Pastas to back-calculate and predict other hydrological variables #670
Replies: 2 comments 2 replies
-
Hi @LucaPiciullo, I'm not entirely sure what you mean by back-calculate, but it's certainly possible to try to build a model that simulates water content or pore-water-pressure, given certain forcing stresses. Be aware that Pastas was mainly tested on groundwater levels, so not all the default choices/settings might be optimal for different variables. As for predictions, once you have a trained model, you can use the estimated parameters to simulate the response of your simulated variable given new forcing stresses. # build model
ml = ps.Model(your_time_series)
sm = ps.StressModel(stress, ...) # make sure your stress contains data for both the calibration period and prediction period
ml.add_stressmodel(sm)
ml.solve(tmin=tmin, tmax=tmax) # set calibration period
prediction = ml.simulate(tmin=tmax) # prediction See also the notebook on uncertainty quantification for more tips on how to get an ensemble of predictions given model parameter uncertainty. Hope that answers your question somewhat? |
Beta Was this translation helpful? Give feedback.
-
Pastas has been applied to soil moisture time series before. I don't know too much about it but there is a paper with more information: https://www.sciencedirect.com/science/article/pii/S1364815220300876 |
Beta Was this translation helpful? Give feedback.
-
HI,
I am wondering if it would be possible to use pastas to back-calculate and predictid the trend of variables different that groundwater level. Specifically I am thinking about water content and pore water pressure.
We have a large dataset available.
best
LPi
Beta Was this translation helpful? Give feedback.
All reactions