weeks. For ACE-FTS and the NDACC stations (combined), these
coincidences occur over several months.
Locations of the first 200 observations of 2012 used in this
study for TANSO-FTS (green), ACE-FTS (red), IMK-IAA MIPAS
(blue), and ESA MIPAS (purple). The NDACC stations are shown in orange.
Averaging kernels
The averaging kernels of a profile retrieval provide information
about the contributions of the retrieval from a priori information
and the measurements. In this study, the retrieval methods for each
data set differ, and the averaging kernel matrices are differently
defined. In general, the rows of the averaging kernel matrix are peaked
functions whose full width at half maximum (FWHM) can be used to
define the vertical resolution of the
measurement. The sum of the rows of the matrix gives the sensitivity,
or response, of the retrieval. A sensitivity close to 1 indicates that
most of the information in the retrieval comes from the measurement,
while sensitivities less than 1 indicate increased reliance on the
a priori in the solution.
The rows of the averaging kernel matrices for the ESA MIPAS, IMK-IAA
MIPAS, TANSO-FTS, and the Eureka NDACC station are shown in
Fig. . Each panel shows the mean from 30 retrievals.
Vertical profiles of pressure associated with each retrieval's
averaging kernel matrix are, in general, unique, so a common
pressure grid was selected for each instrument, and averaging
kernels were interpolated prior to averaging.
Example of averaging kernels for (a) TANSO-FTS, (b) IMK-IAA
MIPAS, (c) ESA MIPAS, and (d) NDACC. Each kernel shown is the
mean from 30 averaging kernel matrices from measurements made over
the Arctic, interpolated to a common pressure grid. Panel (d) shows
the mean averaging kernels from the Eureka station. Panel (e) shows
the sensitivity for the mean averaging kernels shown in each panel:
TANSO-FTS (green), IMK-IAA MIPAS (blue), ESA MIPAS (purple), and
NDACC (orange).
In this study, we treat TANSO-FTS retrievals as having the coarser
vertical resolution in all cases, despite the widths of the kernel
functions shown in Fig. a, which are comparable to
MIPAS and narrower than NDACC.
The peak locations of the TANSO-FTS averaging kernels do not match
the corresponding pressure level of each kernel. Therefore the
FWHM values when considering the location
of the appropriate pressure level are much larger than the
FWHM values for the averaging kernels of the
other instruments.
In the NDACC retrievals, the a priori has a large role, and information
coming from the measurements can hardly distinguish the contribution
coming from the different altitudes. This leads to wide, overlapping
averaging kernels. The IMK-IAA MIPAS retrievals use a form of Tikhonov
regularization without an a priori. The ESA MIPAS retrievals use the
regularizing Levenberg–Marquardt approach (where the parameter setting
has been chosen to leave results largely independent from the
initial-guess profiles) and a posteriori Tikhonov regularization without
an a priori. The ACE-FTS retrievals do not use a regularized matrix
inverse method. Consequently, the ACE-FTS and IMK-IAA MIPAS averaging
kernels are
very narrow, their peak values are close to 1 at each altitude where a
spectrum was acquired, and the solutions do not rely on a priori
information. Very similar averaging kernel are obtained also for ESA
MIPAS, with wider widths at lower altitudes where the retrieval grid
used is coarser than the measurement grid. The sensitivity of both
ACE-FTS and MIPAS, shown in Fig. e, is close to 1
at all altitudes, falling off above 60 or 70 km. ACE-FTS
averaging kernels are under development, and preliminary work is shown
in .
The typical sensitivity of an NDACC retrieval is close to unity
until above 20 km, falling off towards 0 through
60 km. The sensitivity of TANSO-FTS only reaches 0.2–0.3
between 5 and 10 km. The implication of such low values
for sensitivity is that the TANSO-FTS retrievals are highly dependant
on their a priori.
Degrees of freedom for signal for, from top to bottom,
TANSO-FTS, IMK-IAA MIPAS, ESA MIPAS, and NDACC. Each satellite
(and panel) uses a different symbol and colour, but the colour
shades indicate the year the measurement was made in. The
TANSO-FTS and IMK-IAA
MIPAS measurements shown are in coincidence. The ESA MIPAS and
NDACC data are from our analyzed data set but not in coincidence
with the TANSO-FTS data in the top panel. All data are from
the Arctic, 90–60∘ N, with the NDACC measurements from
Eureka, Ny Ålesund, and Thule.
The trace of the averaging kernel matrix gives the DOFS. For
example, DOFS for retrievals made by TANSO-FTS, IMK-IAA
MIPAS, ESA MIPAS, and NDACC from observations over the Arctic,
above 60∘ N, are shown in Fig. .
The IMK-IAA MIPAS and TANSO-FTS data are in
coincidence with one another. The NDACC data come from Eureka,
Ny Ålesund, and Thule. The NDACC and ESA MIPAS data shown are the
TANSO-FTS one-to-one coincidences used throughout this study (but are
not coincident with the TANSO-FTS data shown in the top panel of
Fig. ). The trends visible are seasonal and are related
to opacity and water vapour content. Recreating this figure over
mid-latitudes or the tropics reveals a flat trend over time, while
over Antarctica the trends are reversed in DOFS space.
The mean of the DOFS for the three NDACC stations over the Arctic is
1.98 with a standard deviation, σ, of 0.50. Over the tropics,
considering data from Izaña, Réunion St. Denis, Altzomoni,
and Mauna Loa (Réunion Maïdo only has data from 2013
onward, not shown here), the mean is 2.39 with σ=0.37. The
mean DOFS for IMK-IAA MIPAS are slightly larger than those
for ESA MIPAS. Over the Arctic, their means and standard deviations
are 17.05 and σ=1.06 for IMK-IAA, and 15.76 and σ=0.93 for
and ESA, respectively. Over the tropics, they are 16.10 and
σ=0.33, and 15.88 and σ=1.20.
The TANSO-FTS DOFS are larger at low latitudes, with a mean over the
tropics of 0.72 and σ=0.08, and means over the Arctic and
Antarctic of 0.32 and 0.20, respectively (σ=0.13 and 0.12).
The DOFS for a TANSO-FTS retrieval rarely go above unity. Conversely,
in the coincident NDACC data discussed above, over the tropics and
Arctic, the DOFS never fall below unity. Note that the averaging
kernel matrices for TANSO-FTS, and therefore the DOFS, cover a much
smaller altitude range than for NDACC and MIPAS, which can extend
above 100 km.
VMR vertical profile comparisonsMethodology
Retrievals made by an instrument with fine vertical resolution may
result in structure over its vertical range that is not
distinguishable in retrievals made by an instrument with coarser
vertical resolution.
In order to make the best comparison between two instruments with
differing vertical resolution, it is necessary to smooth the vertical
profiles retrieved from the finer-resolution instrument, in order
to simulate what we could infer from it if it had a sensitivity similar
to that of the other instrument. Smoothing is done using
the a priori CH4 VMR vertical profiles and averaging kernel
matrices of the instrument with lower vertical resolution
:
x^s=xa+A(x^-xa),
where x^
is original higher-resolution retrieved
profile, x^s is the smoothed profile, xa
is the a priori profile of the lower-resolution retrieval, and
A is the averaging kernel matrix of the lower-resolution
retrieval. xa and A are from the TANSO-FTS
retrieval in all cases presented here. The smoothed profile,
x^s, approximates the a priori, xa, when
either the rows of A
are close to 0, or when the retrieval is close to xa.
As can be inferred from Fig. a, above 20–25 kmx^s∼xa.
In order to apply Eq. (), all the variables on the
right-hand side must be interpolated to a common grid. TANSO-FTS retrievals
are done on a retrieved pressure grid. Determining the altitude of its
VMR vertical profiles requires applying the equation of hydrostatic
equilibrium and incorporating a priori temperature and water vapour.
Since pressure is retrieved by ACE-FTS and MIPAS, and the tropospheric
a priori pressure profiles and measured surface pressure are accurate
for NDACC , all comparisons here have been done
on a common pressure grid, as opposed to an altitude grid.
The data products do not always overlap over the entire pressure range
of the common grid. Extrapolation is needed to ensure that the length
of x^ matches the dimensions of A in
Eq. (). For ACE-FTS and MIPAS, we use xa to
extend their retrieved profiles below their altitude range to cover the
full pressure range of the TANSO-FTS averaging kernels. The averaging
kernels at these non-overlapping pressure levels do not contribute to
the smoothed retrieval at higher, overlapping levels.
The following steps are taken to compute vertical profiles of the
mean CH4 VMR differences:
appropriate instrument data quality flags are
applied to each VMR vertical profile in the coincidence pair;
TANSO-FTS a priori and validation target VMR vertical
profiles are interpolated to the TANSO-FTS retrieval pressure
grid;
the interpolated validation target profile is extended as
needed to match the TANSO-FTS pressure range (and vector length)
using the TANSO-FTS a priori;
the interpolated validation target profile is smoothed using
the TANSO-FTS averaging kernel matrix using Eq. ();
TANSO-FTS-retrieved and validation-target-smoothed
VMR vertical profiles are interpolated to a standard
pressure grid, and levels outside the pressure range of the
target's VMR profile are discarded;
the piecewise difference between the
TANSO-FTS and the smoothed validation target VMR vertical
profiles is found;
the means, standard deviations, and
correlation coefficients of the VMR differences are calculated
at each level of the standard pressure grid for all coincidences
within a latitude zone.
For comparison, mean VMR vertical profile differences were also computed
without smoothing by using only steps 1, 5,
6, and 7. Zonally averaged VMR
difference profiles are presented in Sect. , and results
obtained without applying smoothing to the validation targets
are shown in Sect. . The data quality flags in
step 1, referring to variables in the data product
files, were, for TANSO-FTS, CH4ProfileQualityFlag must be 0;
for ACE-FTS, quality_flag must be 0 and cannot be equal to
4, 5, or 6 at any altitude;
for ESA MIPAS, ch4_vmr_validity must be 1, and pressure_error
cannot be NaN (not a number); and for IMK MIPAS, visibility must be 1, and
akm_diagonal must be greater than 0.03.
found that identifying and removing coincident
CH4 VMR vertical profile pairs that may have one or both
profile locations within a polar vortex, and then filtering these events, had little
effect on their vertical profile comparisons below 25 km.
Polar vortex event will have a much smaller effect on this study
since it uses global and year-round data sets. For these two reasons,
our method does not filter for profiles located within a polar vortex.
Arrival Heights may be differently affected by a much stronger
Antarctic polar vortex, but comparison results from this site are
not anomalous and only account for 1.5 % of the NDACC data set,
so they are treated in a consistent manner.
Zonally averaged comparison results. The rows present results
for each zone, from top to bottom: 90–60∘ N,
60–30∘ N, 30∘ N–30∘ S,
30–60∘ S, and
60–90∘ S. In each row, the four panels show,
from left to right, the mean CH4 VMR difference between
retrievals from TANSO-FTS and the validation target at each
pressure level; the mean CH4 VMR differences relative to the
mean CH4 VMR vertical profile of the validation target;
the correlation coefficients R2 of the CH4 VMR differences
for each coincident pair at each pressure level; and the number of
coincidences at each pressure level. Differences are calculated as
TANSO-FTS minus target for each data set compared. In all frames,
ACE-FTS is
shown in red, ESA MIPAS is purple, IMK-IAA MIPAS is blue, and NDACC
stations are shades of orange. Each individual NDACC station with a
zone is shown, and their shades indicated.
Zonally averaged VMR profile differences
Following , we are trying to determine whether there are
any zonal biases in the TANSO-FTS data or zonal dependencies
when making comparisons to other instruments. The mean CH4
VMR differences, averaged zonally, between the TANSO-FTS
vertical profiles and the smoothed vertical profiles from ACE-FTS,
IMK-IAA MIPAS, ESA MIPAS, and each NDACC station are show in
Fig. . Each row in Fig. shows the
results from five latitudinal zones: 90–60∘ N,
60–30∘ N, 30∘ N–30∘ S,
30–60∘ S, and 60–90∘ S.
The left-most column shows the mean differences between the
retrievals from TANSO-FTS and those from the other instruments,
always calculated as TANSO-FTS minus target. One standard
deviation is shown for each instrument comparison with dotted
lines. The middle-left column shows the mean differences as a
percentage of the mean CH4 VMR vertical profile taken
for the target validation instrument in each zone. The number of
VMR measurements used in the mean at each altitude, for each
comparison, is shown in the right-most panel, with ESA MIPAS
always having the most. At each altitude, we also calculated the
Pearson correlation coefficient between the set of TANSO-FTS
CH4 VMR measurements and the coincident set from each
validation instrument. These are shown in the middle-right column
for each panel in Fig. .
For each zone, the mean difference tends towards 0, and the
standard deviation falls off above 100 hPa. This is a
reflection of the TANSO-FTS sensitivity. Above this altitude,
the TANSO-FTS averaging kernels tend to 0, as shown in
Fig. , and the smoothed profiles from each
target instrument begin to approximate the TANSO-FTS a priori.
Likewise, the TANSO-FTS retrieval above this pressure level is also
close to its a priori. Conversely, the number of CH4
VMR measurements in the mean falls off sharply below
10–12 km, or around 80–90 hPa, for the
comparisons to the satellite instruments. For the satellite
instruments and many of the NDACC stations we see the same trend:
a positive bias (TANSO-FTS VMRs are greater than those
of the validation instruments) decreasing with increasing
altitude, with a tropospheric mean of around 20 ppbv, or
1 %. The bias is smallest for the two MIPAS data products
in the tropics, between 30∘ N and 30∘ S.
The bias relative to ACE-FTS is consistent in all the zones. For
three of the NDACC stations – Ny Ålesund, Bremen, and Toronto –
there is a negative bias (TANSO-FTS retrieves less CH4 than
these stations), and for Eureka and Jungfraujoch the bias is close
to 0.
There is a notable feature just below 100 hPa in all the
zones except 30–60∘ S. This feature is a pronounced increase
in the mean difference in the northern zones 60–30∘ N and
90–60∘ N, while it is a decrease in the mean difference
between 30∘ N and 30∘ S and between 60 and
90∘ S. It is around this pressure level,
or altitude, that the VMR of CH4 begins to fall off rapidly
from between 1.8 and 2 ppmv in the troposphere towards
0 ppmv in the upper stratosphere and mesosphere. This
feature indicates that the altitude at which this VMR decrease occurs
differs between instruments. In the Northern Hemisphere this
decrease in CH4 VMR occurs at higher altitudes for
TANSO-FTS than for the other instruments, and in the tropics and
Southern Hemisphere this decrease occurs more rapidly and at
lower altitudes for TANSO-FTS.
For all instruments and in all zones, the correlation coefficients,
R2, at each altitude fall off very sharply, to around 0.2,
below the 90 hPa level (and remain higher in the tropics).
This indicates that biases seen in the mean differences are not uniform across
the coincident data set and that there is significant variability
in the magnitudes of the differences for individual vertical
profile pairs and in the direction of the difference. This
is related not only to the increasing standard deviation of the differences
with decreasing altitude but also to the standard deviations of each
data product in the comparison. The sharpness and altitude of the
decrease are directly related to the TANSO-FTS averaging kernels.
Above the 100 hPa level, the standard deviations of the
TANSO-FTS and
the smoothed validation target fall off very sharply as they both
begin to approximate the a priori (which also explains why R2 is
close to 1).
Averaged comparison results, as in each panel of
Fig. , for all latitudes, without applying smoothing
to the validation instruments' CH4 VMR
vertical profiles. Differences are calculated as
TANSO-FTS minus target for each data set compared
(and ACE-FTS-ESA MIPAS for that case).
Impact of smoothing
This study was also performed without applying any smoothing to
the vertical profiles of the target validation instruments. These
results are shown in Fig. , which has the same
panels as Fig. . The data have not been separated
zonally, and the plots show means for all latitudes. No zonal
biases were observed in the unsmoothed data. The
16 NDACC stations have been combined into a single data set.
Figure shows the mean differences
between the TANSO-FTS data product and those of other instruments,
and the behaviour of the comparisons at higher altitudes when
the validation targets are unaffected by the TANSO-FTS averaging
kernels. Without the smoothing applied, the difference profiles in
Fig. show more consistent behaviour
over the pressure, or altitude, range shown. While the magnitude
of the differences is much greater without smoothing, it is not
consistently biased high or low for all the data products at all
altitudes. When comparing to the satellite instruments in the
upper troposphere, we find that the TANSO-FTS retrieval has
greater CH4 VMRs by around 50 ppbv, or around
3 %.
For context, a comparison between the ACE-FTS and ESA MIPAS data
products, using profiles that were coincident with the same TANSO-FTS
observation, is shown in grey. The mean differences between these
two data products are smaller than those relative to TANSO-FTS but
have comparable standard deviations and a slightly smaller correlation,
with R2=0.5 and 0.6 in the upper troposphere.
The comparison between TANSO-FTS and NDACC extends below the range
of ACE-FTS and MIPAS. NDACC and TANSO-FTS agree very well in this
region, between ±30ppbv, or between ±2 %.
In this case, the NDACC stations retrieve more CH4, on
average. The low-altitude NDACC and TANSO-FTS data are also more
closely linearly correlated, between 50 and 60 %.
It should also be noted that the standard deviation of the TANSO-FTS
and NDACC differences is also less than those for ACE-FTS and MIPAS
at all altitudes.
Partial column comparisonsMethodology
For each CH4 VMR vertical profile in a pair of coincident
measurements, we computed a partial column and compared those from
TANSO-FTS to each of the other instruments to investigate how well
correlated the derived CH4 abundances are. For consistency,
each pair of partial columns must be calculated over the same pressure
range, as the number of molecules in the column strongly depends on
the altitude range (length of the column) of the integral. To
determine the pressure range over which to compute partial columns
for each coincident pair of profiles, we considered
the TANSO-FTS averaging kernels.
Two-dimensional histogram showing the number of TANSO-FTS
CH4 VMR profiles within our data set (z axis) that have
some number of usable pressure levels (y axis) with a sensitivity
greater than some given threshold, s (x axis). The data set
shown here consists of all TANSO-FTS observations that are one-to-one
coincident with a target validation data set. The threshold chosen for
this study was s=0.2.
We investigated the sensitivity of the TANSO-FTS retrievals, as
defined in Sect. to find an altitude range which
minimizes the partial column dependence on a priori information,
ensuring our investigation is focused on retrieved information
from TANSO-FTS. Figure shows a two-dimensional
histogram of the number of TANSO-FTS profiles, for all validation
targets combined for two criteria: setting a requirement that the
sensitivity must be greater than some threshold and the resulting
number of usable pressure levels in the integral for each profile.
We see that the maximum number of usable levels falls off in an
approximately linear manner with increasing sensitivity threshold, and
that for any sensitivity threshold there will be a large number of
TANSO-FTS CH4 VMR vertical profiles that never meet the
sensitivity criteria. Increasing the sensitivity cutoff by
0.05 causes approximately 10 000 additional TANSO-FTS vertical
profiles, or around 6 % of the total data set combining all
validation targets, to fail to meet the requirement at any altitude.
The number of usable pressure levels given a restriction on
sensitivity is not normally distributed, as can be inferred from the
empty area in the upper right of Fig. .
For this study, we have selected a sensitivity threshold of 0.2 and
require a minimum of three integrable pressure levels.
Approximately 23 % of the TANSO-FTS retrievals do not meet these
criteria. In such a case, partial columns are still computed using
three pressure levels surrounding the level with the maximum sensitivity
that are within the range of the target profile (e.g., not below
10 km when comparing to ACE-FTS). These excluded data do
not exhibit a broader distribution, but their computed partial columns
are all very small due to the integration range. Because the overlapping
altitude regions for NDACC and TANSO-FTS measurements extend much
lower in the atmosphere than for ACE-FTS and MIPAS, the number of
TANSO-FTS profiles that do not meet the sensitivity criteria is much
smaller for NDACC.
Partial columns are computed as
column=∫z1z2P(z)kT(z)χ(z)dz,
where z1 and z2 bound the integration range over altitude z,
P is pressure, T is temperature, χ is the CH4 VMR,
and k is the Boltzmann constant. For each instrument, χ(z)
is the retrieved quantity, and either retrievals were performed
on a pressure grid or pressure was retrieved simultaneously.
We compute partial columns from vertical profiles after step 5 in Sect. , so both the TANSO-FTS
and the smoothed validation target profiles have the same
pressure at each level in the integration. Since TANSO-FTS
retrievals do not have an altitude grid, we use that of the
coincident measurement, which corresponds to the pressure levels
and should be very accurate within the altitude range considered
in this study (upper troposphere to lower stratosphere). Thus,
we are integrating over the same altitude range for both instruments.
Since ACE-FTS and both MIPAS data products include retrieved
temperatures, we use their retrieved temperature. For TANSO-FTS and
NDACC, we use their corresponding a priori temperatures.
Several methods of integration were investigated, and the results
presented in Sect. are derived by simple summation
of the integrand multiplied by the bin width of each data point in
kilometers. We also used numerical integration techniques, variations of
Newton–Cotes and Gaussian quadrature formulas. These did not
provide significantly different results due the large size of our
sample (i.e., our results are statistics found from the
least-squares method, and small differences in the individual
partial columns due to different integration methods do not
introduce bias). Since the analytic function being integrated is
not well defined, neither is the uncertainty of the derived
partial column. Propagating reported retrieval uncertainties of
temperature and VMR provides the most appropriate estimate of
uncertainty, which is shown in Fig. .
Partial column (PC) correlation plots comparing TANSO-FTS
CH4 to each
validation instrument. Comparisons to ACE-FTS are red, to
IMK-IAA MIPAS are blue, to ESA MIPAS are purple, and to NDACC are
orange. The vertical range of partial column integration varies
for each pair of coincident profiles based on the criteria described
in Sect. .
The statistics for weighted linear least-squares regression
are shown, with weights equal to 1/(δx2+δy2).
Partial column correlation
The computed partial columns from TANSO-FTS are plotted against
those from each validation instrument in Fig. . The
panels for ACE-FTS, ESA MIPAS, and IMK-IAA MIPAS contain measurements
for all latitudes, and that for NDACC combines results from all
16 stations. Since IMK-IAA retrievals do not extend as low as those of
ESA generally, the altitude range of the partial column integral
is often smaller than those of the other instruments, resulting in
smaller CH4 abundances. Conversely, abundances when comparing
to the NDACC stations are the largest.
The Pearson correlation coefficients, R2, are 0.9986, 0.9965, 0.9968,
and 0.9958 for ACE-FTS, IMK-IAA MIPAS, ESA MIPAS, and NDACC, respectively.
The slopes of the fitted correlation lines are all close to unity,
and a small bias is seen in the y intercept corresponding to between
0.4 and 2.8 % relative to the mean partial columns of the
validation targets, with the greatest corresponding to the NDACC
data. Among the individual NDACC stations, those with the largest
correlation function intercept are Mauna Loa, Jungfraujoch, Bremen,
Izaña, and Zugspitze (1.2×1023–7.5×1023).
TANSO-FTS has a negative intercept only with respect to two stations:
the correlation coefficients for each station are all greater than 0.96, except for
Mauna Loa, Izaña, and Réunion Maïdo, which all
happen to be islands and for which a large number of coincident
TANSO-FTS measurements would have been made over water (see
Sect. ).
Statistics for the partial column integration ranges for
ESA MIPAS, IMK-IAA MIPAS, ACE-FTS, and NDACC stations with the
requirements that the TANSO-FTS sensitivity, s, is greater than
0.2 for at least three pressure levels. The number of coincident
profiles passing this criterion, N, and its percentage of one-to-one
coincidences found in this study are given. Means and standard
deviations are given for the minimum altitudes, min(z); total
integration range, zrange; and number of levels used, n.
TargetProfiles with s>0.2Lowest altitude (km) Altitude range (km) Number of levels InstrumentN(%)min(z)σmin(z)zrangeσzrangenσnESA MIPAS52 01660.98.41.54.61.54.81.1IMK-IAA MIPAS17 78734.811.30.63.50.93.70.6ACE-FTS256259.67.31.45.22.35.41.8Total NDACC18 58798.03.31.011.32.110.41.5
Statistics regarding the distribution of the integration ranges over
altitude are given in Table . This table gives the
number of coincident pairs for each validation instrument
for which the TANSO-FTS CH4 VMR vertical profile passed
the sensitivity requirements. It also gives the
mean and standard deviation of the lower bound of the integral
(lower altitude), the width of the interval (highest altitude minus
the lowest altitude), and the number of pressure levels used. As expected,
the NDACC stations have the widest altitude range,
while the IMK-IAA MIPAS retrievals have the smallest. Note that the
column in Table showing number of levels used does not
correspond to the mode in Fig. since Fig.
considers only the TANSO-FTS averaging kernels and does not reflect
the lack of available comparison data at lower altitudes.
As in Fig. but for partial column correlation
results using unsmoothed CH4 VMR vertical profiles for each
validation instrument.
Repeating the analysis using unsmoothed data from ACE-FTS, ESA and
IMK-IAA MIPAS, and NDACC, the spread in the correlation plots
increases and the biases observed in the intercepts increase, while
the correlation coefficients remain very close to unity.
Figure shows derived partial column correlation
plots for each validation target instrument. The intercept without
smoothing is between 2 and 6 %. The correlation coefficient for
the MIPAS instruments is reduced to 0.97.
Discussion
The objective of this study was to quantitatively assess TANSO-FTS
CH4 VMR vertical profile retrievals compared with other FTS
instruments and to further
investigate whether there were any biases with latitude or other
retrieval parameters. As shown in Sect. , we did not
find a significant difference in mean CH4 VMR profile
differences between latitudinal zones.
To investigate further, we consider the CH4 VMR differences
averaged over altitude for each coincident pair, for each validation
instrument. To choose the altitude range over which to find the
mean, we use the same sensitivity criteria developed in
Sect. . The resulting mean differences between
TANSO-FTS and ACE-FTS, MIPAS, and NDACC are shown as a function of
latitude in Fig. . Weighted least-squares regression
of the combined data sets for each hemisphere reveals a bias at all
latitudes of 13.30±0.06ppbv. There is also a small slope
in the data from each hemisphere, decreasing from the poles to the
tropics. Linear fit parameters for the combined data sets in
each hemisphere are given in Table . This leads to
a bias of around
4 ppbv in the tropics (0.25 % of a tropical
tropospheric VMR value of 1.8–2 ppmv) and of
0.014 and 0.020 ppmv at the North and South Pole,
respectively (or around 1 %). The biases are latitude-dependent
and vary between the tropics and the poles.
Mean CH4 VMR differences between TANSO-FTS and each
validation target data set, averaged vertically using the altitude range
selected for integrating partial columns as a function of latitude.
Differences are calculated as
TANSO-FTS minus target for each data set compared.
Least-squares regression statistics for the data in each
hemisphere plotted in Fig. . Results from
all four validation target data sets are combined.
We also compared the differences shown in Fig. to
TANSO-FTS retrieval parameters: land or sea mask, sunglint flag,
incident angle along the scan path, incident angle along the GOSAT
track path, and observation mode see. Each parameter was compared
to the latitudes and the mean differences in Fig. ,
and the regression and covariance statistics from least-squares
fitting were computed. We found no biases in our
coincident TANSO-FTS data set related to any of these parameters or
whether the observation was made during night or day.
The land or sea mask is an indicator of whether the retrieval was
made over land, water, or a combination in the field of view.
In our data set of all one-to-one coincidences between TANSO-FTS
and the validation targets, 54.0 % of TANSO-FTS measurements
were made over water, 36.3 % were made over land, and
9.6 % were a mixture. The sunglint flag indicates whether the
positions of the sun, satellite, and observation point are related
within a predefined range, qualifying the observation as being
made in sunglint mode. In our data set, only 1.6 % of
TANSO-FTS measurements are sunglint observations, and they
are all over water and within ±45∘ latitude.
Finally, 54.1 % of TIR observations were made at night.
The primary driver of the mean differences found when comparing
TANSO-FTS to other FTS instruments, with and without smoothing,
is the instrument design and observation geometry. TANSO-FTS
is a much more compact and, therefore, coarser-spectral-resolution
FTS than those used in the comparison. The coarser
spectral resolution makes it harder to distinguish closely
spaced absorption lines, leading to poorer vertical sensitivity
and higher uncertainty in the measurements. While the TIR
spectral range of TANSO-FTS is comparable to that of MIPAS, the
mid-infrared ranges of NDACC and ACE-FTS include a very strong
methane absorption band near
3000 cm-1 with little interference from CO2,
increasing their sensitivity and ability to accurately constrain
CH4 retrievals. Furthermore, MIPAS and ACE-FTS observe
the limb of the atmosphere, providing them with more measurements
per retrieved profile, improved vertical resolution, and much
higher sensitivity. While NDACC instruments also only have a
single spectrum per retrieved profile, they observe the sun
directly (as does ACE-FTS), resulting in a very strong signal.
All these factors contribute to TANSO-FTS performing retrievals
on a lower-spectral-resolution measurement of a weaker signal
compared to MIPAS, ACE-FTS, and the NDACC sites. This results in
the sensitivity and DOFS shown in Figs. and .
In Sect. , we examined the variability within each
data set. This gives an idea of some of the sources of error
in our comparison. The coincidence criteria used allow for
the comparison of retrieved CH4 vertical profiles
from different air masses. Our investigation of the NDACC data
provides an estimate of the dependence of the CH4
abundance on time, since we compared profiles retrieved from
the same location using the same retrieval algorithms but at
different times of day. Our result shows that temporal spacing
may contribute around 5 ppbv. Our investigation of the
ACE-FTS variability fixed the instrument and retrieval algorithm
but compared observations of different air masses, and we found
a similar result of only several ppbv. The largest
variability was exhibited when we investigated the MIPAS data
set. This comparison was of the same observations analyzed by
different retrieval algorithms (IMK-IAA and ESA) and resulted
in much larger mean differences on the order of 100 ppbv.
Differences in retrieval algorithms between TANSO-FTS and the
validation instruments may also account for the differences
found in Figs. and . Small
differences in spectroscopic parameters exist; for example,
each instrument's retrieval algorithms use different
editions of the HITRAN line list. Comparisons of these
line lists, and their impact on retrievals, can be found in,
e.g., , , and . The most significant
parameter for TANSO-FTS is its a priori due to the weight given
to the a priori profile by the TANSO-FTS averaging kernels in
the retrieval. In Sect. we compared the
TANSO-FTS-retrieved vertical profiles of CH4 to the corresponding
a priori profiles and found that they differ, on average,
by up to 30 ppbv. This provides a rough minimum of the accuracy
of the a priori profiles required for the retrievals.
Conclusions
The TANSO-FTS TIR CH4 vertical profile data product is an
important and novel data set. Its vertical range extends lower into
the troposphere than other satellite data products, and its spatial
coverage is global with a high density of measurements. We have
investigated the sensitivity and averaging kernels for the TANSO-FTS
data product and done a global comparison with four other
FTS data products. Our comparisons showed that the sensitivity of
the TANSO-FTS retrieval is relatively low at all altitudes and
that there is a limitation on the upper altitude of its data
product of around 15 or 20 km. Unfortunately, the
lower-altitude boundaries of the other satellite-based data products,
between 7 and 15 km, reduce the vertical range over which we
can make comparisons. In the upper troposphere,
we found good agreement between TANSO-FTS and
NDACC, without a bias. The agreement between these two data sets
persisted regardless of whether smoothing was applied to the NDACC
data. Therefore, despite the lower sensitivity of the
TANSO-FTS data product, it remains an important and unique data set
of global tropospheric CH4 measurements.
In the overlapping altitude ranges of the three satellite data
products, we found a small, but consistent, positive bias of around
20 ppbv, or 1 %. We found that the shapes of the
TANSO-FTS CH4 VMR vertical profiles near 15 km,
where the CH4 VMR falls off with increasing
altitude, does not match those of the other instruments, and in a
consistent manner, resulting in a pronounced feature in the
mean difference profiles in Fig. , just below
the 100 hPa level. Despite the large variability in each
data set and in the differences between the TANSO-FTS retrievals
and the others, we found that partial columns computed from
the vertical profiles were very tightly correlated, with and without
smoothing.
When looking for a relationship between latitude and the differences
between data products, we found a small, but statistically
significant, dependence of the vertically averaged differences on
latitude. The TANSO-FTS data product shows better agreement over
the tropics than the poles.
We look forward to future versions of the retrieval which may
feature a greater sensitivity and altitude range, while reducing
the small biases and dependence on the a priori profiles. In a future
release, the a priori will not be changed but remain the outputs
of the NIES-TM. used theoretical simulations to
determine that the Level 1B spectra which were used (V161) to
generate the current TIR CH4 data product had considerable
uncertainties. New Level 1B spectra are due for release in 2018
and should lead to improved retrievals. also
proposed some corrections to the TANSO-FTS TIR L1B spectra which
may be implemented. The spectral line list used (HITRAN 2008)
will be updated. Uncertainties in the surface emissivity over
cold surfaces (snow and ice) affect the retrieval at higher
altitudes and will be improved in the next release. Improvements
are also being made to the way the retrieval handles and
simultaneously retrieves interfering species, such as O3.
All data used in this study except for that from GOSAT
TANSO-FTS are available upon request and accessed through appropriate Web
portals. The ACE-FTS Science Team at the University of Waterloo provided
access to their Level 2 data through https://databace.scisat.ca/level2/
(registration required). Access to the ESA MIPAS Level 2 data was granted and
provided through the ESA Earth Online portal: https://earth.esa.int/
(registration required). The IMK-IAA MIPAS Level 2 data were accessed using
the KIT website: https://www.imk-asf.kit.edu/english/308.php
(registration required). NDACC data have been compiled by many independent
research groups and were accessed through a National Centers for
Environmental Prediction FTP server, with each station being accessible
through https://www.ndsc.ncep.noaa.gov/data/data_tbl/.
Early access to the TANSO-FTS TIR CH4 VMR vertical profiles was provided through the GOSAT User Interface Gateway: https://data.gosat.nies.go.jp/ (no
longer functional). Regular access to the TANSO-FTS data products is
available at https://data2.gosat.nies.go.jp/index_en.html (registration
required).
The work presented here was done
by KSO with input from and under the supervision of
KS, KAW, and NS. KSO prepared the
manuscript with contributions from all co-authors. CH4
retrievals were developed and provided, with additional input, by
NS for TANSO-FTS, CDB for ACE-FTS, PR
for ESA MIPAS, JP
for IMK-IAA MIPAS, MG for Altzomoni, JWH
for Thule and Mauna Loa, FH for Kiruna, NJ for
Wollongong, WB for Jungfraujoch, MdM for
Réunion St. Denis and Maïdo, JN for Bremen
and Ny Ålesund, MS for Izaña, DS for
Lauder and Arrival Heights, SC and KS for Eureka
and Toronto, and RS for Zugspitze.
The authors declare that they have no conflict of
interest.
Acknowledgements
This research was conducted under the framework of the
Japan Aerospace Exploration Agency (JAXA), National Institute
for Environmental Studies (NIES), and the Ministry of the
Environment (MOE) Research Announcement (RA) project “GOSAT
Validation Using Eureka and Toronto Ground-Based Measurements
and ACE, CloudSat, and CALIPSO Satellite Data”, which was
supported by the Canadian Space Agency (CSA), the Natural
Sciences and Engineering Research Council of Canada (NSERC),
and Environment & Climate Change Canada (ECCC).
This work is the result of many long-lasting collaborations, and we
would like to thank our co-authors and collaborators for providing
data and expertise. The GOSAT team provided early access to its TANSO-FTS TIR CH4
VMR vertical profiles through the GOSAT User Interface Gateway. SCISAT/ACE is a Canadian-led mission mainly supported by the CSA and NSERC. The
ACE-FTS Science Team at the University of Waterloo provided access to their
Level 2 data, as well as expert knowledge with its interpretation and quality
management. Access to the ESA MIPAS Level 2 data was granted and provided
through the ESA Earth Online portal. The IMK-IAA MIPAS Level 2 data were
accessed using the KIT website. NDACC data have been compiled by many
independent research groups and were accessed through a National Centers for
Environmental Prediction FTP server.
Measurements at PEARL were made by the Canadian Network for the
Detection of Atmospheric Change (CANDAC), led by James R. Drummond,
and in part by the Canadian Arctic ACE/OSIRIS Validation Campaigns,
led by Kaley A. Walker. Support is provided by AIF/NSRIT, CFI, CFCAS,
CSA, EC, GOC-IPY, NSERC, NSTP, OIT, PCSP, and ORF. Logistical and
operational support is provided by PEARL Site Manager Pierre Fogal,
CANDAC operators, and the ECCC weather station. Measurements at the
University of Toronto Atmospheric Observatory were supported by
CFCAS, ABB Bomem, CFI, CSA, EC, NSERC, ORDCF, PREA, and the
University of Toronto. NDACC data analysis at Toronto and Eureka
was supported by the CAFTON project, funded by the CSA's FAST Program.
The National Institute of Water and Atmospheric Research Ltd (NIWA)
ground-based FTSs are core-funded through New Zealand's
Ministry of Business, Innovation and Employment. We thank Antarctica,
New Zealand, and the Scott Base staff for providing logistical support
at Arrival Heights. Measurements at the Jungfraujoch station are
primarily supported by the Fonds de la Recherche Scientifique
(F.R.S.–FNRS) and the Fédération Wallonie-Bruxelles, both in
Brussels. The Swiss GAW-CH program of MeteoSwiss is further
acknowledged. We thank the International Foundation High Altitude
Research Stations Jungfraujoch and Gornergrat (HFSJG, Bern) for
supporting the facilities needed to perform the observations and the
many colleagues who contributed to FTS data acquisition at that site.
Whitney Bader has received funding from the European Union's Horizon
2020 research and innovation program under the Marie
Skłodowska-Curie actions grant agreement no. 704951. We would like to thank
Alejandro Bezanilla, who operates the Altzomoni site, and Wolfgang Stremme,
who does the data processing for the Altzomoni site and uploads
the data to the NDACC archive. Altzomoni is supported by
Consejo Nacional de Ciencia y Tecnología (CONACYT, grants 239618 and
249374) and la Dirección General de Asuntos del Personal Académico
de la Universidad Nacional Autónoma de México (DGAPA-UNAM, grants
IN109914 and IA101814).
Edited by: Helen Worden
Reviewed by: two anonymous referees
ReferencesBader, W., Bovy, B., Conway, S., Strong, K., Smale, D., Turner,
A. J., Blumenstock, T., Boone, C., Collaud Coen, M., Coulon, A.,
Garcia, O., Griffith, D. W. T., Hase, F., Hausmann, P., Jones, N.,
Krummel, P., Murata, I., Morino, I., Nakajima, H., O'Doherty, S.,
Paton-Walsh, C., Robinson, J., Sandrin, R., Schneider, M., Servais,
C., Sussmann, R., and Mahieu, E.: The recent increase of atmospheric
methane from 10 years of ground-based NDACC FTIR observations since 2005,
Atmos. Chem. Phys., 17, 2255–2277, 10.5194/acp-17-2255-2017, 2017.Baray, J. L., Courcoux, Y., Keckhut, P., Portafaix, T., Tulet, P., Cammas,
J. P., Hauchecorne, A., Godin, S., Beekmann, S., de Mazière, Hermans,
C., Desmet, F., Sellegri, K., Colomb, A., Ramonet, M., Sciare, J.,
Vuillemin, C., Hoareau, C., Dionisi, D., Duflot, V.,
Vérèmes, H., Porteneuve, J., Gabarrot, F., Gaudo, T.,
Metzger, J.-M., Payen, G., Leclair de Bellevue, J., Barthe, C.,
Posny, F., Ricaud, P., Abchiche, A., and Delmas, R.: Maïdo
observatory: a new high-altitude station facility at Reunion Island
(21∘ S, 55∘ E) for long-term atmospheric remote sensing and in
situ measurements, Atmos. Meas. Tech., 6, 2865–2877,
10.5194/amt-6-2865-2013, 2013.Batchelor, R. L., Strong, K., Lindenmaier, R., Mittermeier, R. L.,
Fast, H., Drummond, J. R., and Fogal, P. F.: A new Bruker IFS 125HR
FTIR spectrometer for the Polar Environment Atmospheric Research Laboratory
at Eureka, Canada: measurements and comparison with the existing Bomem DA8
spectrometer, J. Atmos. Ocean. Tech., 26, 1328–1340,
10.1175/2009JTECHA1215.1, 2009.Baylon, J. L., Stremme, W., Plaza, E., Bezanilla, A., Grutter, M.,
Hase, F., and Blumenstock, T.: CO2 total column variability from
ground-based FTIR measurements over central Mexico, in: AGU Fall Meeting,
AGU Fall Meeting, 2014.Bernath, P. F.: The Atmospheric Chemistry Experiment (ACE), J. Quant. Spectrosc. Ra., 186, 3–16, 10.1016/j.jqsrt.2016.04.006,
2017.Bernath, P. F., McElroy, C. T., Abrams, M. C., Boone, C. D., Butler,
M., Camy-Peyret, C., Carleer, M., Clerbaux, C., Coheur, P.-F.,
Colin, R., DeCola, P., de Mazière, M., Drummond, J. R., Dufour,
D., Evans, W. F. J., Fast, H., Fussen, D., Gilbert, K., Jennings,
D. E., Llewellyn, E. J., Lowe, R. P., Mahieu, E., McConnell, J. C.,
McHugh, M., McLeod, S. D., Michaud, R., Midwinter, C., Nassar, R.,
Nichitiu, F., Nowlan, C., Rinsland, C. P., Rochon, Y. J., Rowlands,
N., Semeniuk, K., Simon, P., Skelton, R., Sloan, J. J., Soucy,
M.-A., Strong, K., Tremblay, P., Turnbull, D., Walker, K. A.,
Walkty, I., Wardle, D. A., Wehrle, V., Zander, R., and Zou, J.:
Atmospheric Chemistry Experiment (ACE): Mission overview, Geophys. Res.
Lett., 32, L15S01, 10.1029/2005GL022386, 2005.Blumenstock, T., Kopp, G., Hase, F., Hochschild, G., Mikuteit, S., Raffalski,
U., and Ruhnke, R.: Observation of unusual chlorine activation by
ground-based infrared and microwave spectroscopy in the late Arctic winter
2000/01, Atmos. Chem. Phys., 6, 897–905, 10.5194/acp-6-897-2006,
2006.Boone, C. D., Nassar, R., Walker, K. A., Rochon, Y., McLeod, S. D.,
Rinsland, C. P., and Bernath, P. F.: Retrievals for the atmospheric
chemistry experiment Fourier-transform spectrometer, Appl. Opt., 44,
7218–7231, 10.1364/AO.44.007218, 2005.
Boone, C. D., Walker, K. A., and Bernath, P. F.: Version 3 Retrievals
for the Atmospheric Chemistry Experiment Fourier Transform Spectrometer
(ACE-FTS), in: ACE at 10: Solar Occultation Anthology, edited by: Bernath,
P. F., A. Deepak Publishing, Hampton, Virg., 103–127, 2013.Buchwitz, M., Schneising, O., Burrows, J. P., Bovensmann, H., Reuter,
M., and Notholt, J.: First direct observation of the atmospheric CO2
year-to-year increase from space, Atmos. Chem. Phys., 7, 4249–4256,
10.5194/acp-7-4249-2007, 2007.Côté, J. S., Gravel, S., Méthot, A., Patoine, A., Roch,
M., and Staniforth, A.: The operational CMC-MRB global environmental
multiscale (GEM) model. Part I: Design considerations and formulation, Mon.
Weather Rev., 126, 1373–1395,
10.1175/1520-0493(1998)126<1373:TOCMGE>2.0.CO;2, 1998.de Mazière, M., Vigouroux, C., Bernath, P. F., Baron, P.,
Blumenstock, T., Boone, C., Brogniez, C., Catoire, V., Coffey, M.,
Duchatelet, P., Griffith, D., Hannigan, J., Kasai, Y., Kramer, I.,
Jones, N., Mahieu, E., Manney, G. L., Piccolo, C., Randall, C.,
Robert, C., Senten, C., Strong, K., Taylor, J., Tétard, C.,
Walker, K. A., and Wood, S.: Validation of ACE-FTS v2.2 methane profiles
from the upper troposphere to the lower mesosphere, Atmos. Chem. Phys., 8,
2421–2435, 10.5194/acp-8-2421-2008, 2008.Engel, A., Bönisch, H., Schwarzenberger, T., Haase, H.-P., Grunow, K.,
Abalichin, J., and Sala, S.: Long-term validation of ESA operational
retrieval (version 6.0) of MIPAS Envisat vertical profiles of methane,
nitrous oxide, CFC11, and CFC12 using balloon-borne observations and
trajectory matching, Atmos. Meas. Tech., 9, 1051–1062,
10.5194/amt-9-1051-2016, 2016.Errera, Q., Ceccherini, S., Christophe, Y., Chabrillat, S., Hegglin,
M. I., Lambert, A., Ménard, R., Raspollini, P., Skachko, S., van
Weele, M., and Walker, K. A.: Harmonisation and diagnostics of MIPAS ESA
CH4 and N2O profiles using data assimilation, Atmos. Meas. Tech.,
9, 5895–5909, 10.5194/amt-9-5895-2016, 2016.Fischer, H., Birk, M., Blom, C., Carli, B., Carlotti, M., von
Clarmann, T., Delbouille, L., Dudhia, A., Ehhalt, D., Endemann, M.,
Flaud, J. M., Gessner, R., Kleinert, A., Koopman, R., Langen, J.,
López-Puertas, M., Mosner, P., Nett, H., Oelhaf, H., Perron,
G., Remedios, J., Ridolfi, M., Stiller, G., and Zander, R.: MIPAS:
an instrument for atmospheric and climate research, Atmos. Chem. Phys., 8,
2151–2188, 10.5194/acp-8-2151-2008, 2008.Goldman, A., Paton-Walsh, C., Bell, W., Toon, G. C. andBlavier, J. F.,
Sen, B., Coffey, M. T., Hannigan, J. W., and Mankin, W. G.: Network for
the detection of stratospheric change Fourier transform infrared
intercomparison at Table Mountain Facility, November 1996, J. Geophys. Res.,
104, 30481–30503, 10.1029/1999JD900879, 1999.Hannigan, J. W., Coffey, M. T., and Goldman, A.: Semi-autonomous FTS
observation system for stratospheric and tropospheric gases, J. Atmos.
Ocean. Tech., 26, 1814–1828, 10.1175/2009JTECHA1230.1, 2009.Hase, F., Hannigan, J. W., Coffey, M. T., Goldman, A., Hopfner, M.,
Jones, N. B., Rinsland, C. P., and Wood, S. W.: Intercomparison of
retrieval codes used for the analysis of high-resolution, ground-based FTIR
measurements, J. Quant. Spectrosc. Ra., 87, 25–52,
10.1016/j.jqsrt.2003.12.008, 2004.Holl, G., Walker, K. A., Conway, S., Saitoh, N., Boone, C. D.,
Strong, K., and Drummond, J. R.: Methane cross-validation between three
Fourier transform spectrometers: SCISAT ACE-FTS, GOSAT TANSO-FTS, and
ground-based FTS measurements in the Canadian high Arctic, Atmos. Meas.
Tech., 9, 1961–1980, 10.5194/amt-9-1961-2016, 2016.Ishida, H. and Nakajima, T. Y.: Development of an unbiased cloud detection
algorithm for a spaceborne multispectral imager, J. Geophys. Res., 114,
D07206, 10.1029/2008JD010710, 2009.Ishida, H., Nakjima, T. Y., Yokota, T., Kikuchi, N., and Watanabe, H.:
Investigation of GOSAT TANSO-CAI cloud screening ability through an
intersatellite comparison, J. Appl. Meteorol. Climatol., 50, 1571–1586,
10.1175/2011JAMC2672.1, 2011.Jin, J. J., Semeniuk, K., Beagley, S. R., Fomichev, V. I., Jonsson,
A. I., McConnell, J. C., Urban, J., Murtagh, D., Manney, G. L.,
Boone, C. D., Bernath, P. F., Walker, K. A., Barret, B., Ricaud,
P., and Dupuy, E.: Comparison of CMAM simulations of carbon monoxide (CO),
nitrous oxide (N2O), and methane (CH4) with observations from
Odin/SMR, ACE-FTS, and Aura/MLS, Atmos. Chem. Phys., 9, 3233–3252, 2009.Kohlhepp, R., Ruhnke, R., Chipperfield, M. P., De Mazière, M., Notholt,
J., Barthlott, S., Batchelor, R. L., Blatherwick, R. D., Blumenstock, T.,
Coffey, M. T., Demoulin, P., Fast, H., Feng, W., Goldman, A., Griffith, D.
W. T., Hamann, K., Hannigan, J. W., Hase, F., Jones, N. B., Kagawa, A.,
Kaiser, I., Kasai, Y., Kirner, O., Kouker, W., Lindenmaier, R., Mahieu, E.,
Mittermeier, R. L., Monge-Sanz, B., Morino, I., Murata, I., Nakajima, H.,
Palm, M., Paton-Walsh, C., Raffalski, U., Reddmann, T., Rettinger, M.,
Rinsland, C. P., Rozanov, E., Schneider, M., Senten, C., Servais, C.,
Sinnhuber, B.-M., Smale, D., Strong, K., Sussmann, R., Taylor, J. R.,
Vanhaelewyn, G., Warneke, T., Whaley, C., Wiehle, M., and Wood, S. W.:
Observed and simulated time evolution of HCl, ClONO2, and HF total column
abundances, Atmos. Chem. Phys., 12, 3527–3556,
10.5194/acp-12-3527-2012, 2012.
Kurylo, M. J. and Zander, R.: The NDSC–Its status after 10 years of
operation, in: Proc. 19th Quadrennial Ozone Symp., 167–168, 2000.Kuze, A., Suto, H., Nakajima, M., and Hamazaki, T.: Thermal and near
infrared sensor for carbon observation Fourier-transform spectrometer on the
Greenhouse Gases Observing Satellite for greenhouse gases monitoring, Appl.
Opt., 48, 6716, 10.1364/AO.48.006716, 2009.Kuze, A., Suto, H., Shiomi, K., Urabe, T., Nakajima, M., Yoshida,
J., Kawashima, T., Yamamoto, Y., Kataoka, F., and Buijs, H.: Level 1
algorithms for TANSO on GOSAT: processing and on-orbit calibrations, Atmos.
Meas. Tech., 5, 2447–2467, 10.5194/amt-5-2447-2012, 2012.Kuze, A., Suto, H., Shiomi, K., Kawakami, S., Tanaka, M., Ueda, Y.,
Deguchi, A., Yoshida, J., Yamamoto, Y., Kataoka, F., Taylor, T. E.,
and Buijs, H. L.: Update on GOSAT TANSO-FTS performance, operations, and
data products after more than 6 years in space, Atmos. Meas. Tech., 9,
2445–2461, 10.5194/amt-9-2445-2016, 2016.Laeng, A., Plieninger, J., von Clarmann, T., Grabowski, U., Stiller,
G., Eckert, E., Glatthor, N., Haenel, F., Kellmann, S., Kiefer, M.,
Linden, A., Lossow, S., Deaver, L., Engel, A., Hervig, M., Levin,
I., McHugh, M., Noël, S., Toon, G., and Walker, K.: Validation
of MIPAS IMK/IAA methane profiles, Atmos. Meas. Tech., 8, 5251–5261,
10.5194/amt-8-5251-2015, 2015.
Maksyutov, S., Patra, P. K., Onishi, R. A. S. T., and T., N.: NIES/FRCGC
global atmospheric tracer transport model: Description, validation, and
surface sources and sinks inversion, J. Earth Simul., 9, 3–18, 2008.Notholt, J., Toon, G. C., Stordal, F., S., S., Schmidbauer, N.,
Becker, E., Meier, A., and Sen, B.: Seasonal variations of atmospheric
trace gases in the high Arctic at 79∘ N, J. Geophys. Res., 102,
12855–12861, 10.1029/97JD00337, 1997.Payan, S., Camy-Peyret, C., Oelhaf, H., Wetzel, G., Maucher, G.,
Keim, C., Pirre, M., Huret, N., Engel, A., Volk, M. C.,
Kuellmann, H., Kuttippurath, J., Cortesi, U., Bianchini, G.,
Mencaraglia, F., Raspollini, P., Redaelli, G., Vigouroux, C., de
Mazière, M., Mikuteit, S., Blumenstock, T., Velazco, V.,
Notholt, J., Mahieu, E., Duchatelet, P., Smale, D., Wood, S.,
Jones, N., Piccolo, C., Payne, V., Bracher, A., Glatthor, N.,
Stiller, G., Grunow, K., Jeseck, P., Te, Y., and Butz, A.:
Validation of version-4.61 methane and nitrous oxide observed by MIPAS,
Atmos. Chem. Phys., 9, 413–442, 10.5194/acp-9-413-2009, 2009.Picone, J. M., Hedin, A. E., Drob, D. P., and Aikin, A. C.: NRLMSISE-00
empirical model of the atmosphere: Statistical comparisons and scientific
issues, J. Geophys. Res., 107, 1468, 10.1029/2002JA009430, 2002.Plieninger, J., von Clarmann, T., Stiller, G. P., Grabowski, U.,
Glatthor, N., Kellmann, S., Linden, A., Haenel, F., Kiefer, M.,
Höpfner, M., Laeng, A., and Lossow, S.: Methane and nitrous oxide
retrievals from MIPAS-ENVISAT, Atmos. Meas. Tech., 8, 4657–4670,
10.5194/amt-8-4657-2015, 2015.Plieninger, J., Laeng, A., Lossow, S., von Clarmann, T., Stiller,
G. P., Kellmann, S., Linden, A., Kiefer, M., Walker, K. A.,
Noël, S., Hervig, M. E., McHugh, M., Lambert, A., Urban, J.,
Elkins, J. W., and Murtagh, D.: Validation of revised methane and
nitrous oxide profiles from MIPAS-ENVISAT, Atmos. Meas. Tech., 9, 765–779,
10.5194/amt-9-765-2016, 2016.Pougatchev, N. S., Connor, B. J., and Rinsland, C. P.: Infrared
measurements of the ozone vertical distribution above Kitt Peak, J. Geophys.
Res., 100, 16689–16697, 10.1029/95JD01296, 1995.Raspollini, P., Belotti, C., Burgess, A., Carli, B., Carlotti, M.,
Ceccherini, S., Dinelli, B. M., Dudhia, A., Flaud, J.-M., Funke,
B., Höpfner, M., López-Puertas, M., Payne, V., Piccolo, C.,
Remedios, J. J., Ridolfi, M., and Spang, R.: MIPAS level 2 operational
analysis, Atmos. Chem. Phys., 6, 5605–5630, 10.5194/acp-6-5605-2006,
2006.Raspollini, P., Carli, B., Carlotti, M., Ceccherini, S., Dehn, A.,
Dinelli, B. M., Dudhia, A., Flaud, J.-M., López-Puertas, M.,
Niro, F., Remedios, J. J., Ridolfi, M., Sembhi, H., Sgheri, L., and
von Clarmann, T.: Ten years of MIPAS measurements with ESA Level 2
processor V6 – Part 1: Retrieval algorithm and diagnostics of the products,
Atmos. Meas. Tech., 6, 2419–2439, 10.5194/amt-6-2419-2013, 2013.Raspollini, P., Arnone, E., Barbara, F., Carli, B., Castelli, E., Ceccherini,
S., Dinelli, B. M., Dudhia, A., Kiefer, M., Papandrea, E., and Ridolfi, M.:
Comparison of the MIPAS products obtained by four different level 2
processors, Ann. Geophys., 56, 10.4401/ag-6338, 2014.Remedios, J. J., Leigh, R. J., Waterfall, A. M., Moore, D. P., Sembhi, H.,
Parkes, I., Greenhough, J., Chipperfield, M. P., and Hauglustaine, D.: MIPAS
reference atmospheres and comparisons to V4.61/V4.62 MIPAS level 2
geophysical data sets, Atmos. Chem. Phys., 7, 9973–10017,
10.5194/acpd-7-9973-2007, 2007.Rodgers, C. D. and Connor, B. J.: Intercomparison of remote sounding
instruments, J. Geophys. Res., 108, 2156–2202, 10.1029/2002JD002299,
2003.Rothman, L. S., Rinsland, C. P., Goldman, A., Massie, S. T., Edwards,
D. P., Flaud, J.-M., Perrin, A., Camy-Peyret, C., Dana, V., Mandin,
J.-Y., Schroeder, J., McCann, A., Gamache, R. R., Wattson, R. B.,
Yoshino, K., Chance, K., Jucks, K., Brown, L. R., Nemtchinov, V.,
and Varanasi, P.: The HITRAN Molecular Spectroscopic Database and HAWKS
(HITRAN Atmospheric Workstation): 1996 Edition, J. Quant. Spectrosc. Ra., 60, 665–710, 10.1016/S0022-4073(98)00078-8, 1998.Rothman, L. S., Jacquemart, D., Barbe, A., Chris Benner, D., Birk,
M., Brown, L. R., Carleer, M. R., Chackerian, C., Chance, K.,
Coudert, L. H., Dana, V., Devi, V. M., Flaud, J.-M., Gamache,
R. R., Goldman, A., Hartmann, J.-M., Jucks, K. W., Maki, A. G.,
Mandin, J.-Y., Massie, S. T., Orphal, J., Perrin, A., Rinsland,
C. P., Smith, M. A. H., Tennyson, J., Tolchenov, R. N., Toth, R. A.,
Vander Auwera, J., Varanasi, P., and Wagner, G.: The HITRAN 2004
molecular spectroscopic database, J. Quant. Spectrosc. Ra., 96,
139–204, 10.1016/j.jqsrt.2004.10.008, 2005.Rothman, L. S., Gordon, I. E., Barbe, A., Benner, D. C., Bernath,
P. F., Birk, M., Boudon, V., Brown, L. R., Campargue, A., Champion,
J.-P., Chance, K., Coudert, L. H., Dana, V., Devi, V. M., Fally,
S., Flaud, J.-M., Gamache, R. R., Goldman, A., Jacquemart, D.,
Kleiner, I., Lacome, N., Lafferty, W. J., Mandin, J.-Y., Massie,
S. T., Mikhailenko, S. N., Miller, C. E., Moazzen-Ahmadi, N.,
Naumenko, O. V., Nikitin, A. V., Orphal, J., Perevalov, V. I.,
Perrin, A., Predoi-Cross, A., Rinsland, C. P., Rotger, M., Šimečková, M., Smith, M. A. H., Sung, K., Tashkun, S. A.,
Tennyson, J., Toth, R. A., Vandaele, A. C., and Vander Auwera, J.:
The HITRAN 2008 molecular spectroscopic database, J. Quant. Spectrosc. Ra., 110, 533–572, 10.1016/j.jqsrt.2009.02.013, 2009.Rothman, L. S., Gordon, I. E., Babikov, Y., Barbe, A., Chris Benner,
D., Bernath, P. F., Birk, M., Bizzocchi, L., Boudon, V., Brown,
L. R., Campargue, A., Chance, K., Cohen, E. A., Coudert, L. H.,
Devi, V. M., Drouin, B. J., Fayt, A., Flaud, J.-M., Gamache, R. R.,
Harrison, J. J., Hartmann, J.-M., Hill, C., Hodges, J. T.,
Jacquemart, D., Jolly, A., Lamouroux, J., Le Roy, R. J., Li, G.,
Long, D. A., Lyulin, O. M., Mackie, C. J., Massie, S. T.,
Mikhailenko, S., Müller, H. S. P., Naumenko, O. V., Nikitin,
A. V., Orphal, J., Perevalov, V., Perrin, A., Polovtseva, E. R.,
Richard, C., Smith, M. A. H., Starikova, E., Sung, K., Tashkun, S.,
Tennyson, J., Toon, G. C., Tyuterev, V. G., and Wagner, G.: The
HITRAN2012 molecular spectroscopic database, J. Quant. Spectrosc. Ra., 130, 4–50, 10.1016/j.jqsrt.2013.07.002, 2013.Saeki, T., Saito, R., Belikov, D., and Maksyutov, S.: Global
high-resolution simulations of CO2 and CH4 using a NIES transport
model to produce a priori concentrations for use in satellite data
retrievals, Geosci. Model Dev., 6, 81–100, 10.5194/gmd-6-81-2013,
2013.Saitoh, N., Imasu, R., Ota, Y., and Niwa, Y.: CO2 retrieval
algorithm for the thermal infrared spectra of the Greenhouse Gases Observing
Satellite: Potential of retrieving CO2 vertical profile from
high-resolution FTS sensor, J. Geophys. Res., 114, D17305,
10.1029/2008JD011500, 2009.Saitoh, N., Kimoto, S., Sugimura, R., Imasu, R., Kawakami, S.,
Shiomi, K., Kuze, A., Machida, T., Sawa, Y., and Matsueda, H.:
Algorithm update of the GOSAT/TANSO-FTS thermal infrared CO2 product
(version 1) and validation of the UTLS CO2 data using CONTRAIL
measurements, Atmos. Meas. Tech., 9, 2119–2134,
10.5194/amt-9-2119-2016, 2016.Schneider, M., Blumenstock, T., Chipperfield, M., Hase, F., Kouker, W.,
Reddmann, T., Ruhnke, R., Cuevas, E., and Fischer, H.: Subtropical trace gas
profiles determined by ground-based FTIR spectroscopy at Izaña
(28∘ N, 16∘ W): Five year record, error analysis, and
comparison with 3-D CTMs, Atmos. Chem. Phys., 5, 153–167,
10.5194/acp-5-153-2005, 2005.Senten, C., de Mazière, M., Dils, B., Hermans, C., Kruglanski, M., Neefs,
E., Scolas, F., Vandaele, A. C., Vanhaelewyn, G., Vigouroux, C., Carleer, M.,
Coheur, P. F., Fally, S., Barret, B., Baray, J. L., Delmas, R., Leveau, J.,
Metzger, J. M., Mahieu, E., Boone, C. D., Walker, K. A., Bernath, P. F., and
Strong, K.: Technical Note: New ground-based FTIR measurements at Ile de La
Réunion: observations, error analysis, and comparisons with independent
data, Atmos. Chem. Phys., 8, 3483–3508, 10.5194/acp-8-3483-2008,
2008.Sepúlveda, E., Schneider, M., Hase, F., García, O. E.,
Gomez-Pelaez, A., Dohe, S., Blumenstock, T., and Guerra, J. C.:
Long-term validation of tropospheric column-averaged CH4 mole fractions
obtained by mid-infrared ground-based FTIR spectrometry, Atmos. Meas. Tech.,
5, 1425–1441, 10.5194/amt-5-1425-2012, 2012.Sepúlveda, E., Schneider, M., Hase, F., Barthlott, S., Dubravica,
D., García, O. E., Gomez-Pelaez, A., González, Y., Guerra,
J. C., Gisi, M., Kohlhepp, R., Dohe, S., Blumenstock, T., Strong,
K., Weaver, D., Palm, M., Sadeghi, A., Deutscher, N. M., Warneke,
T., Notholt, J., Jones, N., Griffith, D. W. T., Smale, D.,
Brailsford, G. W., Robinson, J., Meinhardt, F., Steinbacher, M.,
Aalto, T., and Worthy, D.: Tropospheric CH4 signals as observed by
NDACC FTIR at globally distributed sites and comparison to GAW surface in
situ measurements, Atmos. Meas. Tech., 7, 2337–2360,
10.5194/amt-7-2337-2014, 2014.
Sheese, P., Walker, K., and Boone, C.: Atmospheric pseudo-retrievals for
averaging kernel and total uncertainty characterization for ACE-FTS level 2
(PRAKTICAL) data, in: EGU General Assembly, vol. 18 of EGU General
Assembly, p. 17582, 2016.Sheese, P. E., Boone, C. D., and Walker, K. A.: Detecting physically
unrealistic outliers in ACE-FTS atmospheric measurements, Atmos. Meas.
Tech., 8, 741–750, 10.5194/amt-8-741-2015, 2015.Sussmann, R. and Schäfer, K.: Infrared spectroscopy of tropospheric
trace gases: combined analysis of horizontal and vertical column abundances,
Appl. Opt., 36, 735–741, 10.1364/AO.36.000735, 1997.Sussmann, R., Forster, F., Rettinger, M., and Jones, N.: Strategy for
high-accuracy-and-precision retrieval of atmospheric methane from the
mid-infrared FTIR network, Atmos. Meas. Tech., 4, 1943–1964,
10.5194/amt-4-1943-2011, 2011.Sussmann, R., Ostler, A., Forster, F., Rettinger, M., Deutscher,
N. M., Griffith, D. W. T., Hannigan, J. W., Jones, N., and Patra,
P. K.: First intercalibration of column-averaged methane from the Total
Carbon Column Observing Network and the Network for the Detection of
Atmospheric Composition Change, Atmos. Meas. Tech., 6, 397–418,
10.5194/amt-6-397-2013, 2013.Toon, G. C., Blavier, J.-F., Sung, K., Rothman, L. S., and E. Gordon,
I.: HITRAN spectroscopy evaluation using solar occultation FTIR spectra, J. Quant. Spectrosc. Ra., 182, 324–336,
10.1016/j.jqsrt.2016.05.021, 2016.Vincenty, T.: Direct and Inverse Solutions of Geodesics on the Ellipsoid with
application of nested equations, Surv. Rev., 23, 88–93,
10.1179/sre.1975.23.176.88, 1975.von Clarmann, T., Höpfner, M., Kellmann, S., Linden, A., Chauhan,
S., Funke, B., Grabowski, U., Glatthor, N., Kiefer, M.,
Schieferdecker, T., Stiller, G. P., and Versick, S.: Retrieval of
temperature, H2O, O3, HNO3, CH4, N2O, ClONO2
and ClO from MIPAS reduced resolution nominal mode limb emission
measurements, Atmos. Meas. Tech., 2, 159–175, 10.5194/amt-2-159-2009,
2009.
Waymark, C., Walker, K. A., Boone, C. D., and Bernath, P. F.: ACE-FTS
version 3.0 data set: validation and data processing update, Ann. Geophys.,
56, 10.4401/ag-6339, 2013.Wiacek, A., Taylor, J. R., Strong, K., Saari, R., Kerzenmacher, T., Jones,
N. B., and Griffith, D. W. T.: Ground-Based solar absorption FTIR
spectroscopy: characterization of retrievals and first results from a novel
optical design instrument at a New NDACC Complementary Station, J. Atmos.
Ocean. Tech., 24, 432–448, 10.1175/JTECH1962.1, 2007.Wood, S. W., Bodeker, G. E., Boyd, I. S., Jones, N. B., Connor, B. J.,
Johnston, P. V., Matthews, W. A., Nichol, S. E., Murcray, F. J., Nakajima,
H., and Sasano, Y.: Validation of version 5.20 ILAS HNO3, CH4, N2O,
O3, and NO2 using ground-based measurements at Arrival Heights and
Kiruna, J. Geophys. Res., 107, 8208, 10.1029/2001JD000581, 2002.Yokota, T., Yoshida, Y., Eguchi, N., Ota, Y., Tanaka, T., Watanabe, H., and
Maksyutov, S.: Global Concentrations of CO2 and CH4 Retrieved from
GOSAT: First Preliminary Results, Sci. Online Lett. Atmos., 5, 160–163,
10.2151/sola.2009-041, 2009.Zander, R., Mahieu, E., Demoulin, P., Duchatelet, P., Roland, G., Servais, C.,
de Mazière, M., Reimann, S., and Rinsland, C. P.: Our changing
atmosphere: evidence based on long-term infrared solar observations at the
Jungfraujoch since 1950, Sci. Total Environ., 391, 184–195,
10.1016/j.scitotenv.2007.10.018, 2008.Zou, M., Xiong, X., Saitoh, N., Warner, J., Zhang, Y., Chen, L.,
Weng, F., and Fan, M.: Satellite observation of atmospheric methane:
intercomparison between AIRS and GOSAT TANSO-FTS retrievals, Atmos. Meas.
Tech., 9, 3567–3576, 10.5194/amt-9-3567-2016, 2016.