The Distributed Hydrology Soil Vegetation Model in conjunction with synthetic climate data was used to investigate the impact of forest harvesting upon the peak annual discharge regime of two headwater streams, 240 and 241 Creeks, in south-central British Columbia (BC). For return periods (T) in the range of 1.25 to 100 years, simulated quantile magnitude (QT) increases over the control value with increasing clearcut harvest area, represented as a proportion of basin area. This general trend is independent of the temporal scale of discharge, as represented by hourly, daily, and 7-day streamflow. For both 240 and 241 Creeks QT increases are consistently statistically significant (a = 0.05) at proportion harvest areas = 30%. Fitted Generalized Extreme Values (GEV) distributions suggest that for a snowmelt dominated peak discharge regime, relative (to control) post-harvest quantile change (?QT) increases with increasing return period, with no apparent asymptotic limit. However, this trend substantially overestimates the post-harvest response of the most extreme peak discharge event for the simulated record (Control T > 100 years), which is entirely rainfall driven. Pooled ?QT results indicate that the relationship between ?QT and harvest area is quite strong and suggests that, regionally, the peak discharge response can be predicted using only harvest area as the predictor variable. However, for hourly discharge the effect of harvest elevation (categorized as occurring either above or below the median elevation for each basin) ads uncertainty to the pooled ?QT-area relationship. Closer examination of the ?QT response for 240 Creek suggests that at the spatial scale characteristic of the study area the hourly response relates to the relative apportionment of melt runoff into either a fast surface or slow sub-surface component. As such, the elevation dependence of hourly ?QT is related to the significant spatial discontinuity in drainage density that occurs at the H50 elevation; a morphologic characteristic that may or may not be representative of south-central BC. At the daily and 7-day temporal scale differences in flow path travel times are less significant and ?QT exhibits only moderate elevation dependence.
Hydrologic recovery is defined as the process by which the hydrologic characteristics of a watershed that has been subject to harvesting are restored to near pre-harvest condition by forest regeneration. Although the relation between the level of clearcut and hydrology at the watershed scale is not well understood threshold rate of cuts continue to be used to constrain forest management in British Columbia (BC). Until this relation is understood in a quantitative manner we may be unduly restricting forest development at great cost to the industry. The urgency to understand how clearcut affects hydrology has increased with recent pine beetle epidemics and wildfires that are creating disturbances in large watersheds equivalent to 100% ECA. In this project, we use the University of British Columbia Watershed Model (UBCWM) at two interior snow dominated watersheds to quantify the effects of total clearcut of an entire watershed and investigate the nature of the relationship between these effects and hydrologic recovery over time. The main focus of the project is the effects of clearcut on the water yield and peak flow regime.
Charles Q. Luo and Younes Alila.
Luo, Charles Q., Alila, Younes; Strimbu, B.; Schnorbus, Markus A.; Winkler, Rita D.; Spittlehouse, David L.. 2006. Developing thresholds and guidelines for a key hydrologic indicator of watershed function. Forest Investment Account (FIA) - Forest Science Program. Forest Investment Account Report. FIA2006MR118