Transient Traceability Analysis of Land Carbon Storage Dynamics: Procedures and Its Application to Two Forest Ecosystems

Uptake of anthropogenically emitted carbon (C) dioxide by terrestrial ecosystem is critical for determining future climate. However, Earth system models project large uncertainties in future C storage. To help identify sources of uncertainties in model predictions, this study develops a transient traceability framework to trace components of C storage dynamics. Transient C storage (X) can be decomposed into two components, C storage capacity (X_c) and C storage potential (X_p). X_c is the maximum C amount that an ecosystem can potentially store and X_p represents the internal capacity of an ecosystem to equilibrate C input and output for a network of pools. X_c is codetermined by net primary production (NPP) and residence time (τ_N), with the latter being determined by allocation coefficients, transfer coefficients, environmental scalar, and exit rate. X_p is the product of redistribution matrix (τ_ch) and net ecosystem exchange. We applied this framework to two contrasting ecosystems, Duke Forest and Harvard Forest with an ecosystem model. This framework helps identify the mechanisms underlying the responses of carbon cycling in the two forests to climate change. The temporal trajectories of X are similar between the two ecosystems. Using this framework, we found that different mechanisms lead to a similar trajectory between the two ecosystems. This framework has potential to reveal mechanisms behind transient C storage in response to various global change factors. It can also identify sources of uncertainties in predicted transient C storage across models and can therefore be useful for model intercomparison.