Multi-species inversion and IAGOS airborne data for a better constraint of continental-scale fluxes

Abstract

Abstract. Airborne measurements of CO2, CO, and CH4 proposed in the context of IAGOS (In-service Aircraft for a Global Observing System) will provide profiles from take-off and landing of airliners in the vicinity of major metropolitan areas useful for constraining sources and sinks. A proposed improvement of the top-down method to constrain sources and sinks is the use of a multispecies inversion. Different species such as CO2 and CO have partially overlapping emission patterns for given fuel-combustion-related sectors, and thus share part of the uncertainties related both to the a priori knowledge of emissions and to model–data mismatch error. We use a regional modelling framework consisting of the Lagrangian particle dispersion model STILT (Stochastic Time-Inverted Lagrangian Transport) combined with the high-resolution (10 km × 10 km) EDGARv4.3 (Emission Database for Global Atmospheric Research) emission inventory, differentiated by emission sector and fuel type for CO2, CO, and CH4, and combined with the VPRM (Vegetation Photosynthesis and Respiration Model) for biospheric fluxes of CO2. Applying the modelling framework to synthetic IAGOS profile observations, we evaluate the benefits of using correlations between different species' uncertainties on the performance of the atmospheric inversion. The available IAGOS CO observations are used to validate the modelling framework. Prior uncertainty values are conservatively assumed to be 20 %, for CO2 and 50 % for CO and CH4, while those for GEE (gross ecosystem exchange) and respiration are derived from existing literature. Uncertainty reduction for different species is evaluated in a domain encircling 50 % of the profile observations' surface influence over Europe. We found that our modelling framework reproduces the CO observations with an average correlation of 0.56, but simulates lower mixing ratios by a factor of 2.8, reflecting a low bias in the emission inventory. Mean uncertainty reduction achieved for CO2 fossil fuel emissions is roughly 38 %; for photosynthesis and respiration flux it is 41 and 44 % respectively. For CO and CH4 the uncertainty reduction is roughly 63 and 67 % respectively. Considering correlation between different species, posterior uncertainty can be reduced by up to 23 %; such a reduction depends on the assumed error structure of the prior and on the considered time frame. The study suggests a significant uncertainty constraint on regional emissions using multi-species inversions of IAGOS in situ observations.

Optimizing a dynamic fossil fuel CO<sub>2</sub> emission model with CTDAS (v1.0) for an urban area using atmospheric observations of CO<sub>2</sub>, CO, NO<sub><i>x</i></sub>, and SO<sub>2</sub>

“…The idea behind this is that if we reduce the diagonal value (uncertainty) by a factor of 4 the off-diagonal value (parameter correlation) reduces by a factor of 2. This means that the parameter correlation is relatively stronger if the uncertainty is lower (Boschetti et al, 2018). This hypothesis cannot be confirmed by our results, as we only find a correlation of -0.27 between the prior uncertainty and the sensitivity to parameter correlations (defined as the correlation between the posterior uncertainty and the sum of the parameter correlations).…”

“…Previously, Boschetti et al (2018) have used the presence of error correlations between emissions of different species and found that this reduces the posterior uncertainties for all species. They even show that the uncertainty reduction increases with the correlation and that an incorrect definition of the error correlations may cause a systematic bias in the posterior emission estimate.…”

“…Quantification of the uncertainty at an even higher resolution for urban applications has so far been limited (Andres et al, 2016) (Super et al, 2019), also for most local inventories, while a correct definition of the prior error covariance matrix for an inversion is important to get reliable output (Chevallier et al, 2006;Boschetti et al, 2018). This currently complicates the application of DA studies to urban areas.…”

Source Sector Attribution of CO2 Emissions Using an Urban CO/CO2 Bayesian Inversion System

“…The use of additional atmospheric tracer species has been studied for separating CO 2 atmospheric signals into components, such as fossil fuel CO 2 emissions from biogenic signals at continental scales (e.g., Boschetti et al, ) or for whole‐city emissions (e.g., Wong et al, ). Aside from the use of 14 CO 2 to separate out fossil‐fuel‐related CO 2 (CO 2 ff) from biogenic and other CO 2 sources (e.g., Levin et al, ; Turnbull et al, ), studies have also looked at δ 13 C to further distinguish between the gasoline‐related, natural‐gas‐related, and coal‐related influences (e.g., Clark‐Thorne & Yapp, ; Newman et al, ).…”

“…These efforts have found success by focusing on coupling the error correlations to exploit both CO's known relationship to CO 2 and its increased sensitivity to model transport errors. Boschetti et al () included CH 4 to optimize CO 2 broken down by emission sectors and fuel type. Error‐correlation‐based multispecies inversions have shown promise in improving CO 2 emission uncertainties on a large scale.…”