Development of a Simple Framework to Assess Hydrological Extremes using Solely Climate Data

English Abstract:

Extreme flow conditions such as droughts and floods are in general the direct consequences of short- to long-term weather/climate anomalies. For example, in southern Quebec, Canada, summer 7-day low flows are due to summer and fall precipitations. Which prompts the question: is it possible to assess future extreme flow conditions from meteorological/climate indices or should we rely on the classical approach of using outputs of climate models as input to a hydrological model? The objective of this study is to assess six hydrological indices describing extreme flows at the watershed scale (Q_max, Q_min;7d, Q_min;30d for two seasons: winter and summer) using local climate indices without relying on the aforementioned classical approach.

To establish the relationship between climate and hydrological indices, daily precipitations, minimum and maximum temperatures from 89 climate projections are used as inputs to a distributed hydrological model. River flows from five hydrographic regions in Québec are separately calibrated for winter and summer seasons and the watersheds are divided into upper-, middle-, and lower-sections. This allows to better take into account the different driving mechanisms behind the generation of high and low flows, as for the snow-melt or heavy convective rainfall events induced high-flows of late winter and summer. To identify the best predictors between 1961 and 2100, hydrological indices are extracted from the flow series, and climate indices are computed for different time intervals (from a day up to four years).

Overall, preliminary findings clearly illustrate that the change in the hydrological indices can be detected through the concurrent trends in the climate indices. The use of many climate projections ensures the relationships are not simulation-dependent and shows summer events are particularly at risk with increasing high flows (60% of the variability is explained on average) and decreasing low flows.

The development of a simple predictive tool to assess the impact of climate change on flows represents one of the major spin-off benefits of this study and may proove to be useful to municipalities concerned with source water and flood management. Future work includes the use of different sets of climate indices for watersheds clustered according to their physiographic descriptors.