Representing canopy structure dynamics within the LPJ-GUESS dynamic global vegetation model

Several efforts are being pursued to improve the representation of ecological demographic processes that govern terrestrial vegetation canopy structure within Dynamic Global Vegetation Models (DGVM) as it influences critical fluxes of carbon, nutrients, and water. How trees are structured within the canopy determines how they absorb incoming solar radiation and is partitioned between different tree cohorts. Here we present two new schemes with more detailed vegetation canopy structure representation in the DGVM LPJ-GUESS. These new schemes provide a closer linkage to observations to better constrain processes of growth and mortality, improve the representation of species coexistence as well as the capability to represent reestablishment in small canopy gaps following small-scale mortality or selective harvest. LPJ-GUESS is currently structuring its canopy with vertically overlapping cohort crowns without horizontal spatial structure as the crowns are distributed uniformly across the entire patch area. This original approach does not provide a realistic representation of competition between trees of different heights and sizes, nor canopy gaps following mortalities. A first solution to amend the model is to adopt an approach similar to the perfect plasticity approximation in which cohorts are sorted according to tree height and perfectly fill the patch area with each cohort crown area. When the patch is filled an understory layer is created with the next tallest tree and so on for each consecutive layer. A second solution is to explicitly position cohorts within the patch according to forest floor light conditions during establishment. The death of a tree will result in a gap formation which will persist over time and allow new cohorts to establish within the gap exposed to full light conditions. All schemes have been evaluated against aboveground woody biomass, aboveground woody mortality, and aboveground woody productivity split into diameter at breast height size classes and how forestry thinning generates re-establishment of a woody understory. Also, their capability to represent species coexistence has been evaluated. We see improvements in the capability to simulate stand biomass-size distributions, species coexistence, and reestablishment in small canopy gaps with a more detailed canopy structure scheme.