ModelMuse: Flow Observations

Video Transcript
Download Video
Right-click and save to download

Detailed Description

This video shows how to define flow observations in ModelMuse and how flow observations can be helpful in calibrating a model.


Date Taken:

Length: 00:08:39

Location Taken: US


The letters "USGS" and the words "Science for a Changing World" appear in green and fade to black.
Hello. This video will show how flow observations can be helpful in calibrating a model. 
This model is a profile model. It has 1 row,100 columns, and 20 layers.
There is specified head boundary on the right colored orange.
The specified head is 10.
the blue rectangle defines the recharge which is applied to the top active cells at a rate of 2x10^-11 m/second
Ground water is recharged from the top of the model and leaves through the specified head boundary.
Because it is a closed system in steady state, the amount of water entering through recharge
will equally the amount leaving through the specified head boundary.
Hydraulic conductivity is 1x10^-5 m/second.
I will run the model and then import of the calculated heads.
There are some warnings about some cells going dry but we don't need to worry about those.
Now import the heads.
Here are the cells that went dry, the ones that are not colored.
Now let's increase both the recharge rate and the hydraulic conductivity by factor of 100.
Here is the hydraulic conductivity.
Here's the recharger rate.
We'll run the model again and import the new model results.
We have already imported model results from the first run so I'm going to create a new data sets so we have both sets of results.
The old datasets will have the model results for the original model.
The new datasets will have the heads of the new model.
We do again has some dry heads and the heads look much the same. 
Let's check to see how much difference there is in the calculated heads between the two models.
I'm going to create a new date set.
and I will call it "Delta".
I'll set the formula to be the old heads minus the new heads.
Let's make this 3D.
The datasets are...
These are the old datasets.
These are the new heads.
And we'll color the grid with the difference between the two datasets.
Here it's zero.
So let's see what the actual differences is.
So the maximum difference is 0. The minimum difference is 0.00019.
Not very much.
Let's go back and look at the calculated results.
So we'll look at the the original heads.
And these heads are around 17 going down to 10 here.
In the inactive cells its 1x10^-20 because that's  the value for HDRY, the value applied to dry cells.
We can see that the change in head between the old model and the new model is almost nothing.
Thus we've made a huge change in the model but we can't distinguish between the old and new models using the heads alone.
However, we can use flow observations to distinguish between the two models.
We activate the CHOB package.
That's one of the observation package: the Specified Head Flow Observation package.
Now we define an observation.
It's a CHOB observation.
Set the time to zero
The observed value is the recharge rate 2x10^-9 m/s, 
times the width of the model, 1000 m, times the length of the model, 10,000 m, 
which makes it 2x10^-2 cubic meters per second.
Because the water is leaving the system through the specified head boundary, the sign is negative.
so the observed flow at the constant head boundaries -2x10^-2 m³ per second
This observation will include one object
Which is named "Specified_Head".
We'll set the factor to 1 which means that all the flow through the boundary is part of the observation.
We'll run MODFLOW again.
Close ModelMonitor.
Here's the listing file.
We'll look at the end of the listing file.
So here is the observation. 
Here's the observed value. This is the value we entered. 
And here's the simulated value. They're nearly identical. 
That's the difference between the two, 5x10^-7.
Let's close that.
And we'll close this.
Now we'll go back to our previous values of hydraulic conductivity and recharge and let's see what difference it makes.
So "Data|Edit Data Sets..." hydraulic conductivity...
That was 10^-5
and the recharge rate was 10^-11, 2x10^-11
We'll run the model
Look at the results.
Here it is. The observed value is still the same but the simulated value is now approximately 2x10^-4 instead of 2x10^-2.
So that's made a huge difference in the flow rate through the specified head boundary.
In summary, observed heads alone can be and often are insufficient to properly calibrated model.
Flow observations can provide the extra information needed to help calibrate model.
The letters "USGS" and the words "Science for a Changing World" appear in green.