Among all existing methods for estimating extreme floods, the probable maxi
mum flood method is the safest, since it is a flood with a probability of e
xcedance that is theoretically zero. In the early 1970s, this flood was cal
culated as the combination of the probable maximum precipitation (PMP) and
the probable maximum snow accumulation (PMSA). In the 1990s, this combinati
on has been considered to be highly improbable. Experts advise against comb
ining two maximized events and, instead, recommend combining one maximized
event with a relatively typical extreme event. This article presents a sens
itivity analysis on the return period to be used for the typical extreme ev
ent to be combined with the maximized event to obtain a "more realistic" PM
F To achieve this, all the steps of a PMF study were reviewed and applied t
o the Sainte-Marguerite watershed, i.e., calibration and (or) validation of
SSARR model, estimation of the PMP, the PMSA, and the temperature sequence
. Different flood scenarios have been simulated including accumulated snowf
all corresponding to return periods of 50, 100, and 500 years,followed by P
MP and PMSA, followed by precipitation corresponding to return periods of 5
0, 100, and 500 years. It has been noticed that the use of a return period
of 50, 100, or 500 years, to represent the unmaximized extreme event, has l
ittle effect on the hydrologic response of the basin. Based on the results
of this work the use of a return period of 100 years is recommended.