Pick a scenario; every slider stays editable.
Green dots fall into the canopy (photosynthesis, GPP); orange dots rise from trees and soil (respiration). Their number follows the monthly fluxes.
Stack: soil (dark brown), litter, roots, wood and foliage, harvested products (grey). White dashed line: total stock of the undisturbed stand. Yellow line: the age shown in 3D.
The rotation sweep clear-cuts the current forest type on six cycles and compares yield and mean carbon stock with the mean-annual-increment theory. The warming sweep compares old stands at six temperatures.
Running...
The science
Budget. Carbon moves between six pools (foliage F, wood W, roots R, litter L, soil S, harvested wood products P), in g C m-2, with a monthly time step. Net ecosystem exchange is NEE = Ra + Rh - GPP (negative = the forest takes up carbon). Net primary production is NPP = GPP - Ra.
Photosynthesis. A light-use-efficiency model: GPP = ε fT fW PAR (1 - e-k LAI), with LAI = SLA · F and an extinction coefficient k of about 0.5. Deciduous trees have leaves only when the temperature is above about 5 C; the seasonal temperature swing is ±10 C and light follows a sine over the year. ε and the light level are tuned so the temperate stand gets GPP of about 12-14 t C ha-1 yr-1; that is a plausible order of magnitude, not a measurement of any real site.
Respiration. Ra = Rm + Rg: maintenance respiration Rm = Q10(T-10)/10 (rF F + rR R + rW Wsap) on living tissue (sapwood grows only as W0.7), and growth respiration is 25 % of what is left of GPP. As the wood pool grows, maintenance eats a larger share of GPP, so NPP/GPP falls with stand age (here from about 0.58 to 0.46). Published values of NPP/GPP cluster near 0.5 and are uncertain.
Allocation and turnover. NPP is split between foliage, roots and wood with fixed fractions. Foliage lasts 1 year (deciduous) or 3 years (conifer), fine roots turn over at 0.6 per year, wood mortality is 0.3 % per year plus a density term. Dead material goes to litter; litter decays at kL = 0.35 Q10(T-10)/10 per year (20 % of what decays is humified into the soil pool), and soil carbon decays at 0.012 per year times the same temperature factor. Both are scaled down by drought. Heterotrophic respiration Rh is the rest of the decomposed carbon.
Disturbance. A clear-cut removes the chosen share of wood: 45 % becomes long-lived products (35-year mean life) and the rest is emitted at once; slash, stumps and foliage go to the litter pool and the stand is replanted. A wildfire emits 30 % of foliage and litter and 8 % of wood and kills half of the trees. Thinning takes 25 % of the wood. These fractions are illustrative.
Rotation length. The mean annual increment of wood, W(t)/t, peaks where dW/dt = W/t; a clear-cut at that age gives the highest sustained yield. Longer rotations store more carbon in the forest but export less. The sweep compares the numerical optimum with that tangent condition.
The picture. Tree numbers and heights are illustrative allometry: stem density falls from about 2000 to 400 per hectare with age, and height follows 3.0 · B0.36 for a tree wood carbon B in kg. The 40 m × 40 m plot is a sample of the stand.
Honest simplifications. One stand, no species mix, no nutrient limitation, no CO2 fertilisation, no snow, no wood-product substitution benefits or leakage. Only wood products are tracked after harvest. Constants such as Q10, humification, product life and respiration rates are typical-looking placeholders; the soil response in particular is very sensitive to them. This is a teaching model, not a carbon-accounting tool.