Disordered kinetics module for SOC decomposition

[AlexisRenchon]

Is your feature request related to a problem? Please describe. In a transient climate, ESMs need to predict the fraction of CO2 emitted by human activity that will remain in the atmosphere. To date, about 50% of CO2 emissions stays in the atmosphere, the rest is absorbed by ocean and land sink (about 25% each). On land, the pools of carbon can be divided into live and dead organic mass (e.g., vegetation and soil organic matter). ESMs, via LSMs, need to predict the future of these pools in that transient climate (e.g., how increased T and shifts in moisture and precip will modify those pools).

Describe the solution you'd like There is no clear consensus in the best approach to model the fate of soil organic carbon (it is also very difficult to validate as there are no global data of change in SOC). Most LSMs use a model called "CENTURY", which partition SOC into a finite number of pools, e.g., 3 pools, "active", "slow", "passive", where each pool decay at a rate k, k being big or small depending on which pool. Another approach, that we implemented in CliMA, is to a more mechanistic approach with only one pool, were soil C gets decomposed at a rate function of temperature, moisture, and oxygen content in the porous air space. (these variables varying with depth in the discretized soil domain).

Describe alternatives you've considered The new approach of this SDI would be to implement a "disordered kinetics" approach, where an infinite number of soil C pools gets decomposed at different decay rate k, where k follows a probability density function. In this approach, k is determined by the age of organic matter and, in space, environmental variables such as mean temperature or moisture.

Additional context Advantages of disordered kinetics:

• elegant mathematically
• easy to parameterize
• can make use of global datasets of live C turnover and soil C age
• is built to track pool change at larger time scale rather than fluxes on short time scale

Disadvantages of disordered kinetics:

• works well in a steady state world, might not be ideal for a transient climate
• does not partition decomposition of soil C into CO2 and CH4, which is a problem as these 2 molecule have a very different radiative forcing that matters for climate
• require space for time substitution that may or may not be appropriate
• may not work well on short time scale, making it impossible to benchmark at hourly, diurnal and seasonal time scale