Dear forum,

I am trying to simulate root water uptake by Kernza (Intermediate Wheatgrass) using Hydrus 1D. We have TDR sensors installed at 75, 150 and 250 cm depth. We see changes in the volumentric water content at 75cm but not at 150 and 250. By matching the model output to our TDR data at 75cm depth we hope to simulate what happens between 75 and 150 cm and what the maximum depth of water uptake is. We will also use the model to construct a water balance.

I'm running into some things that I don't understand.

1) I have divided the 3m soil profile into 4 materials of which we know the composition, bulk density, TH33 and TH1500. I used the neural network prediciton to predict the water flow parameters. I also have selected hysteresis in retention curve and initially drying curve. I know AlphaW=2*AlphaD. Are there similar approximations for Qm, QsW, or can they be calculated? And can KsW be regarded as equal to KsD?

2) I have a pre-calculated ET0 from meteo data which I want to use as input. Do I have to multiply this ET0 with a crop coefficient before putting it in the boundary condition window, or is using the LAI as input adequate by itself?

3) Am I right to say that solving Hr = exp[h*M*g/RT] for h using daily Hr values gives me daily hCritA values?

4) Since no Feddes parameters are known for Intermediate wheatgrass, should I use the parameters for small grains (eventhough Kernza is a perennial)? For small grains different parameters exist for the vegetative and generative period, how then do i simulate (and get output graphs for) a whole growing season?

5) How do I determine the extinction coefficient or is it always 0.463?

Thanks in advance,

Joost

## Root water uptake Kernza

### Re: Root water uptake Kernza

You have too many questions, which means that your model is too complex. You should start from a simple model and only gradually add complexity.

1. Do not consider hysteresis until you successfully run the model without hysteresis.

2. There are different opinion about this. The crop coefficient partly does what HYDRUS does as well, i.e., reduces potential ET into actual ET. Therefore I prefer not to use the crop coefficient, while others would recommend to use it.

3. Yes.

4. HYDRUS allows only a single set of Feddes parameters. If you want to consider two sets (e.g., for vegetative and generative periods), you need to split the simulation into two and import results of the first simulation as the initial condition for the second simulation.

5. I have no opinion on this.

I'm currently traveling in Tanzania (climbing Kilimanjaro and safari) and thus do not have time to answer in more detail.

J.

1. Do not consider hysteresis until you successfully run the model without hysteresis.

2. There are different opinion about this. The crop coefficient partly does what HYDRUS does as well, i.e., reduces potential ET into actual ET. Therefore I prefer not to use the crop coefficient, while others would recommend to use it.

3. Yes.

4. HYDRUS allows only a single set of Feddes parameters. If you want to consider two sets (e.g., for vegetative and generative periods), you need to split the simulation into two and import results of the first simulation as the initial condition for the second simulation.

5. I have no opinion on this.

I'm currently traveling in Tanzania (climbing Kilimanjaro and safari) and thus do not have time to answer in more detail.

J.