Goals

Vidrio-Sahagún et al. (2024) identified a number of issues with FFA research in Canada. Our framework aims to address these issues. In particular, we hope to achieve:

• Standardization: The United States uses a standardized distribution (LP3) and parameter estimation method (expected moments) for FFA. However, Canada does not provide such guidance. By developing this framework, we hope to limit subjectivity in flood estimation.
• Reproducibility: Many FFA studies in Canada do not provide essential information for reproduction, such information about exploratory data analysis (EDA) performed, the distribution selection mechanism, and even the probability distribution itself.
• Statistical Rigour: Historically, many FFA studies have not performed uncertainty quantification, which makes it difficult to draw accurate conclusion from the results. Our framework automatically performs uncertainty analysis to quantify potential errors.
• Research-to-Practice Translation: The disorganization of FFA research in Canada makes it difficult for regulatory agencies to access cutting-edge advancements in FFA methodologies. By providing a common set of techniques for modellers to use, we hope to bridge the gap between research and practice.

Development is guided by the following principles:

• Software Freedom: Our framework is built on free and open source software.
• Modularity: Users are allowed to use as much or as little of the framework as they like.
• Interoperability: Our framework can be seamlessly integrated with other flood models.
• Flexibility: Users can tailor their analysis to the nuances of individual watersheds.
• Clarity: The source code is easy to read and understand.
• Robustness: The framework can handle datasets with small sample sizes or missing values.