Forest harvesting effects on peak flow : a deterministic-statistical approach

• Main Author: Schnorbus, Markus Albert
• Format: Others
• Language: English
• Published: 2009
• Online Access: http://hdl.handle.net/2429/14257

Logging activities in snowmelt-dominated catchments may alter the magnitude and frequency of annual peak discharge events. It is generally accepted that the forest canopy exerts a strong influence on the processes of snow accumulation and ablation, such that canopy removal will result in locally increased snow accumulation and more rapid melt. The basin-wide impact of canopy removal is less well defined and three issues in particular require greater understanding: a) the relative influence of basin physiography; b) the influence of the duration of the streamflow metric (i.e. hourly, daily, or 7-day peak flow); and c) the influence of event frequency (i.e. return period). To address these issues a simulation study was conducted of the Redfish Creek basin, which drains an area of 26 km² in the Kootenay Mountains of south-eastern British Columbia. The catchment was divided into three elevation bands of equal area and six scenarios simulated 50 and 100% clearcut harvesting of each elevation band and three scenarios simulated the impact of clearcut harvesting of 33, 66 and 100% of the entire watershed area, distributed evenly between the three elevation bands. The peak flow regime for each scenario was described by deriving flood frequency distributions from simulated 100-year series of hourly, daily, and 7-day average streamflow. The scale of peak flow change was established by comparing individual flood quantiles of return periods 1 to 100 years with those derived using pre-harvest land cover. Streamflow simulations were conducted using the Distributed Hydrology-Soil- Vegetation Model (DHSVM), which is an explicit hydrologic model that simulates the land surface water and energy balance simultaneously at individual computational grid elements distributed at a resolution of 25 m. DHSVM is driven by near surface meteorology, provided in the form of time series data; grid cells are parameterized individually to describe soil and vegetation properties; and topographic controls on meteorology, solar radiation, and water movement are represented with a digital elevation model. The paucity of observed meteorological data required that the driving time series of precipitation, air temperature, relative humidity, wind speed, and shortwave and longwave radiation had to be generated using a stochastic weather generation algorithm. Results indicate that harvesting increases the magnitude of peak annual discharge and that these changes are increasingly more significant as the rate and elevation of the harvest treatment and the duration of the streamflow metric increases. Harvesting in the bottom elevation band resulted in a negligible impact to the peak flow regime while harvesting the entire catchment created the most significant impact. Statistically significant (p ≤ 0.05) increases in annual peak discharge were 4 to 13%, 4 to 18%, and 4 to 22% for hourly, daily, and 7-day average discharge, respectively. The peak flow increases were also dependent upon return period, with the significance of the impact decreasing with increasing return period. === Forestry, Faculty of === Graduate