Report for Group Assignment 1.3

CEGM1000 MUDE: Week 1.7, Friday, Oct 18, 2024.

Remember to read the grading and submission instructions in the README.md file thoroughly before finalizing your answers in this document!

Questions

Question 1

Give a short description of the provided dataset in statistical terms. Visualize the data and choose a parametric distribution function for each variable between the indicated ones: choose between Uniform and Gaussian distribution to model the water depth, and between Exponential and Gumbel for the water velocity. Justify your choice based on the previous description and visualization.

You should describe your data with only a few sentences, and be sure to use quantitative information! Refer to that description to choose the parametric distribution function. You can also include some plots that may support your reasoning.

Question 2

Fit and assess the goodness of fit of the selected parametric distribution functions to the dataset. Use (at least) Kolmogorov-Smirnov test and one graphical technique. For the graphical technique, choose between the QQplot and logscale plot. Describe in a few sentences how the chosen parametric distribution functions performs.

Remember to use quantitative information based on the goodness of fit metrics that you have used. Also, you can include some examples of the differences in the computed and observed non-exceedance probabilities.

Question 3

Propagate the uncertainty through the equation to estimate discharge, $q$, using both observations and simulated samples. Discuss the differences between the two obtained distributions.

Question 4

Compare the simulated samples and the observations in a scatter plot. What are the differences? Is there anything you could improve in the analysis? Provide with recommendations to improve the performed analysis. They can be both about the univariate distributions and about the propagation of uncertainty method you have used.

Hint: Compute the correlation coefficient between $h$ and $u$ for both the observations and the simulated samples.

Last Question: How did things go? (Optional)

Use this space to let us know if you ran into any challenges while working on this GA, and if you have any feedback to report.

End of file.

© Copyright 2024 MUDE, TU Delft. This work is licensed under a CC BY 4.0 License.