Steady state of spinup

[luyssaert]

In revision r7040 the vegetation biomass reaches a steady state after around 600 years. This is seems to be quite a lot given that we don't expect many forests to grow older than 200 to 300 years.

Check: whether this long time period to reach steady state can be ascribed to a single of few PFTs. If so check the self-thinning parameters of these PFTs.

In revision r7040 the soil has not yet reached equilibrium after 600 years resulting in an NBP of 0.309 Pg C/yr. This is more than 3 times higher than the equilibrium NBP in tag 2.2.

Check: the soil is expected to be running behind the vegetation. Could be that soil reaches equilibrium a few decades later. In r7040 grasses need to be replanted too often (this hints at a bug -> #728). Crops need to be replanted annually (by conception -> not a bug). This means that we keep adding nitrogen to the soil and thus change the soil fertility. This could result in some transient behavior (although a steady state should occur after some time but we don't know how long "some" is).

[luyssaert]

The biomass of PFT8 increases for almost 600 years after which it saturates. PFT8 contributes considerably to the global biomass (see attached figure, Y-axis is in Pg C). The biomass of PFT9 is still linearly increasing after 650 years but its global contribution is much smaller (see attached figure, Y-axis is in Pg C). If we could speed up the period in which PFT8 could reach a steady state biomass, PFT2 would become the bottleneck because its biomass increases (slowly) between 300 and 650 years (see figure). Given that PFT2 dominates the global biomass, a relatively small increase could still be seen in the NBP.

[mmcgrath]

An extended spinup to 1000 years has been finished for r7013.

​https://vesg.ipsl.upmc.fr/thredds/fileServer/work/mcgrathm/OL2/PROD/spinup/FG1.FM1.r7013/MONITORING/index.html#tabs-2

The NBP is not going to zero, instead staying positive, even when the average is taken over 300 years.

yes? list NBP_FLUX_C_LANDS[l=700:1000@ave]
             T        : 699.5 to 1000.5 (averaged)
          0.4065

yes? list NBP_FLUX_C_LANDS[l=800:1000@ave]
             T        : 799.5 to 1000.5 (averaged)
          0.4183

yes? list NBP_FLUX_C_LANDS[l=900:1000@ave]
             T        : 899.5 to 1000.5 (averaged)
          0.3957

[luyssaert]

Committed code to reduce synchronicity during the spinup in r7143.

[luyssaert]

With r7278 the NBP is stable and of the same order of magnitude albeit of different sign as ORC2 and ORC3. ORC4 simulates as ink of 0.136 Pg C y-1. Following parameterization of sugar_loading and autothropic respiration, the labile+reserve carbon and biomass reached equilibrium (reflected in the stable NBP). The remaining sink is thought to be due to the fact that the N-cycle is not completely closed and that new N enters the system when replanting occurs. For crops this is annually. For semi-arid and arid grasslands this annually or once every couple of years. A possible solution is proposed in ticket #793.