Implications of climatic variations in temporal precipitation patterns for the development of design storms in Newfoundland and Labrador

Abstract

The distribution of precipitation in time is an important aspect for the development of design storms for storm water infrastructure design. The current set of mass curves used throughout the province of Newfoundland and Labrador (NL) may not be justified. To identify variation in mass curves across NL, and compare results with existing mass curves, hourly precipitation data from 10 stations were used. Bayesian k-means clustering was used to identify dimensionless mass curves to represent precipitation patterns. Eight distinct temporal patterns of precipitation were identified and further regrouped into four, useful for making recommendations on the choice of mass curve. Crosstabulation applied to the patterns were found to be significantly influenced by event duration, depth, and climate zone. Results support the conclusion that climate was an important determinant of temporal distribution of precipitation, and it is important to determine which pattern is dominant in a given region.