TDS: Tree-based Direct Sampling (on 2D Topography modeling)

Code Reference -

• Zuo, C., Yin, Z., Pan, Z., MacKie, E. J., & Caers, J. ( 2020). A tree‐based direct sampling method for stochastic surface and subsurface hydrological modeling. Water Resources Research, 56, e2019WR026130.

This is the first version of the Tree-based Direct Sampling (TDS), with 2D Antarctica Topography modeling case as example.

Structure of the TDS setup

The TDS has two steps:

• Step 1: Training Phase, run under directory TDS_TrainingPhase. This step builds a tree using Training Image (TI)
• Step 2: Simulation Phase, run under TDS_SimulationPhase2.0. This step simulates direct sampling realizations using the tree built at Step 1.

Install and Run TDS

Please follow the provided User Manual to install and run TDS

To run the TDS, the following dependencies must be met:

• Jupyter
• Python 3.7
• We recommend to install the above two together via Anaconda.
• Java
• Eclipse IDE for Java

Tutorial

A Tutorial example is provided in the User Manual.

Contact

Please contact David Zhen Yin or Chen Zuo in case of questions.

Licensing

This repository is released under the MIT License.