GHG Session 3

RAL has developed an optimal estimation scheme to retrieve height resolved information on CH4 from IASI measurements in the thermal infrared (7.9 micron) band. This scheme has been extensively validated in the context of UK national (NERC/NCEO) and Eumetsat funded projects and shown to perform well in comparison with column-averages from both ground-based observations and the short-wave sounder GOSAT. The scheme extracts two independent pieces of information on the profile, with sensitivity extending into the lower troposphere. Column-averaged mixing ratios with a precision of 20-40ppb can be derived, however sensitivity to the boundary layer is weak compared to short-wave sounders (and dependent on the atmospheric/surface state). IASI also measures methane lines in the 3.7 micron spectral range, where there is significant solar contribution to measurements in the daytime over land; increasing sensitivity to methane in the boundary layer. However, this spectral range has not yet been exploited due to the complexity of modelling both solar and thermal contributions to the measurements and the relatively high noise level of IASI in this spectral region. In this paper we will summarise the performance of the thermal infrared scheme and go on to present findings from an investigation into the potential of adding information on near-surface methane by exploiting the 3.7 micron band.

Carbonyl chloride (COCl₂), better known as phosgene, is a toxic gas and it was mainly used by chemical industry in the preparation of insecticides, pharmaceuticals and herbicides. Its usage has been reduced over the years due to its high toxicity. To study possible seasonal and latitudinal variations of COCl₂ in the upper troposphere / lower stratosphere (UTLS) we used the measurements from the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS). MIPAS is a Fourier Transform Spectrometer that measured atmospheric limb emission spectra from a polar orbit on board the ESA ENVISAT satellite in the time span from June 2002 to April 2012. Retrievals of phosgene were carried-out using the ν₅ band of COCl₂ in the 830-860 cm^-1  spectral region. In the same spectral region we find the ν₄ band of CCl₃F (CFC-11), that is much stronger than the COCl₂ signal. Therefore phosgene spectral features are hidden by the CFC-11 absorbing band. To cope with this strong interference, CFC-11 and phosgene abundances have been retrieved simultaneously. The retrieval scheme used in this work is based on the so called Optimized Retrieval Model (ORM), the scientific version of ESA level 2 processor, upgraded with the Multi-Target Retrieval (MTR) functionality and with the possibility to use Optimal Estimation (OE) to apply external constraints to the state vector. We carry-out the forward simulations using state-of-the art spectroscopic line data calculated at LISA-CNRS, not yet included in the HITRAN database. We show the global distribution of phosgene in the UTLS region and an analysis of its seasonal and latitudinal variability.

The Carbon Cycle Data Assimilation System (CCDAS, Rayner 2005, Knorr 2008) is used to increase our knowledge about the global carbon cycle by reducing the uncertainties in regional sources/sinks in the terrestrial biosphere. This is achieved through the simultaneous assimilation of complementing satellite data streams, including column integrated CO2 (XCO2), into the land biosphere model BETHY (Knorr 2000) and a subsequent propagation of posterior uncertainties to terrestrial fluxes.

BETHY simulates carbon assimilation and plant and soil respiration, embedded within a full energy and water balance. Atmospheric CO2 concentrations are derived from CO2 surface fluxes using the precomputed Jacobian of the Transport Model TM3. The data assimilation system uses the Bayesian approach by defining a cost function which quantifies the misfit  between observations and their model equivalence, assuming Gaussian error distribution. The cost function depends on a number of process parameters of BETHY and the mean global CO2 concentration. The best fit of the model to the observations is at the minimum of the cost function. It is found by a BFGS algorithm using the gradient information computed by the adjoint model. The adjoint code is generated by FastOpt's Automatic Differentiation tool Transformation of Algorithm in Fortran (TAF). By this procedure dynamically consistent Net and Gross Surface CO2 fluxes are found.

Data streams integrated in CCDAS are:
- Fraction of Absorbed Photosynthetic Active Radiation (fapar) and Leaf Area Index (LAI) from 2Stream Inversion Package (TIP-GlobAlbedo)
- soil moisture from SMOS
- surface CO2 concentration from FluxNet sites
- XCO2 satellite product from SCIAMACHY

All model-data misfits are weighted by the inverse error estimates. The TIP GlobAlbedo fapar and LAI data misfits are weighted by their inverse error correlation matrix (preliminary results of QA4ECV). Prior estimates of the process parameters are used and deviations are weighted with assumed uncertainties. Posterior error estimates of the parameters are derived from the Hessian of the cost function at the minimum. This error is propagated by the adjoint and tangent of the prediction operator to error estimates of predicted quantities (Scholze 2007). Again these derivative codes (Hessian,adjoint,tangent) are generated by TAF.

Preliminary result show a reduction in the posterior error estimates of net terrestrial carbon fluxes in 2010 compared to assimilation without the XCO2 data.

The inter-annual variations of fire, wetland, and terrestrial ecosystem carbon fluxes – namely CO₂, CH₄ and CO - remain poorly understood in the tropics. For example, the roles of wetland extent, carbon availability and temperature in controlling wetland CH₄ emissions are uncertain; similarly, the temporal effects of drought on fire emissions and fire traits (such as combustion efficiency and completeness) are not well quantified. Our study is focused on tropical South America, which accounts for 30-55% of wetland CH₄ and 5-35% of fire CH₄ in the tropics. To de-convolve South America seasonal and inter-annual CH₄ emissions during 2003-2013, we characterize the temporal variability of wetland and fire CH₄ fluxes based on satellite data. We derive an ensemble of wetland CH₄ scenarios based on ERS, AVHRR, SSM/I, QuikSCAT & ASCAT wetland extent dataset, and carbon fluxes from the MsTMIP terrestrial biosphere model ensemble. We use a Bayesian fusion of land surface and atmospheric data to constrain inter-annual variations in fire carbon fluxes, and associated CH₄ fluxes: our fire emissions analysis is based on TES CO/CH₄ observations, MOPITT CO measurements, MODIS burned area OMI NO₂, GOME-2 fluorescence and GLAS-derived above-ground biomass. To evaluate the role of fire and wetland emissions in temporal variability of South America CH₄, we use SCIAMACHY CH₄, GOSAT CH₄ and TES CH₄/CO data to confront an ensemble of atmospheric chemistry and transport model (GEOS-Chem) runs. We conclude by demonstrating the future synergetic potential of OCO-2 CO₂ data to constrain fire traits, carbon fluxes and the overall tropical carbon budget. 

Interannual variations in the atmospheric growth rate of CO2 have been attributed to the tropical regions and the controls are correlated with  temperature anomalies.  We investigate the spatial drivers of the atmospheric growth rate and the processes controlling them over the exceptional period of 2010-2011.  This period was marked by a marked shift from an El Nino to La Nina period resulting in historically high sea surface temperature anomalies in the tropical Atlantic leading to serious droughts in the Amazon.  However, in 2011, unusual precipitation in Australia was linked to gross primary productivity anomalies in semi-arid regions.  We use satellite observations of CO2, CO, and solar induced fluorescence assimilated into the NASA Carbon Monitoring System Project (CMS-Flux) to attribute the atmospheric growth rate to global, spatially resolved fluxes.  This system is based upon observationally-constrained “bottom-up” estimates from the Fossil Fuel Data Assimilation System (FFDAS), the ECCO2­-Darwin physical and biogeochemical adjoint ocean state estimation system, and CASA-GFED3 land-surface biogeochemical model.  The system is used to compute regional tropical and extra-tropical fluxes  and quantify the role of biomass burning and gross primary productivity in controlling those fluxes.  The potential of OCO-2 and Sentinel-5p to continue these critical observations used in this analysis is explored.  

and the CarbonSat Team

M. Buchwitz ^a, M. Reuter ^a, K. Gerilowski ^a, T. Krings ^a, O. Schneising ^a, J. P. Burrows ^a

H. Bösch^b, D. Brunner ^c, P. Ciais ^d, F.M. Breon ^d, D. Crisp ^e, H. Dolman ^f, G. Hayman ^g, S. Houweling ^h, J. Landgraf ^h, G. Lichtenberg ⁱ, P. Ingmann ^j, B. Sierk ^j, A. Loescher ^j, Y. Meijer ^j,

^a University of Bremen, Institute of Environmental Physics, Bremen, Germany

^b University of Leicester, UK

^c EMPA, Dübendorf, CH

^d LSCE, Gif-sur-Yvette, F

^e Caltech/JPL, Pasadena, USA

^f Free University, Amsterdam, NL

^g Centre for Ecology and Hydrology, Wallingford, UK

^h SRON, Utrecht, NL

ⁱ DLR-IMF, Oberpfaffenhofen, D

^j ESA ESTEC, Noordwijk, NL

The greenhouse gases (GHG), carbon dioxide (CO₂) and methane (CH₄), are the main drivers of climate change, the most important environmental challenge in the 21^st century. Understanding and quantifying climate feedback and forcing mechanisms involving these two dominant anthropogenic greenhouse gases requires the discrimination of natural and anthropogenic CO₂ and CH₄ fluxes globally, with regional to local spatial scale resolution. To address this need, CarbonSat was selected by ESA as one of two candidates for the Earth Explorer Opportunity mission (EE8) to be the next step in the evolution of space based GHG observations. The main goal of the proposed ESA CarbonSat Mission is to advance our knowledge on the natural and man-made sources and sinks of these greenhouse gases over this range of scales. The unique feature of the CarbonSat mission concept is its “GHG imaging capability”, which is achieved by combining high spatial resolution (6 km²) and good spatial coverage (up to 240 km wide swath, with contiguous ground sampling). CarbonSat aims to deliver spatially-resolved, time varying global estimates of dry column mixing ratios of CO₂ and CH₄ with high precision (~1 to 2 ppm and ~12 ppb, respectively) and relative accuracy (<0.5 ppm and < 5 ppb, respectively on regional scales). Benefiting from its imaging capabilities along and across track, CarbonSat will provide at least an order of magnitude larger number of cloud-free CO₂ soundings than OCO-2. CarbonSat will be the first satellite mission to image small scale emission hot spots of CO₂ (e.g., cities, volcanoes, industrial areas) and CH₄ (e.g., fossil fuel production, landfills, seeps) and to quantify their emissions and discriminate them from surrounding biospheric fluxes. CarbonSat will also reveal regional details of natural fluxes, allowing discriminating among regions and across seasons the largest carbon sinks and the areas that may loose CO₂ or CH₄ to the atmosphere.This will allow a better separation of natural and anthropogenic GHG sources and sinks also on larger scales. The latter will be further supported by CarbonSat’s ability to constrain the fluxes of CO₂ exchanged between the land biosphere and the atmosphere by simultaneously measuring CO₂ and solar induced chlorophyll fluorescence (SIF), a process associated with Gross Primary Production (GPP). Source/sink information will be derived from the retrieved atmospheric column-averaged mole fractions of CO₂ and CH₄ via inverse modelling.  

 Recent results from the CarbonSat scientific studies and supporting campaigns documenting the expected data quality and potential application areas will be presented.