CO2M Cal/Val requirements science support – Phase I

The success of the CO2M mission will critically depend on its ability to reach the ambitious precision requirements of the Copernicus CO2 Monitoring & Verification Support capacity space component: better than 0.7ppm for XCO2, better than 10ppb for XCH4, and better than 1.5*10^15 molec/cm2 for NO2. To meet the stringent end-to-end performance requirements, in-orbit verification and traceability to on-ground calibration and characterisation is key for the space component [Pinty et al., 2019].

At the start of this service the Cal/Val concept for CO2M was in the early stages of its definition. The service was designed to inform this process and feed into the formal Calibration and Validation Plan (along with contributions from other experts, studies and initiatives relevant to CO2M).

The service addressed the definition of methods, tasks, and data sources (including ground-based fiducial reference data, model data or static data) relevant for routine product monitoring and scientific validation. Where applicable, it also made proposals for improving existing fiducial reference dataset quality and network architecture, as a follow-on to a previous study on ground-based network capacity analysis (Ground-based network capacity analysis for CO2M | EUMETSAT).

The study was performed by Ludwig-Maximilians-Universität München (LMU), in partnership with the Royal Belgian Institute for Space Aeronomy, the University of Bremen, and The Inversion Lab, with LMU also providing the service management.

Objectives

• Identification of datasets for continuous Cal/Val and monitoring of operational GHG products.
• Proposal for improvement of ground-based fiducial reference dataset quality, network architecture and data-provision where needed.
• Definition of scientific methodologies for monitoring and Cal/Val for all proposed datasets.

Overview

The continuous Cal/Val and monitoring function

The continuous Cal/Val and monitoring function (C-CVMF) within the CO2M ground segment will carry out continuous Cal/Val and monitoring tasks on the CO2M products. While Cal/Val will require a considerable amount of time for gathering enough statistics from both on-ground and in-orbit data, continuous monitoring should be able to identify deviations from the norm in near-real time. To support the development of this function, the datasets required, their availability and schematics for data flows and processing needs for both tasks must be specified. The C-CVMF will be implemented on the existing architecture of the EUMETSAT Multi-Mission Elements (MME) for operational monitoring and cal/val, providing rolling archives of all data, processing power and large database capacities, along with automated and user-driven direct access to data flows and databases for user-driven and automated reporting and monitoring.

Identification of datasets for continuous Cal/Val and monitoring of operational GHG products

An assessment of existing ground-based data sets relevant for continuous Cal/Val and monitoring of CO2M products was undertaken. The assessment describes the characteristics of the datasets and their suitability for the purpose of continuous Cal/Val and monitoring of operational data products from the CO2M mission. Special attention was paid to their uncertainty characterisation, availability (timeliness, data usage policy), fitness-for purpose, and sustainability.

The assessment included the following networks and datasets:

• Total Carbon Column Observing Network (TCCON)
• COllaborative Carbon Column Observing Network (COCCON)
• Network for the Detection of Atmospheric Composition Change (NDACC)

Improvement of Fiducial reference data — set quality, network architecture and data provision

CO2M Cal/Val activities will rely to a large extent on ground-based fiducial reference datasets collected by Fourier transform infrared (FTIR) measurements. The usefulness of these datasets depends on various aspects, including their spatio-temporal representativeness and local characteristics of the observational locations that influence the accuracy of the measurements, such as cloudiness/latitude, surface albedo, and aerosol contamination. A database was created describing the station characteristics and instrument parameters, as well as their data products, data quality, and network characteristics of existing stations in the TCCON, COCCON, and NDACC networks. This will be used as input for future network design studies.

Of particular relevance was the footprint of any given measurement, i.e. the geographical region that is 'observed' by the measurement, in the sense that the measurement is sensitive to the surface fluxes over that region. Therefore, footprints were computed for the current total column network, based on the atmospheric transport model TM5.

The other aspect required by a network design study is the specification of a metric that quantifies the quality of a network. Potential target quantities for CO2M are sectoral fossil fuel emissions at the scale of a country or a mega city. The decision on the target quantity is expected to have a major impact on the design of an optimal FTIR network. A focus on city scale may favour multiple FTIR measurement locations over an exemplary city, while global coverage might be more relevant at a country scale.

Visualisation tool

A Geoserver providing visualization tools to show the combined results of the above activities was created and is available at co2m.aeronomie.be. With the completion of this study this tool is no longer being maintained, but accounts can still be requested to view the functionality created to date.

Definition of scientific methodologies for monitoring and cal/val

Methodologies for the quasi-NRT monitoring of CO2M XCO2 and XCH4 were proposed. The methodologies were developed and tested via real satellite data using GOSAT and OCO-2 XCO2 and S5P XCH4. The main application for CO2M is NRT monitoring, in terms of fast and robust identification of potential data quality issues by performing various checks, including comparing actual retrievals with 'expected results'. Reference data from ground-based networks are not expected to be available in NRT. Instead, simple models and past satellite data will be used for comparison. Based on detailed validation results from past periods it will be estimated if validity of expected random and systematic errors are still valid or if degradation of the instrument has occurred. The main purpose of this activity is to raise flags in NRT for issues affecting the XCO2 and XCH4 data quality.

Results

The results from the above activities are summarised in the documentation available below.