Improve magnitud of simulated vs measured soil moisture data fitting

A discussion forum for Hydrus-1D users.
Post Reply
luribar
Posts: 4
Joined: Thu Sep 26, 2019 10:31 pm
Location: Chile

Improve magnitud of simulated vs measured soil moisture data fitting

Post by luribar » Sat Jan 04, 2020 7:17 pm

Hi everyone,

I'm modelling the soil moisture at three different depths using HYDRUS 1D in an avocado orchard in central Chile, i have calibrated different soil parameters and thus improving the simulated vs measured time serie chart but i can not improve the magnitud of the simulated values changing any soil parameters. The output data is sensitive to the increase and decrease water content but fail in the magnitude.

Someone have had this situation?

First_obs_node.JPG
First_obs_node.JPG (86.45 KiB) Viewed 66 times
(I have eliminated the winter season because the simulated/measured data do not fit, we think that the input data could be bad for this period)


Best regards,
Lucas Rivero Iribarne

crhansen
Posts: 25
Joined: Thu Dec 03, 2015 4:21 pm
Location: Canada

Re: Improve magnitud of simulated vs measured soil moisture data fitting

Post by crhansen » Sun Jan 05, 2020 12:25 am

Hi Lucas:

It looks your HYDRUS-1D model is underestimating the soil moisture low and high values. Is that correct ? In any case, it would be helpful if you could provide the following additional information:

- the number and type of soil layers ?
- the time units you specify in the model (seconds, minutes, days, hours) ?
- the soil hydraulic parameters and how you selected these
- what are you using for rainfall and evapotranspiration inputs ?
- what is the bottom boundary condition ?
- what do you have for observed soil water content ?
- are you modelling root water uptake ?

Regards,

Colin Hansen

luribar
Posts: 4
Joined: Thu Sep 26, 2019 10:31 pm
Location: Chile

Re: Improve magnitud of simulated vs measured soil moisture data fitting

Post by luribar » Sun Jan 05, 2020 8:42 pm

Hi Colin,

First to all, thank you for your help.
I will try to answer your questions one by one:

- the number and type of soil layers ?

the soil consist in 6 layers
soil_description.JPG
soil_description.JPG (40.66 KiB) Viewed 58 times
(Mat: soil material, prof.: depth, clay, silt and clay: % of clay, silt and clay, Clase: soil textural clase from USDA textural triangle, Da: bulk density, 2,60,330,1000 and 1500: water content at 2,60,330, 1000 and 1500 kPa.

- the time units you specify in the model (seconds, minutes, days, hours) ?

I'm using days for this simulation

- the soil hydraulic parameters and how you selected these

I used the "Rosseta" module in HYDRUS with the soil description of the above image and then i ran HYDRUS a lot of times looking for improve the simulation changing the soil parameters (inverse solution)

- what are you using for rainfall and evapotranspiration inputs ?

for rainfall i'm using the rainfall data obtained from the nearest meteorogical station to the site (1.7 km)
for evapotranspiration i'm using Eto from the same station that i mentioned for the rainfall and then i use the dual coefficient aproach from FAO 56 to retrieve evaporation and transpiration, for the Ke and Kcb (Ke * ETo = evaporation and Kcb * ETo = transpiration) i use values retrieved from remote sensing data that is the aim of my thesis. I'm using HYDRUS 1D to compare the Ke and Kcb values.

- what is the bottom boundary condition ?

Free drainage

- what do you have for observed soil water content ?

There is a FDR probe installed in the site, which show daily volumetric water content for the whole period

- are you modelling root water uptake ?

yes, for the root distribution i used the description of the soil, which contain the presence of roots in the profile

Again, thank you for your help,
Lucas Rivero Iribarne

Post Reply