Mineral nitrogen depletion and odd GPP diurnal behavior in summer: Artefact or real?

[jan_ivar_korsbakken@cicero_oslo_no]

GPP displays odd bumps during the early morning in summer in many temperate areas, when looking at hourly history output. I am trying to figure out whether it is a real phenomenon or a model artefact. It appears to be related to depletion of soil mineral nitrogen in the model, and there is a plausible physical mechanism to explain it (see below), but colleagues who have been studying Fluxnet data from the relevant areas insist they have never seen signs of such a phenomenon. Any input from either model developers or people familiar with relevant empirical data would be highly welcome.  The main issue can be seen in the attached graph of GPP, from around day 160 (early June) for a patch of forest in southern Norway. GPP has fairly smooth diurnal behavior and a sensible-looking increasing trend up to that point, but gets depressed after that, with weird spikes in the morning.  The depression is caused by downregulation, as INIT_GPP is identical to GPP until the funny behavior starts (see attached figure), but different afterwards. Soil mineral  nitrogen gets depleted at the same time (see graph of SMINN), so this seems consistent. The weird morning bumps in GPP appear to be caused by a small amount of mineral nitrogen that builds up in the soil during the night (see the small nightly wiggles in SMINN), and which supports increased photosynthesis for a few hours after sunrise, but then gets depleted again by mid-morning.  I have checked many forested temperate areas in different parts of the world, and the same thing happens in most of them at the beginning of summer (earlier for lower latitudes, as you would expect). I run an ICRUCLM45BGC case. Total annual global GPP is reasonable (about 120 PgC), so it doesn't appear to be a spin-up issue or anything like that.  Any input welcome, in particular there are two questions:  Does soil mineral nitrogen typically get completely depleted like this in the real world? I am having a surprisingly hard time finding empirical data on this. Most results are about total soil nitrogen, including organic nitrogen, which is not relevant here (I assume?)  Does anyone know empirical data that shows whether the odd diurnal GPP behavior actually happens in the real world (or doesn't)? [/list]  Best regards, Jan Ivar Korsbakken Senior Researcher, CICERO Center for International Climate Research