Sensitivity of PFT definitions to carbon cycle simulations in DGVMs

Background/Question/Methods

Earth System Models (ESMs) represent vegetation by grouping them into 10-20 plant functional types (PFTs) based on a phenomenological species classification of characteristics including photosynthetic pathway, leaf phenology and their growth locations. An important question to ask, however, is whether these PFTs are adequate for the simulation of global carbon cycle under current and future climate conditions. To address this question, we design three different scenarios to assess the effects of PFT definitions on carbon cycle simulations by comparing the model predictions based on one general PFT and two PFTs with very different physiological components. Specifically, we are comparing the model performances to predict vegetation responses to drought, warming, and carbon allocation with different representations of drought tolerances, nitrogen use strategies and carbon allocations to roots among PFTs, respectively.

Results/Conclusions

We expect that the single-PFT representation is able to be adjusted to simulate the same magnitude of the carbon cycles as in the two-PFT representation; however, it will fail to simulate the vegetation responses under a changing climate condition. In summary, we conclude that it may be adequate to simulate the vegetation dynamic under current climate conditions based on phenomenological classification through parameter adjustment. To correctly simulate the vegetation response under future climatic conditions, it is important to have more advanced definitions of PFTs that consider key physiological components that are relevant under future climatic change.