Pervasive shifts in forest vegetation dynamics are already happening and are expected to accelerate under future global changes (McDowell et al., 2020). Forest dynamics are changing due to anthropogenic-driven drivers and transient disturbances. Drought-induced mortality is considered one major cause of the recent widespread tree mortality events, which profoundly damage ecosystems and many ecological processes (Anderegg et al., 2015). Among carbon resources, non-structural carbon (NSC) is commonly considered a repository depending on the balance between the supply of assimilated carbon and carbon demand. Hence, the size of stored C pools can be considered an indicator of the carbon balance of the plant. Additionally, NSC could be the threshold of the conceptual “carbon starvation” as one of the mechanisms after drought (McDowell et al., 2008). However, no ecosystem model can calculate the NSC dynamics and the effect of carbon starvation clearly preventing from understanding future carbon cycling. The objectives of the research are 1) to apply the process representing NSC dynamics to the process-based Spatially Explicit Individual-Based Dynamic Global Vegetation Model (SEIB-DGVM; Sato et al., 2007), 2) to simulate the effect of carbon starvation on forest dynamics, carbon cycling by using the new process.