The cumulative effects of forest management on peak streamflows during rain-on-snow events

Rain-on-snow (ROS) events occur when warm, windyrainstorms pass over a relatively ripe snowpack melting itpartially or completely. In western Washington ROS eventsoccur most frequently within the transient snow zone ofthe Cascade Range. The location of the transient snowzone varies locally, but in the middle Cascades thetransient snow zone extends from roughly 300 m to 900 m inelevation.
The effects of clear-cutting on the rate and volumeof water delivery to the soil during ROS events have beendocumented recently. The forest canopy affects both theaccumulation and melting of snowpacks. While snow tendsto accumulate in clear-cuts, it can be intercepted by theforest canopy where it melts and falls to the forest flooras water or in clumps of wet snow. Higher wind speeds inclear-cuts result in higher melting rates in clear-cutsthan in forests during ROS events. The combination ofgreater snow accumulation and higher melting rates canresult in higher rates and volumes of water delivery tothe soil in a clear-cut than in the forest.
The main objective of this study is to investigatehow forest management affects the size of peak flows fromwatersheds in the transient snow zone during ROS events.Observed meteorological data, an energy balance equation,and a conceptual watershed model were used to demonstratehow clear-cutting affects the size of peak flows fromhillslopes (z 16 ha) and how this effect varies over the
landscape during ROS events. A conceptual hydrologicmodel that has the ability to route water inputs through alinked hillslope and stream network was used to determinehow harvest pattern and harvest intensity affect peakflows in a watershed (74 6400 ha) in the transient snowzone during ROS events.
The effect of clear-cutting on the size of hillslopepeak flows during ROS events was shown to be dependent onthe storm and the initial snowpack conditions under eachforest cover. This effect varies over the landscape withaspect, elevation, and hillslope hydrological behavior.Hillslopes that produce a more peaked hydrograph responseto soil water inputs experience greater increases in thesize of hillslope peak flows following harvest thanhillslopes that generate a more attenuated response. Theincrease in the size of peak flows following clear-cuttingis greater on windier aspects. The increase in the sizeof peak flows following clear-cutting at differentelevations depends on the snow conditions and the weatherat each elevation. Since clear-cutting has a variety ofeffects on the size of hillslope peak flows in awatershed, harvest intensity alone is an inadequate indexfor forest harvest effects on peak flows during ROSevents. The aspect and elevation and the hydrologicbehavior of individual harvest units must also beconsidered.
The results suggested that harvest pattern has littleinherent effect on the size of peak flows from watershedswith drainage areas (PJ 6400 ha), topography, and drainagenetworks similar to the study watershed. However, harvestpattern indirectly affects the size of peak flows to theextent that the different harvest patterns containdifferent aspects and elevations and hillslopes withdifferent hydrologic behavior.
The results also suggested that harvest intensity has
a significant, inherent effect on peak flows fromwatersheds with drainage areas (= 6400 ha), topography,and drainage networks similar to the study watershed. Fora particular ROS event, the ratio of harvested peak flowto forested peak flow increased with harvest intensity.However, there was significant scatter in the ratio amongevents for a particular harvest intensity. This suggeststhat there is a wide range of forest harvest effects onthe size of peak flows during ROS events depending on themeteorological conditions during the storm and the initialsnowpack conditions.