Inversely estimating temperature sensitivity of soil carbon decomposition by assimilating a turnover model and long-term field data

Gen Sakurai, Mayuko Jomura, Seiichiro Yonemura, Toshichika Iizumi, Yasuhito Shirato, Masayuki Yokozawa

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Change in temperature sensitivity of soil organic carbon (SOC) decomposition with change in soil qualities (i.e. decomposability or lability) is one of the most important issues to be evaluated for projection of future CO2 emissions from soils. We inversely estimated the temperature sensitivity of SOC decomposition rate by applying a hybrid of the Metropolis-Hasting algorithm and the particle filter method to the extended Rothamsted carbon model (RothC), together with long-term (9 years) experimental data on SOC obtained at five sites in Japanese upland soils. Contrary to the prediction of the Arrhenius kinetics theory, we found no significant differences in temperature sensitivity among soils with different qualities (represented as soil compartments in the RothC model). We also confirmed that there was a positive correlation between the relative temperature sensitivity of the humus compartment and future total CO2 emissions. The RothC model with default parameterization tended to overestimate future total CO2 emissions relative to the calibrated model, and the degree of overestimation was larger than that of underestimation.

Original languageEnglish
Pages (from-to)191-199
Number of pages9
JournalSoil Biology and Biochemistry
Volume46
DOIs
Publication statusPublished - 2012 Mar 1
Externally publishedYes

Keywords

  • Data assimilation
  • Decomposition rate
  • Long-term field experiment
  • Metropolis-Hasting algorithm
  • Particle filter
  • Soil organic carbon
  • Temperature sensitivity
  • Turnover model

ASJC Scopus subject areas

  • Microbiology
  • Soil Science

Fingerprint Dive into the research topics of 'Inversely estimating temperature sensitivity of soil carbon decomposition by assimilating a turnover model and long-term field data'. Together they form a unique fingerprint.

  • Cite this