Variational Data Assimilation in the Constructor of Dynamic Soil Carbon Models

Siumbel K. Shangareeva; Victor M. Stepanenko; George M. Faykin; Alexander I. Medvedev; Irina M. Ryzhova; Vladimir A. Romanenkov

doi:10.14529/jsfi250406

Authors

Siumbel K. Shangareeva Lomonosov Moscow State University, Research Computing Center, Moscow, Russian Federation https://orcid.org/0009-0004-0717-2320
Victor M. Stepanenko Lomonosov Moscow State University, Research Computing Center, Moscow, Russian Federation; Moscow Center of Fundamental and Applied Mathematics, Moscow, Russian Federation https://orcid.org/0000-0003-3033-6712
George M. Faykin Lomonosov Moscow State University, Research Computing Center, Moscow, Russian Federation https://orcid.org/0000-0002-3595-2808
Alexander I. Medvedev Lomonosov Moscow State University, Research Computing Center, Moscow, Russian Federation https://orcid.org/0009-0001-2045-5446
Irina M. Ryzhova Lomonosov Moscow State University, Soil Science Faculty, Moscow, Russian Federation https://orcid.org/0000-0002-8325-5222
Vladimir A. Romanenkov Lomonosov Moscow State University, Soil Science Faculty, Moscow, Russian Federation https://orcid.org/0000-0002-8967-4225

DOI:

https://doi.org/10.14529/jsfi250406

Keywords:

data assimilation, carbon dynamic models, adjoint model, automatic differentiation

Abstract

This work presents an automatic adjoint-model construction within the Carbon Cycle Model Constructor (CCMC) that enables variational data assimilation (VDA) for estimating the initial state of soil dynamic carbon models. The adjoint is generated once from the generic pool-flux representation used in CCMC, which allows efficient gradient evaluation and iterative optimization of the initial pool vector without constructing a model-specific adjoint. The proposed approach is tested with two soil carbon models: SOCS (Soil Organic Carbon Saturation) and RothC (Rothamsted model). Data assimilation experiments are performed using long-term field observations of soil carbon content. The entire VDA workflow, including the adjoint solver and optimization algorithm, is implemented in the same Fortran code base as CCMC. CCMC+VDA implementation is fully compatible with the MPI+OpenMP TerM land surface model and provides a reusable, scalable foundation for variational soil-carbon data assimilation on modern supercomputers.

References

Pryanishnikov Institute of Agrochemistry. https://www.vniia-pr.ru/laboratorii/otdl-geoseti/lab-geogr-seti/ (2025), accessed: 2025-09-30

Benavides Pinjosovsky, H.S., Thiria, S., Ottlé, C., et al.: Variational assimilation of land surface temperature within the ORCHIDEE Land Surface Model Version 1.2.6. Geoscientific Model Development 10(1), 85–104 (2017). https://doi.org/10.5194/gmd-2016-64

Coleman, K., Jenkinson, D.S.: RothC-26.3-A Model for the turnover of carbon in soil. In: Evaluation of soil organic matter models: Using existing long-term datasets, pp. 237–246. Springer (1996)

Duchi, J., Hazan, E., Singer, Y.: Adaptive subgradient methods for online learning and stochastic optimization. Journal of Machine Learning Research 12, 2121–2159 (2011). https://doi.org/10.5555/1953048.2021068

Faykin, G.M., Stepanenko, V.M., Medvedev, A.I., et al.: Constructor of soil carbon dynamic models. Numerical Methods and Programming 26(3), 281–303 (2025). https://doi.org/10.26089/NumMet.v26r320

Herbst, M., Welp, G., Macdonald, A., et al.: Correspondence of measured soil carbon fractions and RothC pools for equilibrium and non-equilibrium states. Geoderma 314, 37–46 (2018). https://doi.org/10.1016/j.geoderma.2017.10.047

Kaminski, T., Knorr, W., Schürmann, G., et al.: The BETHY/JSBACH carbon cycle data assimilation system: Experiences and challenges. Journal of Geophysical Research: Biogeosciences 118(4), 1414–1426 (2013). https://doi.org/10.1002/jgrg.20118

Kuppel, S., Peylin, P., Chevallier, F., et al.: Constraining a global ecosystem model with multi-site eddy-covariance data. Biogeosciences 9(10), 3757–3776 (2012). https://doi.org/10.5194/bg-9-3757-2012

Lions, J.L.: Contrôle optimal de systèmes gouvernés par des équations aux dérivées partielles. Dunod Gauthier-Villars (1968)

Marchuk, G.I., Orlov, V.V.: On the theory of conjugate functions. In: Krupchinsky, P.A. (ed.) Neutron physics. Sat. articles. Atomizdat, Moscow. p. 30–45. (1961)

Peylin, P., Bacour, C., MacBean, N., et al.: A new stepwise carbon cycle data assimilation system using multiple data streams to constrain the simulated land surface carbon cycle. Geoscientific Model Development 9(9), 3321–3346 (2016). https://doi.org/10.5194/gmd-9-3321-2016

Raoult, N.M., Jupp, T.E., Cox, P.M., Luke, C.M.: Land-surface parameter optimisation using data assimilation techniques: the adJULES system V1.0. Geoscientific Model Development 9(8), 2833–2852 (2016). https://doi.org/10.5194/gmd-9-2833-2016

Ryzhova, I.M.: Analysis of soil organic matter dynamics based on minimal carboncycle models. Russia’s Soils-Strategic Resource: Abstracts of the 8th Congress of the V. V. Dokuchaev Soil Science Society and the School of Young Scientists on Soil Morphology and Classification (Syktyvkar, 2020–2022) 2, 130–131 (2021)

Sasaki, Y.: An objective analysis based on the variational method. Meteor. Soc. Japan (1958)

Sasaki, Y.: Some basic formalisms in numerical variational analysis. Monthly Weather Review 98(12), 875–883 (1970). https://doi.org/10.1175/1520-0493(1970)098<0875:SBFINV>2.3.CO;2

Scholze, M., Kaminski, T., Knorr, W., et al.: Simultaneous assimilation of SMOS soil moisture and atmospheric CO2 in-situ observations to constrain the global terrestrial carbon cycle. Remote sensing of environment 180, 334–345 (2016). https://doi.org/10.1016/j.rse.2016.02.058

Scholze, M., Kaminski, T., Rayner, P., et al.: Propagating uncertainty through prognostic carbon cycle data assimilation system simulations. Journal of Geophysical Research: Atmospheres 112(D17) (2007). https://doi.org/10.1029/2007JD008642

Shutyaev, V.P.: Methods for observation data assimilation in problems of physics of atmosphere and ocean. Izvestiya, Atmospheric and Oceanic Physics 55(1), 17–34 (2019). https://doi.org/10.1134/S0001433819010080

Stepanenko, V.M.: River routing in the INM RAS-MSU land surface model: Numerical scheme and parallel implementation on hybrid supercomputers. Supercomputing Frontiers and Innovations 9(1), 32–48 (2022). https://doi.org/10.14529/jsfi220103

Stepanenko, V.M., Medvedev, A.I., Bogomolov, V.Y., et al.: Land surface scheme TerM: the model formulation, code architecture and applications. Russian Journal of Numerical Analysis and Mathematical Modelling 39(6), 363–377 (2024). https://doi.org/10.1515/rnam-2024-0031

Taghizadeh-Toosi, A., Cong, W.F., Eriksen, J., et al.: Visiting dark sides of model simulation of carbon stocks in European temperate agricultural soils: allometric function and model initialization. Plant and Soil 450(1), 255–272 (2020). https://doi.org/10.1007/s11104-020-04500-9

Thum, T., Zaehle, S., Köhler, P., et al.: Modelling sun-induced fluorescence and photosynthesis with a land surface model at local and regional scales in northern Europe. Biogeosciences 14(7), 1969–1987 (2017). https://doi.org/10.5194/bg-14-1969-2017

Verbeeck, H., Peylin, P., Bacour, C., et al.: Seasonal patterns of CO2 fluxes in Amazon forests: Fusion of eddy covariance data and the ORCHIDEE model. Journal of Geophysical Research: Biogeosciences 116(G2) (2011). https://doi.org/10.1029/2010JG001544

Wutzler, T., Reichstein, M.: Soils apart from equilibrium–consequences for soil carbon balance modelling. Biogeosciences 4(1), 125–136 (2007). https://doi.org/10.5194/bg-4-125-2007