Room for the River: An Analysis of Integrated Water Basin Management Strategies for the Province of Overijsssel in the Netherlands

Released by: EPA1361 Group 23

Name	Student Number
Yaren Aslan	5257514
Ludovica Bindi	5469856
Alexandre Curley	5500125
Imam Masud	5276993
Aspasia Panagiotidou	5631211
Dorukhan Yesilli	5539501

---

Introduction

This repository contains our analysis to address the case study of the upper branch of the IJssel River, which aimed at developing a flood risk management plan for the Province of Overijsssel.

We conducted a Multi-Scenario Many Objective Robust Decision Making (MS-MORDM) (Bartholomew & Kwakkel, 2020) using the model developed by Kwakkel (2021). This analysis is performed thanks to the functionalities developed in the ema_workbench Python package developed by Kwakkel (2022a).

How to Use

The repository contains both the provided model files (Kwakkel, 2021b) and scripts and notebooks created to carry out our analysis (all the files that start with final_step).

To perform the policy, run the the analysis folder in order according to the step number. The results of one step are saved via csv or pickle file and opened in the next file.

To plot the pictures:

• the results of the re-evaluations of the optimal policies (steps 2 and 6) are analyzed and plotted in the final_parallel_coordinate_plot.ipynb
• the plots of the results of the robustness metrics computation (step 3 and 7) are created in their corresponding notebook files (final_step2_reevaluate_policies.ipynb and final_step7_compute_robustness_second_optimization.ipynb)
• the plots for the scenario discovery (step 4) can also be found in the corresponding notebook file (final_step4_scenario_discovery.ipynb)

Files

MBDM-Final_Project

• README.md: this markdown document
• dike_model_function.py: contains the model creation functions
• funs_dikes.py: contains functions for the dike physics sub-model
• _funs_economy.py: contains functions for the costs sub-model
• funs_generate_network.py: contains functions for generating a network of the dike rings
• funs_hydrostat.py: contains functions for the waves sub-model
• problem_formulation.py: contains a function to attach the model to the ema_workbench
• final_parallel_coordinate_plot.ipynb: used to plot Parallel Coordinate Plots for the re-evaluation results
• final_step1_optimization.py: optimization under reference scenario
• final_step2_reevaluate_policies.py: re-evaluation of the optimal policies under sampled scenarios
• final_step3_compute_robustness.ipynb: computation of robustness for the optimal policies
• final_step4_scenario_discovery.ipynb: scenario discovery based on the optimal policies performances
• final_step5_multi_scenario_optimization.py: optimization under scenarios found via scenario discovery
• final_step6_second_reevaluate_policies.py: re-evaluation of the new optimal policies
• final_step7:compute_robustness_second_optimization.ipynb: computation of robustness for the new optimal policies
• report
  • report.pdf: report of our analysis
• data: contains datasets used by the model functions
• intermediate outputs: contains datasets to pass results from one step to the followings
  • step5 - multi scenario optimization results: contains the results of step 5
  • bin: older versions of the results
• output pictures: contains the pictures created for the analysis
• bin: contains older versions of the final files and temporary files

Reference

Bartholomew, E., & Kwakkel, J. H. (2020). On considering robustness in the search phase of Robust Decision Making: A comparison of Many-Objective Robust Decision Making, multi-scenario Many-Objective Robust Decision Making, and Many Objective Robust Optimization. In Environmental Modelling & Software (Vol. 127, p. 104699). Elsevier BV. https://doi.org/10.1016/j.envsoft.2020.104699 Kwakkel, J. H. (2022a). ema_workbench - Docs - A High Level Overview. Retrieved from https://emaworkbench.readthedocs.io/en/latest/overview.html Kwakkel, J. H. (2022b). Epa1361_open, final assignment. Retrieved from https://github.com/quaquel/epa1361_open