Figure 1.  Photographs of control (forested) and gap (open) study sites on 8 May 2011.

Forest Management to Maximize Snow Retention under Climate Change

Figure 1. Photographs of control (forested) and gap (open) study sites on 8 May 2011.


The mountain watersheds of the Pacific Northwest are expected to experience a decrease in snowpack depth and earlier snowmelt in coming decades.  Snowpack functions as a reservoir, storing water in the winter and releasing it during the spring.  Thus, earlier snowmelt will decrease late summer runoff, leading to warmer stream temperatures, diminishing fish habitat, and less-reliable water supply to cities.  However, land management also influences snowpack accumulation and melt.  Forests shelter snow from solar radiation and wind, but contribute to snow melt by forest-emitted longwave radiation, and to snow loss by canopy interception.   Forest management strategies such as thinning or cutting small gaps may increase snow water storage and lengthen the snowmelt season.  However, the optimal forest structure for maximizing snow retention in Pacific Northwest climate conditions is not known.  Our objective is to quantify the overall influence of forest structure on snow in order to help land managers  balance the needs of future water supply, ecosystem health, timber harvest, and fire prevention in mountain watersheds.


Figure 2. Average values of manual surveys of snow depth in four forest types.

In order to understand current snow processes and to predict the effects of climate change and forest change on snow retention and streamflow, we use a combined observational and modeling approach.  We measure snow and meteorology in a range of forest types, including old growth, dense 2nd growth, thinned 2nd growth, and gaps (Figure 1).  Our sites are located in the Cedar River Municipal Watershed, which is managed by Seattle Public Utilities (SPU) and provides water to 1.4 million people.  The watershed is located 50 km east of Seattle in the intermittent snow zone, where the snow is close to freezing at all times, making it particularly sensitive to variations in temperature.  We collect data on snow depth with manual transect measurements, on snow duration with inexpensive ground temperature sensors and fiber optic cable (Lutz et al., 2012), and on snow interception by trees with  a unique instrument that measures the compression in tree trunks to determine the weight of the snow in the tree canopy  (Martin et al., 2012).

Figure 3. Temperature (in degrees C, see colorbar on left) as a function of time and distance along the forest floor at a second-growth forest study site. Solid dark blue lines (0C) illustrate the presence of snow cover, which insulates the ground and prevents diurnal temperature fluctuations. Forest treatments are identified on the right, where both ‘Forest’ and ‘Control’ indicate untreated 80-yr old forest. While snow persisted longer in the gaps than the untreated and thinned forest plots, the variability between different gaps was more than one week.

Combined, these observations let us calculate the net effect of forest type and structure on snow by comparing peak snow water equivalent (SWE) and snow duration.  Preliminary observations indicate that maximum snow retention occurs in gaps and small clearings in the Cedar River watershed, but snow under the thinned forest is the same as under the control forest (Figure 2).  More snow accumulates in these areas due to the lack of canopy interception losses. The melting rate is not lessened in forests by shading from solar radiation and wind, perhaps due to cloud cover during much of the melt season.  Continuing observations will allow us to assess these relationships in high and low snow years, and to extend observations to more forest types.  Our observations also provide evidence to validate distributed hydrology models.   In 2012, we acquired airborne LiDAR to more accurately characterize the forest canopy and to parameterize forest structures on the watershed scale.


Lutz, J. A., Martin, K. A., and  J. D. Lundquist, 2012. Using fiber optic cable to measure surface temperature in heterogeneous forests, Northwest Science, 86(2) 108-121. DOI:

Martin, K.A., Van Stan II, J.T., Dickerson-Lange, S.E., Lutz, J. A.,Berman, J.W., Gersonde, R, and  J. D. Lundquist, 2012 (accepted).   Development and testing of a snow interceptometer to quantify canopy water storage and interception processes in the rain/snow transition zone of the North Cascades, Washington, USA.  Water Resources Research.


This work is funded by the National Science Foundation (NSF) under CBET-0931780 and through an NSF Graduate Research Diversity Supplement (GRDS).

LiDAR flights were funded by the National Center for Airborne Laser Mapping (NCALM), Graduate Student Seed Grant for LiDAR Acquisition, 2012.


Jessica Lundquist (Department of Civil and Environmental Engineering, University of Washington)

Jim Lutz (School of Environmental and Forest Sciences, University of Washington)