Changes between Version 105 and Version 106 of Documentation/TrunkFunctionality4

09/06/21 20:34:59 (10 months ago)



  • Documentation/TrunkFunctionality4

    v105 v106  
    422422Temperature is a very consistent climatological variable in the sense that each year the summer is warmer than the winter (even in so-called cold years). This implies that temperature driven phenology is very robust in the sense that the PFT will develop a canopy each year. Soil moisture, however, is much less predictable and in very dry or very yet years the seasonal patterns may differ a lot. Especially when soil moisture criteria are combined with other criteria (pheno model: moigdd) there are years that the criteria are never met and that the PFT never develops leaves. This is not at all observed. Even farmers would plant crops when the conditions are far from optimal (typically because they have to plant the crops before they know that the conditions will be far from optimal). If set the yes, the flag'''ok_force_pheno''' will start the growth of a canopy a predefined number of days (force_pheno_mtc) after the average budbreak data for that PFT at that location. Test simulations over the Americas showed that over a 30 year period forced phenology was used 5 to 10% of the years between 1901 and 1930 for PFT10, 12 and 13. This flag did little or nothing to the phenology of the other PFTs with the pheno model moigdd. The variable qc_pheno_event registers if budbreak happened during the year. The variable qc_forced_pheno_event registers if the phenology event had to be forced rather then being simulated as a function of the climatological variables. The ratio of these output variables can be used to plot in which regions and years phenology had to be forced. Not surprisingly it is most used in dry regions of the PFT distribution. 
     424=== Pgap === 
     425the Pgap calculation in r7278 was checked line by line by three persons. No obvious problems were found. Why was it checked in the first place? Pgap based veget is higher than Lambert-Beer used veget. Given the differences in definitions/approach this is not really a suprise but it seemed worth to check. This line by line check revealed two possible issues: (a) The Pgap that goes into the calculation of veget assumes zenith solar angle. This might be problematic for radiation-related uses of Pgap. Should we introduce two vegets? one for zenith angle to be used for interception and one for the real solar angle to be used for shading and temperature related processes? (b) It needs to be checked how overlap is dealt with in Pgap. The approach is very insensitive to the canopy porosity this might be cause by the fact that overlap between canopies in not properly accounted for (thus the cumulative crown volume may exceed the entire space of the layer for which we calculate the crown volume). A ticket was made for these issues.