We combined the methods and advantages of stochastic hydrology and paleohydrology to
estimate 900 years of weekly flows for the North and South Saskatchewan Rivers at Edmonton and Medicine
Hat, Alberta, respectively. Regression models of water-year streamflow were constructed using historical
naturalized flow data and a pool of 196 tree-ring (earlywood, latewood, and annual) ring-width chronologies
from 76 sites. The tree-ring models accounted for up to 80% of the interannual variability in historical naturalized flows. We developed a new algorithm for generating stochastic time series of weekly flows constrained by the statistical properties of both the historical record and proxy streamflow data, and by the
necessary condition that weekly flows correlate between the end of a year and the start of the next.