Improved/fixed sorghum co2 response

[hol430]

Working on #572

@jbrider - with this change we get some co2 response but not much, and we probably will need to reparameterize the two functions I copied from maize. We may not necessarily want to merge this change, I'm mainly interested to see what effect it has on the stats (if any).

I tried to test the sorghum model model from old apsim with variable co2 levels, but it fails because there's no CO2xTE response defined in Sorghum.xml.

[Keith-Pembleton]

have a look at https://doi.org/10.1017/S0021859615001185 for some sorghum parameters. They were developed for forage sorghum in classic but they will at least give a starting point to work from.

[jbrider]

thanks @Keith-Pembleton. We can use this to help test NextGen as well.
There should have been defaults in there that must have been removed at some stage - will need to look back in version control to find them.

[jbrider]

@hol430 You will probably need to bring the CO2 code across from sorghum classic in the first instance as it looks like it works differently to nextgen maize. I think the CO2 response is integrated with MicroClimate which may not work with sorghum yet. I have forwarded a couple of papers that Graeme sent me - which should be a reflection of what is implemented in the sorghum and maize classic code.

The shifting influence of drought and heat stress for crops in northeast Australia - Lobell, David & Hammer, G. & Chenu, Karine & Zheng, Bangyou & McLean, Greg & Chapman, Scott. (2015)

Designing crops for adaptation to the drought and high temperature risks anticipated in future climates. Hammer, G. & McLean, Greg & Oosterom, Erik & Chapman, S. & Zheng, Bangyou & Wu, Alex & Doherty, Al & Jordan, D.. (2020).

[hol430]

Thanks Jason. From what I can see, the only place that co2 has an effect in old sorghum model is in the calculation of RUE

// Plant.cpp:274
float rueToday = (float)(rue * rue_co2_modifier());
...
// Plant.cpp:539
double Plant::rue_co2_modifier(void)                 //!CO2 level (ppm)
   {
   //  Purpose : Calculation of the CO2 modification on rue
   const double scale = 1.0 / 350.0 * 0.05;
   return (scale * this->co2 + 0.95); //Mark Howden, personal communication
   }

This is already implemented in next gen as the RUECO2Function:

// RUECO2Function.cs:67
else if (PhotosyntheticPathway == "C4")
{
    return 0.000143 * MetData.CO2 + 0.95; //Mark Howden, personal communication
}

That's one of the changes I've made in this pull request - I've changed the FCO2 function (previously a constant of 1) in next gen sorghum to use the RUECO2Function with the pathway set to "C4". The other changes I made were because GSMax (max daily stomatal conductance) was always returning 0, due to GSMax350 and FRGR always being 0.

• I set GSMax350 to constant of 0.009, which is what maize does
• I connected the FRGR property to the child function of the same name (previously this property was just returning 0).
• I changed StomatalConductanceMod (previously constant of 1) to use the imlpementation in maize (FCO2/([IWeather].CO2 - [Leaf].CO2internal)/(350 - [Leaf].CO2internal)). This may not have been the right thing to do, as the old sorghum doesn't seem to have this modifier as far as I can tell

[Keith-Pembleton]

@hol430 from memory you are right sorghum in classic was structured different to the other models. We used the forage/sweet sorghum model in the past. I think adding what maize does to sorghum gives us a place to start from. @JJguri, I and others will be working on it over the next few months.

[hol430]

It appears that sorghum's co2 response curve in old apsim was removed in 2012, but even before the parameters were removed from the xml file, there was no response - it was just a flat line. When the parameters were removed from the xml file, this was changed to be an optional parameter, so the model will work with them or without them. Maize in classic is only different in the sense that these parameters are not optional, and it provides some default values.

In new apsim, we had a co2 modifier on RUE but not on TE. I've now added a CO2xTE response, but the question is how do we want to parameterize it. For now I've just copied the response used in the maize model in apsim classic but I'm happy to change this to whatever you guys reckon is best, just let me know. There are quite a few parameterizations in the paper linked by Keith which we could try. If anyone wants to have a play, the CO2TE experiment in the sensibility folder should give a decent graph

[hol430]

Ok, after discussion with @jbrider I have changed the sorghum CO2xTE response to match what old apsim did - ie there is no response, it's a flat line. For those of you wanting to change/experiment with this, it lives at Sorghum.Leaf.PotentialBiomassTEFunction.CO2Modifier.

[hol430]

@hol353, @jbrider I think we can accept the stats on this one - nse for Leaf.Dead.Wt has changed by ~0.01.

[jbrider]

@Keith-Pembleton @JJguri Let us know if you have trouble with setting your own values for it.

[JJguri]

@Keith-Pembleton @JJguri Let us know if you have trouble with setting your own values for it.

Thanks @jbrider, we will be working on it during the next weeks.

[hol430]

@hol353 should be safe to accept stats here (as per above comment, NSE for one sorghum variable has changed by ~0.01).

[hol353]

retest this please jenkins.

Want to rerun this because I merged another PR this morning.