Toward High‐Resolution Global Atmospheric Inverse Modeling Using Graphics Accelerators

Chevallier, F.; Lloret, Zoé; Takache, Sakina; Remaud, Marine

doi:10.1029/2022gl102135

Cited by 6 publications

(4 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ERA5 data are accessible at (Hersbach et al, 2022). CAMS global inversion-optimized greenhouse gas concentrations are accessible at the Copernicus Climate Change Service (C3S) Climate Data Store (CDS) (Chevallier, 2013;Chevallier et al, 2019Chevallier et al, , 2023. The ODIAC emission inventory is documented in Oda and Maksyutov (2015).…”

Section: Data Availability Statementmentioning

confidence: 99%

Urban XCO₂ Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Che,

Lauvaux,

Taquet

et al. 2024

JGR Atmospheres

View full text Add to dashboard Cite

Satellite measurements of urban CO2 plumes offer a global approach to track CO2 emissions for large cities. To examine and to quantify the feasibility of space‐based monitoring, an intensive measurement campaign (MERCI‐CO2) using seven solar‐tracking Fourier transform infrared (FTIR) spectrometers has been conducted over the Mexico City Metropolitan Area (MCMA) to monitor urban emissions and to evaluate Snapshot Area Map (SAM) observations from the NASA's Orbiting Carbon Observatory‐3 (OCO‐3) mission. Once adjusted for their respective averaging kernels, we diagnosed a positive difference between OCO‐3 and FTIR column measurements (1.06 ppm). Thanks to this unprecedented amount of column observations over a large city, we demonstrate that XCO2 gradients within OCO‐3 SAMs align with the inter‐calibrated FTIR measurements (mean bias of 0.3 ppm), confirming the potential to track CO2 emissions from space over large metropolitan areas. XCO2 urban‐rural differences across the FTIR network, show a strong correlation with observed gradients, with Pearson's correlation coefficients (R) around 0.92. The correlation is significantly lower when considering intra‐urban gradients, where R drop to around 0.24. Simulated XCO2 enhancements (ΔXCO2) based on X‐STILT for both FTIR and OCO‐3 show relatively high correlations (R is around 0.6) using high‐resolution footprints and two gridded inventories. Spatial correlations with OCO‐3 improve when aggregating satellite retrievals at coarser resolutions (10 km). Our study demonstrates the capabilities of detecting urban gradients by FTIR network and OCO‐3 SAM observations over MCMA, a promising result to evaluate the evolution of MCMA's emissions over the coming decade.

show abstract

Section: Data Availability Statementmentioning

confidence: 99%

Urban XCO₂ Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Che,

Lauvaux,

Taquet

et al. 2024

JGR Atmospheres

View full text Add to dashboard Cite

show abstract

“…The off-line version only solves tracer transport equations, driven by pre-computed air mass fluxes from a reference run of the full GCM nudged to the 3-hourly horizontal winds from the fifth generation ECMWF reanalysis (ERA5). The code of the off-line transport model corresponds to the one made public by Chevallier et al (2023) with some memory optimizations in order to accommodate the larger arrays of the new resolution. The inversion system, coded in Python and run on CPU, orchestrates the connection across monthly runs of the transport model, coded in Fortran and basically run on GPU, ensuring the coherence and continuity of the inversion process.…”

Section: Inversion Systemmentioning

confidence: 99%

“…The vertical profiles of CO concentration were not ₂ significantly affected by changes in resolution unlike the XCO fields, especially around emission hotspots. ₂ The high computing cost associated with this resolution increase delayed its implementation in the production chain of the CAMS CO inversion product until the code was ported on Graphics Processing ₂ Units (GPUs) in 2023 (Chevallier et al, 2023). The migration also opened the possibility of further resolution increases while maintaining a processing time, or "time to solution", compatible with operational constraints.…”

Section: Introductionmentioning

confidence: 99%

Refining the Global Picture: the Impact of Increased Resolution on CO₂ Atmospheric Inversions using OCO-2 XCO₂ retrievals

Lloret,

Chevallier,

Cozic

2024

Preprint

View full text Add to dashboard Cite

The threat posed by the increasing concentration of carbon dioxide (CO₂) in the atmosphere motivates a detailed and precise estimation of CO₂ emissions and absorptions over the globe. This study refines the spatial resolution of the CAMS/LSCE inversion system, achieving a global resolution of 0.7° latitude and 1.4° longitude, or three times as many grid boxes as the current operational setup. In a two-year inversion assimilating the midday clear-sky retrievals of the column-average dry-air mole fraction of carbon dioxide (XCO₂) from NASA’s second Orbiting Carbon Observatory (OCO-2), the elevated resolution demonstrates an improvement in the representation of atmospheric CO₂, particularly at the synoptic time scale, as validated against independent surface measurements. Vertical profiles of the CO₂ concentration differ slightly above 22 km between resolutions compared to AirCore profiles, and highlight differences in the vertical distribution of CO₂ between resolutions. However, this disparity is not evident for XCO₂, as evaluated against independent reference ground-based observations. Global and regional estimates of natural fluxes for 2015-2016 are similar between the two resolutions, but with North America exhibiting a higher natural sink at high-resolution for 2016. Overall, both inversions seem to yield reasonable estimates of global and regional natural carbon fluxes. The increase in calculation time is less than the increase in the number of operations and in the volume of input data, revealing greater efficiency of the code executed on a Graphics Processing Unit. This allows us to make this higher resolution the new standard for the CAMS/LSCE system.

show abstract

“…Readers interested in running the model should follow the guidelines at http://orchidee.ipsl.fr/index.php/you%E2%80%90orchidee. The version of the LMDz global transport model as used in this study is publicly available from Zenodo at https://doi.org/10.5281/zenodo.7324039 (Chevallier et al., 2023).…”

Section: Data Availability Statementmentioning

confidence: 99%

Temperature dependence of spring carbon uptake in northern high latitudes during the past four decades

Zhu,

Wang,

Ciais

et al. 2023

Global Change Biology

Self Cite

View full text Add to dashboard Cite

In the northern high latitudes, warmer spring temperatures generally lead to earlier leaf onsets, higher vegetation production, and enhanced spring carbon uptake. Yet, whether this positive linkage has diminished under climate change remains debated. Here, we used atmospheric CO2 measurements at Barrow (Alaska) during 1979–2020 to investigate the strength of temperature dependence of spring carbon uptake reflected by two indicators, spring zero‐crossing date (SZC) and CO2 drawdown (SCC). We found a fall and rise in the interannual correlation of temperature with SZC and SCC (RSZC‐T and RSCC‐T), showing a recent reversal of the previously reported weakening trend of RSZC‐T and RSCC‐T. We used a terrestrial biosphere model coupled with an atmospheric transport model to reproduce this fall and rise phenomenon and conducted factorial simulations to explore its potential causes. We found that a strong–weak–strong spatial synchrony of spring temperature anomalies per se has contributed to the fall and rise trend in RSZC‐T and RSCC‐T, despite an overall unbroken temperature control on net ecosystem CO2 fluxes at local scale. Our results provide an alternative explanation for the apparent drop of RSZC‐T and RSCC‐T during the late 1990s and 2000s, and suggest a continued positive linkage between spring carbon uptake and temperature during the past four decades. We thus caution the interpretation of apparent climate sensitivities of carbon cycle retrieved from spatially aggregated signals.

show abstract

Toward High‐Resolution Global Atmospheric Inverse Modeling Using Graphics Accelerators

Cited by 6 publications

References 30 publications

Urban XCO₂ Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Urban XCO₂ Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Refining the Global Picture: the Impact of Increased Resolution on CO₂ Atmospheric Inversions using OCO-2 XCO₂ retrievals

Temperature dependence of spring carbon uptake in northern high latitudes during the past four decades

Contact Info

Product

Resources

About

Toward High‐Resolution Global Atmospheric Inverse Modeling Using Graphics Accelerators

Cited by 6 publications

References 30 publications

Urban XCO2 Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Urban XCO2 Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Refining the Global Picture: the Impact of Increased Resolution on CO₂ Atmospheric Inversions using OCO-2 XCO₂ retrievals

Temperature dependence of spring carbon uptake in northern high latitudes during the past four decades

Contact Info

Product

Resources

About

Urban XCO₂ Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City

Urban XCO₂ Gradients From a Dense Network of Solar Absorption Spectrometers and OCO‐3 Over Mexico City