Hillslope hydrology in the Rennell Sound area of the Queen Charlotte Islands

This experiment was designed to study the hydrologic response of three first order watersheds to the inputs of precipitation, especially those inputs received during storm events. The study was conducted in the Gregory Creek watershed in the Rennell Sound area of the Queen Charlotte Islands. There were six data collection sites, two in each of three sub-basins. Groundwater levels and surface flow were monitored over an 82 day period from November 26th, 1991 to February 15th, 1992. Precipitation data were collected over a 40 day period from November 26th, 1991 to January 4th 1992. The data were analyzed by comparing precipitation data to groundwater levels and surface flow using time-series and regression analysis. Storm precipitation was compared to groundwater levels and surface flow (for the same time as the storm event). Including antecedent precipitation in this analysis did not increase the number of significant regressions relationships. Time-series analysis revealed strong correlations between groundwater levels at wells within the same site. The two sites in watershed number one produced more significant relations between storm precipitation and groundwater response than watersheds two and three when analyzed using time-series analysis. The watershed that produced significant results using time-series analysis did not produce significant regression results. Alternatively, one watershed which did not produce significant time-series results did produce significant regression results. Mean groundwater level changes during storm events were significantly greater at the upper site in one watershed, the lower site in another and inconsistent in the third. There was a wide range in the mean time to peak for all sites during storm events (5.6 hr - 12.1 hr), but the differences were not statistically significant. An analysis of covariance indicated that there was not a common regression equation that defined the relation of changes in groundwater levels or time to peak during storm events for all three watersheds in the study. It is hypothesized that the differences between sites and watersheds may be a result of individual differences in geologic characteristics at each site and instrument location.Forestry, Faculty ofGraduat