Among the more than 20 ground-based FTIR (Fourier transform infrared) stations currently operating around the globe, only a few have provided formaldehyde (HCHO) total column time series until now. Although several independent studies have shown that the FTIR measurements can provide formaldehyde total columns with good precision, the spatial coverage has not been optimal for providing good diagnostics for satellite or model validation. Furthermore, these past studies used different retrieval settings, and biases as large as 50 % can be observed in the HCHO total columns depending on these retrieval choices, which is also a weakness for validation studies combining data from different ground-based stations.
For the present work, the HCHO retrieval settings have been optimized based
on experience gained from past studies and have been applied consistently at
the 21 participating stations. Most of them are either part of the Network
for the Detection of Atmospheric Composition Change (NDACC) or under
consideration for membership. We provide the harmonized settings and a
characterization of the HCHO FTIR products. Depending on the station, the
total systematic and random uncertainties of an individual HCHO total column
measurement lie between 12 % and 27 % and between 1 and
This unprecedented harmonized formaldehyde data set from 21 ground-based FTIR
stations is presented and its comparison with a global chemistry transport
model shows consistency in absolute values as well as in seasonal
cycles. The network covers very different concentration levels of
formaldehyde, from very clean levels at the limit of detection (few
Through reactions with hydroxyl radical (OH) and
At present, validation studies of HCHO satellite products have taken place
at a few locations only, mainly using aircraft data
Up to now, time series of HCHO total columns have been studied at only six
FTIR stations out of more than 20 FTIR sites currently in operation:
Ny-Ålesund
Characteristics of the FTIR stations contributing to the present work: location and altitude (in km a.s.l.), time period used in the present study, instrument type, retrieval code, and team.
Table
The formaldehyde spectral signatures used in ground-based infrared
measurements lie in the 3.6
Locations of the FTIR stations providing HCHO total columns.
HBr or
Summary of the HCHO harmonized forward-model and retrieval
parameters. The micro-window limits are given in
We refer to
We summarize the forward-model and retrieval
parameters that have been harmonized in Table
The dominant source of systematic uncertainty being the spectroscopic
parameters, it is crucial that all stations use the same spectroscopic
database. We use the compilation from Geoffrey Toon (JPL), the atm16 line
list, which is available at
Retrieved contributions of all fitted species in the four
MWs
Residuals (calculated – observed spectrum) in each of the four MWs
for the retrieval of a spectrum recorded on 12 February 2014 at Maïdo and
corresponding to a retrieved HCHO total column of
To avoid any bias between the stations due to different spectroscopic
parameters, it is also mandatory to harmonize the spectral micro-windows (MWs)
containing the HCHO signatures. The challenge of the HCHO retrievals is that
this species has very weak absorption signatures in the infrared (below 1 %),
and it is therefore very important to minimize the impact of the interfering
gases with more intense signatures, either by avoiding MWs with
strong interfering lines when feasible or by including them only if
they are very well fitted (e.g. no large residuals remain due to bad
spectroscopic or incorrect ILS parameters). In past studies, while the
micro-window spectral widths differ, some common HCHO signatures were used:
the two more intense signatures at about 2778.5 and 2781.0 cm
In Fig.
In SFIT4 and PROFFIT retrieval codes, based on optimal estimation, a priori
information (profile and regularization matrix) needs to be provided. In this
work, the a priori HCHO profile, as well as all interfering species except
water vapour and its isotopologues, were provided for each station from the
v4 of the model WACCM
It is worth noting that another important forward-model parameter is the
instrumental line shape (ILS) since it impacts the gases absorption line
shapes. The treatment of ILS in the retrievals has not been harmonized yet
among the stations because the stability and quality of the alignment is site
dependent and/or the instrument's PIs have their own preferences. This is,
however, another step toward full harmonization that should be done in the
future within NDACC. At present, there are three options for considering the
ILS, and we refer to
Mean of the HCHO total columns (TC) in
The vertical resolution and sensitivity of the retrieved HCHO products can be
characterized by the averaging kernel matrix
We give the trace of this averaging kernel matrix
The uncertainty budget is calculated following the formalism of
The random uncertainty given in Table
The forward-model parameters error covariance matrices
Effort has been made in this study to harmonize the uncertainty budget at all
sites. This is done by calculating the errors from the same forward-model
parameters (solar zenith angle, temperature, spectroscopic line parameters,
baseline, etc.) across the network and by choosing the same
After the measurement noise error (and the channelling for PROFFIT users),
the largest contributions to the random uncertainty due to the forward-model parameters come from the temperature,
the interfering species, and the off-set baseline. For temperature, the
If one uses the FTIR HCHO measurements to validate a model or a satellite
with a fine vertical resolution, considering the random and systematic
uncertainties (without smoothing) in Table
The smoothing systematic uncertainty, reflecting the bias that would occur on
the retrieved profile if the a priori
The dominating systematic uncertainty sources are the spectroscopic
parameters: the line intensities and the pressure broadening coefficients of
the absorption lines present in our MWs. For the HCHO spectroscopic
parameters, the line list in atm16 follows HITRAN 2012
The other systematic error sources due to forward-model parameters are lower
or within a few percent (ILS, temperature), except for the PROFFIT channelling
source (from 7 % to 17 %), which also has a systematic component. We see from
Table
Overview of the individual HCHO total columns (molec cm
In Fig.
Diurnal cycles of HCHO total columns (molec cm
Second, we show the intermediate concentration sites (with the same vertical
axis with maximum
As explained in the introduction, to reconcile the different results obtained
using satellites observing at different times (e.g. SCIAMACHY and GOME-2
measuring in the morning and OMI in the afternoon), it is crucial to have
ground-based observations of the HCHO diurnal cycles
We see from Fig.
The length of the HCHO time series allows trends to be derived for some
stations. We have calculated the trends at each station using the monthly
mean time series
It turned out that, due to the very high variability of HCHO, the
uncertainties in the trends are often too large to obtain significant values.
A more sophisticated multi-regression model might be able to reduce the
uncertainties, but this is beyond the scope of this paper. However, for a few
stations, significant trends are found. They are mainly negative: at
St Petersburg (
For the longest time series, we observe a very good agreement with previous
studies in general. The negative trends observed over the European stations
St Petersburg and Zugspitze are in agreement with the negative trends
observed by OMI (2004–2014) over St Petersburg and Germany
In this study, we do not aim to validate the model input parameters or attribute different emission sources at the different stations. We use the model to assess the internal consistency of the network using harmonized retrieval settings. This means that we expect that, for the same latitude regions and/or type of sites (polluted; clean), the comparisons with the model will give consistent biases. In the Supplement we provide a global map of IMAGES climatological daytime HCHO columns (2005–2015) together with the mean columns observed at the FTIR stations (Fig. S3). This map illustrates the very different levels covered by the FTIR stations and the overall good agreement with the calculated levels of IMAGES. However, Fig. S3 can only provide a qualitative comparison due to the different measurement periods covered. We give quantitative comparisons in the present section.
The IMAGESv2 global model calculates the distribution of 170 chemical
compounds gases with a time step of 6 h at
Anthropogenic emissions of
The chemical degradation mechanism of pyrogenic NMVOCs is described in
Monthly means of HCHO total columns (molec cm
The calculation of the model columns at the FTIR stations accounts for its
location in the horizontal (nearest model pixel) for the FTIR a
priori profiles and averaging kernels as prescribed in
Seasonal cycle of HCHO total columns (molec cm
Correlation (Corr), bias
We compare the monthly means of FTIR HCHO total columns at each station with
the IMAGES columns calculated for the 2003–2016 period. The time series of
both products are shown in Fig.
For each station the correlation, the bias and the standard deviation (SD) of
the statistical comparisons between the monthly means,
The median of IMAGES and FTIR differences is small (
We distinguish two groups of Arctic sites, Eureka, Ny-Ålesund and Thule, which
are very remote (77–80
Very similar biases
(
The North American sites Toronto and Boulder give similar biases (
The mountain sites are more difficult to model, especially when they are close
to cities. They are often very clean sites, but the model cannot reproduce
this at the current resolution (2
At the mountain site of Izaña, located in a clean marine area, the model
and FTIR are in overall good agreement (
A moderate positive model bias is calculated at Mauna Loa (
The model falls short in reproducing the enhanced HCHO levels observed at
Mexico City (ca.
Comparison at two sites in South America, the coastal site of Paramaribo and
the Porto Velho site at the edge of the Amazon rainforest, indicates a
consistent model overestimation (
The two marine sites at Reunion Island (Saint-Denis at sea level, and Maïdo at
2.2 km altitude) show a small model bias (
The Wollongong site shows the same behaviour as most of the Northern
Hemisphere sites: an overall underestimation of the model (
Since the time series at Saint-Denis, Wollongong and Lauder have been
published in the past using different retrieval strategies
Only five NDACC FTIR sites have delivered HCHO time series until now
We have presented the retrieval settings that have been optimized for this
challenging species, and the FTIR HCHO products have been characterized by
their averaging kernels and their uncertainty budget. The systematic
uncertainty of an individual HCHO total column measurement lies between
12 % and 27 %, with some differences remaining between the SFIT4 code
users (12 %–15 %) and the PROFFIT users (12 %–27 %), which
needs to be investigated in the future within the NDACC InfraRed Working
Group. The random uncertainty lies between 1 and
In addition to the well-defined seasonal cycles, the diurnal cycles were
presented at each site. These observations are crucial for interpreting the
differences observed between satellites measuring at different local times.
For example, the diurnal cycle at Porto Velho which shows insignificant
variations suggests that the negative bias observed over Rondônia between
OMI (13:30 LT) and GOME-2 (09:30 LT)
The monthly mean time series as well as the seasonal cycles have been
compared to the IMAGES model. We did not aim to evaluate the model but show
that the FTIR network provides coherent absolute values and seasonal cycles.
We observed an overall good agreement with IMAGES, which usually (but not
always) underestimated the HCHO total columns (median bias
These HCHO time series, harmonized and well characterized, provide an important data set for past and present satellites, and model validation. They are continuously extended by new measurements and will be used in the coming years for the validation of new satellites, such as Sentinel 5P and Sentinel 4.
The FTIR data sets can be provided in the public NDACC
repository (
The authors declare that they have no conflict of interest.
This study has been supported by the ESA PRODEX project TROVA (2016–2018) funded by the Belgian Science Policy Office (Belspo). NCAR is supported by the National Science Foundation. The NCAR FTS observation programmes at Thule and Mauna Loa are supported by a contract with the National Aeronautics and Space Administration (NASA). The Thule work is also supported by the NSF Office of Polar Programs (OPP). We wish to thank the Danish Meteorological Institute for support at the Thule site and NOAA for support at the Mauna Loa site. Eureka measurements were made at the Polar Environment Atmospheric Research Laboratory (PEARL) under the CANDAC and PAHA projects led by James R. Drummond, and in part by the Canadian Arctic ACE/OSIRIS Validation Campaigns, led by Kaley A. Walker. Funding was provided by AIF/NSRIT, CFI, CFCAS, CSA, ECCC, GOC-IPY, NSERC, NSTP, OIT, PCSP, and ORF. Logistical and operational support was provided by PEARL Site Manager Pierre Fogal, the CANDAC operators, and the ECCC Weather Station. Toronto measurements were made at the University of Toronto Atmospheric Observatory, supported by CFCAS, ABB Bomem, CFI, CSA, ECCC, NSERC, ORDCF, PREA, and the University of Toronto. The measurements at Reunion Island have been also supported by the Université de La Réunion and CNRS (LACy-UMR8105 and UMS3365), and at Porto Velho by the BRAIN-pioneer project IKARE, funded by Belspo. The measurements at Paramaribo have been supported by the BMBF (German Ministry of Education and Research) in the project 5 O3CHEM (01LG1214A). We thank the Meteorological Service Suriname for support. The measurements and data analysis at Bremen are supported by the Senate of Bremen. The measurements at the St Petersburg site (SPbU) have been supported by the Russian Science Foundation (project no. 14-17-00096). Observational facilities have been provided by the Centre for Geo-Environmental Research and Modelling (GEOMODEL) of SPbU. Analysis of FTIR data acquired at SPbU has been performed with the financial support of the Russian Foundation for Basic Research (project no. 18-05-00011). The measurements at Lauder are core-funded by NIWA, through New Zealand's Ministry of Business, Innovation and Employment. We are grateful to Sorbonne Université and Région Île-de-France for their financial contributions as well as to Institut Pierre-Simon Laplace for support and facilities. The Altzomoni and Mexico City measurements have been funded by DGAPA, PAPIIT (nos. IN112216 and IN111418) as well as CONACYT (nos. 275239 and 239618). The German partners acknowledge BMWi for support in HCHO data analysis. The authors would like to thank essential people for the FTIR measurements (Cristian Hermans, Nicolas Kumps, Francis Scolas, Minqiang Zhou, BIRA-IASB; Christiane Silvestrini de Morais, IFRO; Uwe Raffalski, IRF; Eliezer Sepulveda, AEMET; Sukarni Mitro: Meteorological Service of Suriname; Pascal Jeseck, LERMA-IPSL; Alejandro Bezanilla, Omar Lopez, Miguel Angel Robles, Alfredo Rodriguez Manjarrez, Delibes Flores Roman: CCA-UNAM). We thank the station personnel at the AWIPEV research base in Ny-Ålesund, Spitsbergen, for taking the measurements. We also thank the AWI Bremerhaven for logistical support. Edited by: Michel Van Roozendael Reviewed by: three anonymous referees