Variable number of soil layers for each HRU

[LausitzModeler]

Hello everybody,

i am quite new in Raven. Our research team would like to build the raven model for a catchment, which is already modeled with another watershed model. So with the goal of model intercomparison.

My first question adresses the number of soil layers:
Our catchment is divided into subbasins and HRU's. And each HRU has a specific number of soil layers with variable depth. Is it possible in Raven that the HRU's have a different number of soil layers (eg. 3-11) ? And how is it implemented in the raven-files?

And my second question: Has Raven the ability to model preferential flow along soil macropores from top soil directly to groundwater? In the manual I found "split process". We want to set a volumetric percentage of preferential flow for each HRU and soil layer (or it's calculated from soil layer attributes).

Thanks!
LausitzModeler

[rchlumsk]

Hi there

Welcome to using Raven, hope that you have a good experience in using it.

To the first question, it is possible to setup HRUs with different numbers of soil layers, but you will need to setup different soil profiles (:SoilProfiles in the rvp file) that define the depth of each soil layer, and then define the soil profile used in each HRU. If you look at the Raven Tutorial Files (downloadable from the http://raven.uwaterloo.ca/Downloads.html webpage) at the Nith model, you will see an example of multiple soil profiles setup in the model, and how they are specified in the corresponding rvh file. You would simply have a variable number of layers in different profiles.

On the second question, yes and there are a few ways to do that. For example, as you suggested you could use the split process to move a percentage of water from the top soil layer to a deeper soil layer or surface water directly, e.g. moving 10% of the top soil layer to the third soil layer, bypassing the second entirely.

Code: Select all

:Split RAVEN_DEFAULT SOIL[0] SOIL[2] 0.1

Another option would be to add a special percolation process, perhaps making it conditional on a certain HRU type (if the preferential flow only occurs in some HRU types). The flush command could also be used to move all water from the topsoil to a deeper soil layer if the macropore flow is dominant. See the Template Files (Appendix D) in the Raven manual for some examples of using these commands.

Hope that all makes sense and is helpful. Please respond here if you have other questions to discuss.

Cheers,
Rob

[LausitzModeler]

Hi Rob,

thanks for your response. It helped me a lot!

Meanwhile, I have an other question regarding the soil parameter porosity. In the manual the parameter is declared as "porosity" and in the appendix of the manual (soil parameter table) as "effectiv porosity". When I use the equations on page 30 only the total porosity makes sense. Here my example calculation for one soil layer with thickness of 2 m, total porosity 42 Vol%, field capacity 33 Vol% and wilting point 17 Vol%:

theta_max = 2000 mm * 0.42 = 840 mm
theta_fc = 2000 mm *0.33 = 660 mm
theta_pwp = 2000 mm * 0.17 = 340 mm
S_fc = theta_fc / theta_max = 0.79
S_wilt = theta_pwp/theta_max = 0.4
theta_tense = theta_fc - theta_pwp = 320 mm

So is it right to use total porosity in the rvp-file? I would be very happy to receive an answer!
Thanks!

LausitzModeler

[rchlumsk]

Hi there

That's a good question with a bit of nuance. The porosity provided is corrected by the stone fraction, so if a stone fraction is specified, the porosity provided will be corrected by the stone fraction (SF) to get an effective porosity and correct for unusable space in the soil (i.e. equation 3.2 in Section 3.2 Infiltration on page 30 of the manual). If the stone fraction is not specified, it is autocalculated as zero, essentially implying that total porosity and effective porosity are equal.

In short, you can specify the porosity as total porosity and also specify a stone fraction to correct for effective porosity, or you can specify the parameter POROSITY as the effective porosity and skip specifying the stone fraction.

Hope that helps, let me know if you have any other questions!

Cheers,
Rob

[LausitzModeler]

Hi Rob,

Thank you for the quick response! I am afraid I need a little bit more clarification. Sorry.

I think there might be different definitions for total porosity and effective porosity in different languages. For context: I am working with an idealized soil, where there are no dead end pores and no stones present in my profile.

In german total porosity is defined as basically all pores, that are present in a soil. Which also means that the value of total porosity is always bigger than the field capacity of a soil.

Effective porosity under given circumstances is however defined as the porosity where water can be moved by gravitational forces and is available for vegetation. Water bound by molecular and surface-tension forces (immobile water) is excluded, because it cannot take part in any hydrologic processes.

Especially for clay soils the value of total porosity and effective porosity can be vastly differnet. Which one I use has therefore a huge impact on hydrologic processes like infiltration.

I do not know the inner workings of Raven, so I cannot tell if there are any functions that reduce total porosity by the fraction of immobile water automatically or not, because there is the posibility of autogenerating a lot of soil parameters.
So I don´t know which one of the peviously defined porosities I should use for the parameter POROSITY in my model.


Thanks a lot!
LausitzModeler

[jrcraig]

You can think of POROSITY in Raven as total porosity.

Effective porosity in raven is addressed in a few (but not all) algorithms via the use of a wilting point saturation, below which water cannot be removed via gravity or transpiration.

In most cases, however, porosity and effective porosity are equivalent because there is no minimum saturation (greater than zero) considered in the algorithms for infiltration or evapotranspiration.