Effects of monthly versus annual mean temperature data for modelling soil organic matter decomposition
The comparatively low seasonal temperature range of maritime climates is one factor that causes relatively low annual soil organic matter decomposition rates, compared to regions experiencing higher seasonal temperature range.
Global models generally agree that warming will lead to a loss of soil organic C, while increasing CO2 is expected to stimulate plant net primary productivity and increase C stocks. However, predictions of the combined effect of increased CO2 and climate change vary greatly between models.
This is aptly illustrated in a study by Sitch et al. (2008) who ran five global models to the end of the 21st century with the same projected climatic changes as key input (Fig. 1). The models disagreed considerably on the magnitude and even the direction of net changes in soil C. The magnitude of the disagreement amounted to over 200 GtC (billion tons of C) between the most extreme simulations. That extent of uncertainty is greater than all the C released globally from all land-use changes to date.
This large disagreement is most disconcerting as it has a direct bearing on the urgency with which climate change mitigation has to be addressed. If the more benign simulations are correct then we have more time and flexibility in responding to climate change, whereas the more pessimistic simulations would call for much more urgent and drastic mitigation responses.
A detailed analysis of the simulations revealed that much of the differences between simulations critically depended on the temperature sensitivity of organic matter decomposition. Much of our understanding of such sensitivities is based on short-term studies that generally show a strong temperature sensitivity of organic matter decomposition. However, many climate change models use longer (annual) time steps. Dr Miko Kirschbaum has investigated the extent to which short-term temperature response functions can be used for long-term global applications, by determining how predictions of changes in soil C stocks are affected by the inclusion or omission of inter-annual temperature variations.
Results indicate that for global applications, it is critical to consider seasonal temperature variations. Short-term measures of temperature dependence cannot be applied at different temporal scales without explicitly considering the variability of temperature over the longer temporal scale. Using short time steps of at least months in simulations results in greater annual decomposition activity, especially for cold regions, than would be predicted by simulations that use only annual time steps (Fig. 2).
So what does it all mean? The nature of feedbacks from soil carbon storage strongly affects the extent of ultimate climate change caused by a given amount of fossil-fuel emissions. Inclusion of the seasonal temperature cycle somewhat reduces the concerns that global warming could lead to significant additional losses of soil carbon stocks, especially in cold regions. It is thus one bit of good news that somewhat allays concerns that human fossil-fuel emissions could be compounded by natural feedback processes.
More generally, the work showed that the temperature dependence of organic matter decomposition cannot be quantified without specifying the spatial and temporal scale over which it is used. There is a wealth of observations at small and short scales. Long-term and global responses can be derived from these short-scale observations, but only with further analyses to understand and incorporate the factors that operate and interact at these various scales of interest.
New Zealand’s generally more maritime climate, and consequent low seasonal temperature range, is one factor that causes low annual decomposition rates. This, in turn, leads to higher soil C stocks than for regions with similar mean temperature but higher seasonal temperature range.
References
Kirschbaum MUF 2010. Global Change Biology 16: 2117–2119
Sitch S et al. Global Change Biology 14: 2015–2039.
Miko Kirschbaum
Figure 1. Five dynamic global vegetation models (as specified in the Figure) run under the SRES A1F1 climate change scenario, which includes changes in CO2 concentration as well as climatic driver.
Figure 2. The effect of using annual mean air temperature or monthly mean air temperature to estimate annual decomposition activity for systems with different annual mean temperatures and temperature ranges. All data are expressed as a percentage of the decomposition activity calculated from daily soil temperatures.
