PLT - An Evaluation of Uncertainty Associated with Generalized PMP Estimation Methods

When a dam is classified as High (with dam break consequences involving loss of life or extreme economic, environmental or cultural incremental damages) the Inflow Design Flood (IDF) is typically in the range of very large floods defined by the Probable Maximum Flood (PMF). This flood is typically calculated using an estimate of the Probable Maximum Precipitation (PMP), for which, among other approaches, generalized methods based on existing studies can be used. There are uncertainties in the application of these methods, that could have serious implications when assessing the adequacy of a particular dam to pass its required IDF.

We compared three of the generalized methods of PMP estimation that are available for a small watershed (25 km2) in Ontario, Canada: the Bruce Graphical Method (1957), the findings of the study PMP for Ontario (IBI, 2006), and the National Weather Service Hydrometeorological Report (HMR) 51 (USACE, 1978).

To provide context to the results obtained with the generalized methods, we reviewed reports for other areas in Ontario for which more detailed hydrometeorological PMP analyses had been conducted. This review indicated that the results of the three generalized methods varied with the size of the watershed as follows:

• The IBI method was the closest to the hydrometeorological analyses for small watersheds (areas in the range of 25 to 500 km2). The HMR 51 approach generally resulted in the highest values for small watersheds.
• The Bruce Method resulted in the lowest PMP depth values for small and large watersheds.
• The IBI and HMR 51 methods provided similar results for large watersheds, which were also in the same range as those obtained with the more rigorous hydrometeorological analyses.

These results highlight the effect of the watershed size, on the degree of uncertainty associated with the PMP estimates obtained from the generalized methods, and its implications on assessing spill capacity adequacy.