Integrating profile water content over depth

A discussion forum for Hydrus-1D users.
VIVA
Posts: 10
Joined: Tue Feb 28, 2012 1:11 pm
Location: Denmark

Integrating profile water content over depth

Post by VIVA » Wed Jan 16, 2013 9:56 am

Dear all,

Is there an output file where you can see water content integrated over a desired depth?. I have such a dataset (vertical TDR data)and would like to get an output directly from Hydrus to compare with simulations. Any suggestions are welcome. Thank you.

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Mon May 08, 2017 4:03 pm

Although this question has been posted long ago, it interesting to me as well.
I will formulate my current approach and someone who has some remarks on that can post them.

Yes, you can create a subregion for the section so it calculates theta ('WatBalT' in the balance.OUT file).
However, this is limited to the maximum amount of print times (250). So if you work with a time series that is a year you are already in trouble (assuming you want a daily theta value).
Thus, I just equally distribute the max. amount of observation nodes: 10 nodes (f.e. a top soil layer of 11 cm would lead to a observation node every 1 cm) and take an average of those values for 'time' value in the file Obs_Node.OUT.

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Mon May 08, 2017 6:42 pm

Note that the program also calculates and prints into T_level.out the water storage in the entire profile.
Although the GUI support only 10 observation nodes, the computational module supports 100 observation nodes. You can thus modify the Profile.dat (and HYDRUS1D.DAT) input file manually, you will get information for up to 100 observation nodes.

J.

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Tue May 30, 2017 11:25 am

Hi Jirka,

Thanks for your reply that is useful to know!
I actually had three questions that I hoped you could answer.
The first relates to CO2 production in the UNSATCHEM module. That is, the fs(h) = 1 and the (h3) = set at -10^5. I was wondering whether there is a way to calculate the optimum pressure head (h2) from data or should this be gained through optimization? Currently, the CO2 respiration follows the trend of the measurements but its absolute values are not correct.
I imagine that I would need measurements of respiration and water content where temperature is a constant, do you have an example by any chance?
Second, my project is about modelling heterotrophic respiration. In the manual I read that it is assumed that root production takes up 40% under optimal conditions. Just to check: do I simply multiple the cvTop value with a factor of 0.6 to gain the soil respiration output? On page 41 it is mentioned that ‘the values of the optimal production Ypo and Yso must be adjusted accordingly’ in relation to the oxygen deficit reduction factor. Therefore, I wasn’t sure whether Hydrus calculates their ratio as constant over the modelled time series or not.
Third, a colleague of mine wanted to know whether there have been articles published in which Hydrus is coupled with another model using matlab or fortran. I was wondering whether you knew some examples from the top of your head before we dive in the references list.

Regards,

Timo

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Tue May 30, 2017 9:41 pm

1. You would need to have a specialized CO2 production experiment where you would measure CO2 production at different levels of water content. If you have transient field data, then you need to optimize this parameter.
2. Optimal CO2 production rates by plants and soil are input parameters. It is really on the users what they specify. In the examples that I provided (e.g., Buyan81) we assumed that the total optimal CO2 production is divided between the soil and plant respirations with a ratio of 40:60. This is however true only for the optimal production. Actual CO2 productions by plants and soil may have a different ratio. Once CO2 is produced in the soil profile, we do not distinguish the sources any longer. I'm not sure what you refer to on page 41.
3. There are many publications like this.
J.

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Thu Jun 15, 2017 5:07 pm

Ok thanks Jirka. I was thinking that the ratio soil:plant respiration might be variable over time but I understood later that these are described indirectly by either diurnal/seasonal fluxes of other reduction factors.

Anyway, as it is not as warm here in the UK as in the Californian climate I would like to modify the UNSC.exe source code for the reference temperature (for optimal production) within the CO2 module.
As I will not modify this myself I was wondering whether;
- The link to the source code download includes the UNSC.exe file.
- What FORTRAN compiler is needed to do this (I understood there are MS, Digital and Intel compiler types?)
- It would be perfect if perhaps you know which line this is on
- Then I had one strange finding. When I use the meteo.out values for E & T, convert them to cm and use them to run the CO2 module, theta and temperature are equal but the pressure head values are completely different (by a factor of 4x, for instance in a very dry summer -80.0000 (h) becomes -20.000 (h) using the CO2 module. In both cases there are 10 observation nodes placed in the top 10 cm. Do you know why this could be?

Regards,

Timo

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Thu Jun 15, 2017 7:20 pm

I do not share the UnsatChem code with the public.
If the theta is the same, then the pressure heads must be the same as well. It is possible that there is the same theta (at least down to 3 digits, which are printed) for h=-20,000 or -80,000 (cm or m?). You may be using different limiting pressure heads (hCritA). J.

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Fri Jun 16, 2017 3:34 pm

Hi Jirka,

Ok sorry to hear, although I had a discussion and think I understand so with the correct Ea value you can still have a coefficient for the underlying equation that is higher than 1 for atmospheric temperatures below the reference temperature of 20 o C? The only difference would be perhaps that I can estimate the Ea value with more certainty as I would have more data points that lie around the annual mean temperature I guess..
The pressure heads are in cm and hCritA is equal, I see now that there is a slight deviation in water content but the deviation in pressure heads is quite larger as you can see. (orange/yellow: unsatchem, blue/grey: hydrus) .Unfortunately I cannot upload the zipped hydrus projects here as they exceed the maximum limit. Do you have an idea why this could be?
Attachments
16-06.png
16-06.png (40 KiB) Viewed 17632 times

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Sat Jun 17, 2017 2:48 am

The temperature dependence of CO2 production uses an Arhenius equation, which needs to have a rate at a reference temperature (it is the same as pretty much in every chemical/thermodynamical database - e.g., PHREEQC, WATEQ, Minteq, etc; ditto for pesticide degradation rates). One needs to specify a reference rate at a reference temperature. One then uses the Arhenius equation (and the activation energy of the reaction) to get rates for any other temperature. The rate can then be either higher or lower than the reference rate depending on temperature. That's a standard practice of dealing with temperature dependence. J.

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Fri Jun 23, 2017 3:16 pm

Hi Jirka ok got it. I have calculated the activation energy using 293.15 K as a reference temperature, it turns out rather low (around 3800) compared to the default values, but that might be due then to lower average temperatures?
I was wondering whether you could provide me with some tips on how I could convert the respiration output of Hydrus M3 m2 day-1 to mg C/m2/day. I have been reading a lot on it, but I did not understand whether the respiration is given in reference to air or not. That is, if I could convert it to ppm I could simply adjust the ideal gas law to convert it to g C / cm3 and from there on it will be easier.
Following on that, should I take into account as well that the modelled area is on a hillside (i.e. the weather station gives values around 760 mbar compared around 1013 stp)?

I also wondered whether you would have an idea on why this pressure head value difference can occur between unsatchem and hydrus?
Looking forward to your reply

Timo

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Wed Jun 28, 2017 2:17 pm

I’m currently traveling in Peru and will not be answering discussion forum questions until I come back to US. JS

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Mon Jul 03, 2017 2:27 pm

Ok, thanks. Looking forward to your reply when you have the chance.
Enjoy your travels!

Timo

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Mon Jul 03, 2017 8:26 pm

Activation energy in the Arrhenius equation is a single value that is independent on temperature.

Conversions: You need to look closely on units that are used in UnsatChem, and units into which you want to do conversions. The CO2 concentrations in UnsatChem are expressed on a volumetrical basis, i.e., [cm^3 of CO2 per cm^3 of air]. The CO2 production rate is then correspondingly expressed in units of [cm^3 of CO2 per cm^3 of air per unit of time].

The conversion of concentration units from [cm3/cm3] to partial pressure in atm may be done as follows (using ideal gas law):
R=8.314 ! universal gas constant (J/K/mol)
Temp=273.15 ! temperature (K) 0C
V=0.022414 ! Volume (m3)
P=101325 ! atmospheric pressure (Pa)
pCO2=CO2(i)*R*Temp/V/P

J.

TimoSamuel
Posts: 11
Joined: Mon Feb 27, 2017 1:48 pm
Location: United Kingdom

Re: Integrating profile water content over depth

Post by TimoSamuel » Sat Jul 08, 2017 1:02 am

Hi Jirka,
Thanks again so much for your reply. I have one of my last questions on this project which I hoped you could help me with.

Within Both the overall fit and the difference between earlier years and later years can possibly be explained by either an increase in the Michaelis’ constant, or in the Y¬so parameter thus, a basis in the rate of oxygen uptake or co2 production. (And thus a higher respiratory quotient). Particularly the MC parameter appears relevant as the min and max values are maintained as compared to an overall increase in the time series with higher Yso parameters.

However, from the manual, I do not understand what it means that “it is assumed that the respiratory quotient is equal to unity”.
The soil modelled is a peat type soil with 0.89 θs and 73% OM content in the top soil layer. It is a hydric system that shows high saturation (even water-logged) conditions in winter.

Thus my question is; a higher M constant of 0.21 or 0.22 g physically speaking incorrect due to 21% oxygen composition in soil air, however, how does the representation for a higher RQ in water-logged conditions relate to this? I ask this, as the CO2 production strongly increases up till a MC value of 0.27 whereas it should actually level off according to the last plot.

Also is there a way to solve the equation 5.25 from the Unsatchem manual so I can check where the current RQ lies in the michaelis-menten plot, however O2 concentration is not an output is that correct?

Then, a quite general question: Could you elaborate or give a view on how I should interpret the interaction between optimal production and oxygen limitation within the unsatchem module as the sensitivity analysis I have done assumes steady state of other variables?

Regards,

Timo
Attachments
sensitivity co2 hydrus.pptx
(572.55 KiB) Downloaded 248 times

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Integrating profile water content over depth

Post by Jirka » Sun Jul 09, 2017 12:10 am

Equation 5.22 gives the dependence of CO2 production on O2 deficiency. Since the code does not know what the O2 concentration is it needs to make some assumption about it, such as O2=0.21-CO2 and that O2 consumed is replace with CO2 produced (the respiratory quotient = 1). Equation 5.23 is then expressed in concentration of CO2, with the assumption that the respiratory quotient (the ratio of carbon dioxide produced to oxygen consumed) is equal to 1.

There is no equation 5.25 in the manual and thus I'm not sure what you are referring to. However, I assume to eq. 5.22 and 5.23, which I discussed above.

J.

Post Reply