The fast atmospheric trace gas retrieval for OCO-2 (FOCAL-OCO2) has been setup to retrieve XCO2
(the column-average dry-air mole fraction of atmospheric CO2) by analyzing hyper spectral solar
backscattered radiance measurements of NASA's OCO-2 satellite.
FOCAL includes a radiative transfer model which has been developed to approximate light
scattering effects by multiple scattering at an optically thin scattering layer.
This reduces the computational costs by several orders of magnitude.
FOCAL's radiative transfer model is utilized to simulate the radiance in all three OCO-2
spectral bands allowing the simultaneous retrieval of CO2, H2O, and solar induced chlorophyll
fluorescence.
FOCAL's accuracy and precision has been determined by comparison with collocated ground based TCCON (Total Column Carbon Observing Network) FTS (Fourier transform spectrometer) measurements.
Regional-scale (station-to-station) biases amount to about 0.6ppm.
The single measurement precision (standard deviation of the difference to TCCON) is about 1.8ppm.
Extension of the data set now covering 09/2014-02/2022.
Improvement of bias correction.
Improvement of uncertainty estimates.
v10
Generation of a global data set (09/2014-12/2021) based on OCO-2 v10 L1b data.
Relaxation of the latitude preprocessor filter from 70° to 80°.
Removal of the OMI L3 based aerosol preprocessor filter.
Replacement of the MODIS based preprocessor cloud filter by a random
forest classifier which is trained with the MODIS MYD35 cloud mask and
which analyzes OCO-2 L1b radiances.
Removal of the preprocessor filter for the number of spectral spikes.
Spectral spikes are now handled equally to bad detector pixels by masking
affected parts of the spectra in the retrieval.
Improvement of the RT model by assuming an isotropic instead of a
Lambertian scattering phase function at the scattering layer.
Introduction of HDO in the RT model and usage of delta-d as state vector element.
Update of the CO2, O2, and H2O cross section tables to ABSCO v5.1
and adding H2O as absorber in the O2 fit window.
Update of the CO2 a priori to the Simple cLImatological Model for
atmospheric CO2 (SLIMCO2) which bases on a climatology data base constructed
from 16 years of NOAA CarbonTracker data (Noël et al., 2021b).
Update of the CO2 a priori error covariance matrix corresponding to
SLIMCO2 scaled to an XCO2 a priori uncertainty of 7.5ppm.
Replacement of the postprocessing bias correction based on the small area
assumption by a random forest regressor trained with a reference data
base consisting of NOAA CarbonTracker model data in regions justified
with TCCON (Noël et al., 2021a).
Improvement of the computational efficiency.
v09
Generation of a global 5-years data set (2015-2019) based on OCO-2 v8 L1b data and extension till 05/2020 based on OCO-2 v10 L1b data.
Migration of L2 processor from IDL to Python.
Usage of previous results as first guess state vector (except for albedo) in order to improve convergence behavior.
This acceleration is only applied for soundings of the same orbit having distances below 25km.
Additionally, the maximum number of successive accelerated soundings is limited to 25.
v08
Generation of a global 4-years data set (2015-2018).
Improved cross section data bases with finer temperature, pressure, and wavelength grid in the wCO2 (0.0026nm) and sCO2 (0.0044nm) band.
Quadratic wavelength and linear pressure interpolation of the cross section data base.
Usage of HITRAN2016 as H2O spectroscopy.
Allowing negative values of ps for improved convergence behavior.
Widened limits for improved convergence behavior.
Improved smoothing and noise error diagnostics.
Usage of ECMWF ERA5 meteorological data.
v06
First application to a larger global dataset of actually measured OCO-2 data and validation as described by Reuter et al. (2017b).
Development of a preprocessor including filtering, adaptation of the noise model, and zero level offset correction.
Development of a postprocessor including filtering and bias correction.
Implementation of a Levenberg-Marquardt minimizer.
v01
Initial version as described by Reuter et al. (2017a).
Analyzes of simulated OCO-2 measurements.
Data Access
The data-format is NCDF, one file per day, and organized as ESA's GHG CCI CRDP#4 data format.
All NCDF files are compressed into annual tar.gz files.
For more information on the data format, algorithm theoretical basis, validation, etc. please have a look at the literature section of this website.
Condition of Use
When using FOCAL data, you agree ...
... to inform us prior to any publication where FOCAL data products are planned to be used,
... to offer us co-authorship for any planned peer-reviewed publication based on FOCAL data products (for non peer-reviewed publications it is sufficient if you add an appropriate acknowledgement),
... not to distribute the FOCAL data products to any third party (the only exception being colleagues working in your institute, in this case you agree to inform them about the conditions listed here and that they also have to accept these conditions).
Acknowledgements
FOCAL is in part funded by the European Union via the Copernicus Climate Change Service and the H2020 projects CHE (Grant Agreement No. 776186) and VERIFY (Grant Agreement No. 776810), ESA/ESRIN (GHG-CCI/GHG-CCI+ projects), and the State and the University of Bremen.
ECMWF provided the used meteorological profiles.
The OCO-2 data were produced by the OCO-2 project at the Jet Propulsion Laboratory, California Institute of Technology, and obtained from the OCO-2 data archive maintained at the NASA Goddard Earth Science Data and Information Services Center.
Additionally, NASA provided the MODIS L2 collection 6 MYD35 cloud mask data, the OMI L3 OMAERUVd v003 UV aerosol index data, and the absorption cross section database ABSCO.
The used solar spectra were made available by R. L.Kurucz and G.C. Toon.
The used chlorophyll fluorescence spectrum has been published by U.Rascher.
TCCON data were obtained from the TCCON Data Archive, hosted by the Carbon Dioxide Information Analysis Center (CDIAC).
