Issues

SOLVED

MAJOR ISSUE: vegetation runs full of labile carbon

The burning of excess labile carbon and reserve carbon is not like in OCN. The increase in autotrophic respiration when labile carbon stocks get unrealistic high is deactivated by DOFOCO team. When using nutrient this is not acceptable, as vegetation runs full of sugar under nutrient stress on allocation. We introduced the burning of excess labile cabron following Zaehle & Friend (2010). It seems to work, but affects NPP which might be problematic. Effect on phenology is not yet checked.

MAJOR ISSUE: the soil and litter CN ratios for spinup_analytic

The running averages of the stoichiometry of the respective pools are reset every spinup cycle to zero. Therefore, the running averages are always too small. According to NV, this doesn't happen in his setup. I modified the calculation and ensured that the variables are written out in the restart file in a clean way. The modifications should ensure that the running averages work for all setups.

UNRESOLVED

Reading in lithology

The nearest neighbour interpolation of the get_lithology routines doesn't recognize missing values. I assigned an average lithology to the grids with a missing values. It would be better if the routine could recognize missing values and remove them from the interpolation.

Reading in N inputs

The variables storing the N inputs in the restart file are all undef. Not quite sure why. I set them all to zero.

MINOR, BUT FUNDAMENTAL ISSUE: Soil type inconsistency in ORCHIDEE-CN-P

The phosphorus model needs soil type specific parameters. Therefore, a new input routine is introduced which reads in USDA soil orders. According to the soil order the parameter values are set for each grid box.

The problem is that ORCHIDEE already reads in fields of soil texture classes which were originally derived from soil orders. According to the soil texture classes, parameter related to soil hydrology are set. As the original data cannot be found anymore, ORCHIDEE will run with two independent sets of soil order specific parameters. Thereby inconsistencies between the phosphorus related parameters and the the hydrology related parameters can be introduced. The practice of reading in soil texture classes instead of soil orders is in general problematic. The ORCHIDEE project members are informed but they seem not to care about this.

MINOR ISSUE: Stoichiometric flexibility

Meyerholdt & Zaehle (2015) showed that model predictions best matched available data when the C:N ratios in all ecosystem pools except the wood pool were treated as flexible. Thus, we should make the CNP ratio of wood fixed; this needs substantial modifications in the code. These modification are at the moment not feasible. I suggest to work on that as as soon as the model works well.

Effect of nutrient stress on root/leaf ratio

The effect of long-term nitrogen stress on the ratio between roots and leaves of O-CN is commented out in the ORCHIDEE-CN. We could reintroduce the effect of nutrient stress on that ratio. However, SL warns of the introducing modifiers on the allocation ratio. DOFOCO does not account for any water stress effect on root/leaf ratio (unlike described in Naudts et al. 2014). Before introducing any effects on the allocation ratios, the way forward has to be clarified and harmonized with NV and SL. For now this is on ice.

heartwood

I experience that the heartwood pool is still accumulating biomass after 500 yr of simulation reaching a stock of appr. 45kg/m2. This is very high compared to estimates of total biomass ranging between 3-15 kg/m2 (Houghton et al. 2005). Keith et al. (2009) shows that forest can have more than 20 kg/m2 aboveground biomass, but PFT averages are close to the values of Houghton et al. (2005).

When using the value for tau_sap of the trunk version of ORCHIDEE (730days) instead of ORCHIDEE-CAN (11680days) the heartwood pool reaches after 250 years 5kg/m2 instead of 30kg/m2. Also all other vegetation pools are substantially reduced, too, and we end up having far too less carbon. An updated set of parameters from NV resulted in intermediated carbon stock. The Krinner et al (2005) shows that ORCHIDEE simulates high total biomass stocks of up to 70+kg/m2 and the biomass. Thus for now, we accept that ORCHIDEE simulates high biomass stocks. In any case, we need more than 250 years to equilibrate the vegetation carbon stocks when using the ORCHIDEE-CAN.

See attached file "Heartwood_3.pdf" for more information