Interception of two models

A discussion forum for Hydrus-1D users.
Post Reply
M_Vontobel
Posts: 1
Joined: Fri May 08, 2020 9:13 pm
Location: Switzerland

Interception of two models

Post by M_Vontobel » Fri May 15, 2020 10:33 am

Dear Hydrus-1D team,

I am currently modeling groundwater recharge using two Hydrus-1D models, which have the same atmospheric boundary conditions but vary largely in terms of interception. I hope someone can help me with this issue as I cannot find an explanation.

For both models, in the window "Water Flow Boundary Conditions" I ticked "Input PET and LAI", chose an extinction coef. of 0.39 and also ticked "Interception" and set the value for a = 0.25.
As for PET, I provide both Hydrus models with the same FAO reference grass evapotranspiration (ET0) and LAI which is based on MODIS remote sensing data.
However, I have to mention that my dataset is fairly large so that I cannot enter the values for time, precipitation, PET, LAI, hcrit, GWL and root depth via the "Variable Boundary Conditions" window in Hydrus.
I have to close Hydrus, enter all values manually in the ATMOSPH text file and then start Hydrus again to execute it.
The Hydrus output I attached here:
The bold lines all refer to the model for which I also conducted a soil moisture calibration. As a result the soil moisture in the topmost layers is different compared to the other model. The topmost layers are much more moist which explains the higher cum. root water uptake (light green, bold) – but this should not have an impact on interception because interception occurs before water infiltrates and is available for the roots, or for evaporation or for percolation.
Interception I computed as cum. precipitation minus cum. infiltration. The "spikes" in the cum interception (dark green, bold) can then be explained by ponding that occurs due to delayed infiltration.

But why is interception almost equal to zero for the calibrated model? For the uncalibrated model interception is about 330 mm (dark green, thin line).

Many thanks for your inputs,
Martin

Post Reply