Combination_Uncertainty

Title

Quantifying uncertainty in flood predictions due to estimated river bathymetry and arbitrary conventions in grid alignment

Authors

Martin Nguyen, Prof. Matthew D. Wilson, Dr. Emily M. Lane, Prof. James Brasington, and Dr. Rose A. Pearson

Overview

Digital elevation models (DEMs) generated by sampling and interpolating LiDAR data onto a square grid can produce reliable flood predictions. However, the estimated river bathymetry and arbitrary conventions in grid alignment that can introduce uncertainty in flood predictions are frequently overlooked. Hence, our research quantified this uncertainty using a Monte Carlo approach and flood model LISFLOOD-FP to generate multiple flood simulations for analysis.

Description

This work is used to produce data for developing several machine learning models here. To generate the uncertainty in flood modelling, we generate the uncertainty arising from the estimated river bathymetry and arbitrary conventions in grid alignment. The final uncertainty data from Buller, Waikanae, and Esk River sites (New Zealand) will then be used to train and test the machine learning models. The workflow is as below:

Reproducibility

The project was written under modules format (for three sites) and has not been updated to a class format that can be used for multiple sites. If further information is needed please contact with the Github author.

Environment installation

Download files forbuller&waikanae_packages.yaml and foresk_packages.yaml and use the following command (reference from the answer of @merv) to recreate two anaconda environments (only support Windows at the moment).

conda env create -n forbuller&waikanae -f forbuller&waikanae_packages.yaml

and

conda env create -n foresk -f foresk_packages.yaml

Data

This project includes 3 main folders as described below.

• Buller: Includes three sub-folders Script_development for background scripts, Script_execution for executing the background scripts to generate simulations, and Script_extra for editing the results if there are any faults.

• Waikanae: Includes three sub-folders Script_development for background scripts and Script_execution for executing the background scripts to generate simulations.

• Esk: Includes three sub-folders Script_development for background scripts, Script_execution for executing the background script to generate simulations, and Script_extra for editing the results if there are any faults.

Acknowledgement

This project is part of the NIWA-led national programme, "Reducing flood inundation hazard and risk across Aotearoa - New Zealand", funded by the Ministry for Business, Innovation and Employment (MBIE) Endeavour Programme. A PhD scholarship was provided from this programme to Martin Nguyen. The LISFLOOD-FP flood model was developed at the University of Bristol and is available here.

Contacts

Github author: Martin Nguyen martinnguyen20494@gmail.com