An estimate of the terrestrial carbon budget of Russia using inventory based, eddy covariance and inversion methods

A. J. Dolman, А. Shvidenko, Dmitry Schepaschenko, Philippe Ciais, N. M. Tchebakova, T. Chen, M. K. van der Molen, Luca Belelli Marchesini, T. C. Maximov, Shamil Maksyutov, Ernst‐Detlef Schulze2012 · DOI

TL;DR

This paper provides a comprehensive estimate of Russia's terrestrial carbon budget by integrating multiple methodologies, including inventory-based approaches, eddy covariance, and atmospheric inversions. The findings reveal that Russia's terrestrial ecosystems act as a significant carbon sink, primarily due to its vast forests, with an estimated net uptake of -0.66 Pg C yr⁻¹. This work is crucial for understanding global carbon cycles and informing climate policy.

What Problem Does This Paper Address?

Accurately quantifying the global carbon budget is essential for understanding and mitigating climate change. Terrestrial ecosystems play a vital role in this budget, acting as either sources or sinks of atmospheric carbon dioxide. Russia, with its immense landmass and diverse ecosystems, particularly its vast forests and permafrost regions, is a critical component of this global carbon cycle. However, estimating the net carbon exchange for such a large and complex region presents significant challenges. Previous studies often relied on single methods or limited geographical coverage, leading to uncertainties. This paper aims to address this gap by providing a more robust and integrated estimate of Russia's terrestrial carbon budget, considering multiple data sources and methodologies to improve accuracy and reduce uncertainty.

Methods

The authors employed a multi-faceted approach to estimate Russia's carbon budget, integrating four key methods:

  • Landscape Ecosystem Approach (LEA): This bottom-up method uses inventory data and empirical models to account for carbon stocks and fluxes across different land classes. It combines flux-based calculations (e.g., Net Ecosystem Production) with pool-based assessments of carbon changes.
  • Eddy Covariance (EC): This technique directly measures the exchange of carbon dioxide between the land surface and the atmosphere at specific sites. Data from 14 sites representing various ecosystems were used and then upscaled.
  • Dynamic Global Vegetation Models (DGVMs): These are process-based models that simulate vegetation dynamics and carbon cycling under changing climate and CO2 conditions. The study utilized results from 8 DGVMs.
  • Inverse Modeling: This top-down approach uses atmospheric CO2 concentration measurements to infer the net flux of CO2 from the surface to the atmosphere. The study incorporated results from 11 different inversion models.

The paper also accounts for smaller, but important, fluxes such as those from fires, trade, river export, and land-use change.

Datasets

The study utilized a variety of datasets:

  • Inventory Data: Official Russian statistics, forest inventory data, and agricultural land use data.
  • Remote Sensing Data: NOAA AVHRR data for vegetation indices and burned area.
  • Eddy Covariance Data: Measurements from 14 ecological observatories across Russia.
  • Climate Data: Merged CRU-NCEP forcing dataset for DGVM simulations.
  • Global Datasets: Global Land Cover 2000 (GLC 2000), GFED-3 (Global Fire Emissions Database), and COSCAT (catchments database).
  • Fossil Fuel Emissions Data: Provided by UNFCCC.

Key Insights and Ideas

  • Forest Dominance: Forests are the primary drivers of Russia's carbon sink, accounting for approximately 90-95% of the net biosphere to atmosphere uptake. This highlights the critical role of forest conservation and management.
  • Methodological Convergence: A key finding is the remarkable agreement between the different approaches (bottom-up LEA, eddy covariance, and top-down inversions). This convergence lends strong confidence to the overall estimate of Russia's carbon balance.
  • Permafrost Vulnerability: The study suggests that some areas, particularly in permafrost regions, might switch from being carbon sinks to sources under regional warming due to increased soil respiration. This points to a potential positive feedback loop in the climate system.
  • Land Use Change Impact: The abandonment of agricultural lands since the 1990s has contributed to a significant CO2 sink due to the recovery of soil organic carbon. However, the consumption of agricultural products can offset this sink, turning some areas into carbon sources.
  • Fire Variability: Fire emissions exhibit substantial interannual variability, with fluctuations of up to a factor of 0.5 to 3, underscoring the challenge of accurately modeling this component of the carbon cycle.

Key Results

  • Net Biosphere to Atmosphere Flux: The best estimate for Russia's net biosphere to atmosphere flux is -0.66 Pg C yr⁻¹.
  • Forest Sink: Forests are estimated to take up -0.69 Pg C yr⁻¹.
  • Inverse Model Estimate: An average net biosphere to atmosphere flux of -653 Tg C yr⁻¹ was derived from inverse models, with an interannual variability of 35%.
  • Bottom-Up Estimate: The LEA system provided a bottom-up estimate of -563 Tg C yr⁻¹.
  • Landscape Approach Estimate: An independent landscape approach yielded -761 Tg C yr⁻¹.
  • Net Land to Atmosphere Flux (including fossil fuels): The estimated range is -145 to -73 Tg C yr⁻¹.
  • Methane Emissions: Estimates for methane flux vary, with one inventory-based estimate at 12.6 Tg C-CH4 yr⁻¹ and a model estimate for boreal and Arctic Eurasia at 27.6 Tg C-CH4 yr⁻¹.

Limitations and Future Work

  • Data Uncertainty: The authors acknowledge that uncertainties exist due to the lack of knowledge and insufficient empirical data, particularly concerning soil carbon pools and the reliability of older forest inventory data.
  • Upscaling Eddy Covariance Data: The upscaling of eddy covariance data is limited by the small number of measurement sites and the lack of reliable gradients for extrapolation.
  • DGVM Simplifications: DGVMs tend to oversimplify land cover, may ignore crucial land classes (like bogs), and underestimate disturbances.
  • Regional Completeness: Reliable carbon budget information for Ukraine, Belarus, and Kazakhstan is scarce, limiting the completeness of the regional analysis.
  • Future Work: The paper suggests that further research is needed to improve the accuracy of estimates, particularly for fluxes related to disturbances, and to better understand the vulnerability of permafrost regions to climate change. The authors also highlight the need for improved methodologies to assess uncertainties.

Summary updated

Abstract

We determine the carbon balance of Russia, including Ukraine, Belarus and Kazakhstan using inventory based, eddy covariance, Dynamic Global Vegetation Models (DGVM), and inversion methods. Our current best estimate of the net biosphere to atmosphere flux is −0.66 Pg C yr−1. This sink is primarily caused by forests that using two independent methods are estimated to take up −0.69 Pg C yr−1. Using inverse models yields an average net biopshere to atmosphere flux of the same value with a interannual variability of 35 % (1σ). The total estimated biosphere to atmosphere flux from eddy covariance observations over a limited number of sites amounts to −1 Pg C yr−1. Fires emit 137 to 121 Tg C yr−1 using two different methods. The interannual variability of fire emissions is large, up to a factor 0.5 to 3. Smaller fluxes to the ocean and inland lakes, trade are also accounted for. Our best estimate for the Russian net biosphere to atmosphere flux then amounts to −659 Tg C yr−1 as the average of the inverse models of −653 Tg C yr−1, bottom up −563 Tg C yr−1 and the independent landscape approach of −761 Tg C yr−1. These three methods agree well within their error bounds, so there is good consistency between bottom up and top down methods. The best estimate of the net land to atmosphere flux, including the fossil fuel emissions is −145 to −73 Tg C yr−1. Estimated methane emissions vary considerably with one inventory-based estimate providing a net land to atmosphere flux of 12.6 Tg C-CH4yr −1 and an independent model estimate for the boreal and Arctic zones of Eurasia of 27.6 Tg C-CH4 yr−1.
BibTeX
@misc{dolman2012estimate,
  author = {A. J. Dolman and А. Shvidenko and Dmitry Schepaschenko and Philippe Ciais and N. M. Tchebakova and T. Chen and M. K. van der Molen and Luca Belelli Marchesini and T. C. Maximov and Shamil Maksyutov and Ernst‐Detlef Schulze},
  title = {{An estimate of the terrestrial carbon budget of Russia using inventory based, eddy covariance and inversion methods}},
  year = {2012},
  doi = {10.5194/bgd-9-6579-2012},
}
Figure