New Pastas Article: Time Series Analysis of Nonlinear Head Dynamics Using Synthetic Data Generated with a Variably Saturated Model

[martinvonk]

The head response of the groundwater system to precipitation and evaporation can be highly nonlinear due to processes in the unsaturated zone. Pastas' linear time series models ignore these nonlinear processes, while the nonlinear time series models approximate them with a bucket type approach. But how do these time series models compare, and when should you use which? An approach to test the performance of these time series models is by using synthetic head data generated with a variably saturated groundwater model. Read the Groundwater article from @mbakker7, @raoulcollenteur, @sorab, @FransSchaars, and I if you want to know more: Time Series Analysis of Nonlinear Head Dynamics Using Synthetic Data Generated with a Variably Saturated Model (doi: 10.1111/gwat.13403).

The synthetic dataset can be reproduced via Zenodo or called using pastas.load_datasets("vonk_2024")