T_Level Output Discrepancy

A discussion forum for Hydrus-1D users.
Post Reply
dg34
Posts: 5
Joined: Wed Nov 28, 2018 6:21 pm
Location: Chile

T_Level Output Discrepancy

Post by dg34 » Thu Jan 17, 2019 6:08 pm

I am using HYDRUS to look at infiltration of meteoric waters (precipitation) into the wall of a tailings dam. Thus, I am primarily interested in net infiltration. My model simulation is for just over three years, with hourly data for precipitation and evaporation calculated by HYDRUS from solar radiation data. I am also comparing the simulation of the bare wall with various potential cover scenarios.

I understand that for cumlative net infiltration, sum(vTop) is the output of interest. However, I originally calculated this myself from the difference between sum(Infil) and sum(Evap). In the case of the bare wall, this difference and the value of sum(vTop) are the same. However, for the cover scenarios there are small (~2%) differences between the two. The water balance errors for each of the simulations are good (<<1%). My assumption is that this discrepancy is solely due to small errors in how these values are calculated, but I also intuitively think that they should be the same.

Am I overthinking this, and correct to assume its just calculation error? I am unsure how/why this discrepancy exists.

Any clarification is greatly appreciated.

Thanks,

Dan

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

Re: T_Level Output Discrepancy

Post by Jirka » Thu Jan 17, 2019 7:23 pm

The sum(Infil) and sum(Evap) values are post-processing values (i.e., calculated independently of the actual solution of the Richards equation; and thus do not affect the actual solution). Values that come directly (and thus are more important) from this solution are vTop, vBot, and vRoot (and thus sum(vTop), sum(vBot), sum(vRoot)) and reported mass balance errors. The Infil and Evap (and thus sum(Infil) and sum(Evap)) values are calculated after each time step from the rTop, vTop, Prec, and rSoil (also surface layer if considered) values and thus should correspond with sum(vTop). However, as I said above, it is a post-processing process, and it is entirely possible that I may have missed some of the many possible combination of rTop, vTop, Prec, rSoil, WLayer.

J.

Post Reply