Linking snowpack dynamics and summer streamflow in transient and seasonal snow zones of the Western Cascades

Despite extensive research on snow in the western United States, the effects of climate and snowpack on streamflow are poorly understood in forested mountain watersheds of the Pacific Northwest, USA. This study assessed relationships among air temperature, precipitation, snow, and streamflow over the period 1998 to 2020 from nine watersheds (8.5 to 580 ha) dominated by conifer forest, spanning transient (410 to 800 m) and seasonal (800 to 1630 m) snow zones in the H.J. Andrews Experimental Forest, Oregon. Data were summarized by season—fall (Oct–Dec), winter (Jan–Mar), spring (Apr–Jun), and summer (Jul–Sep) – and analyzed using linear regression and all-subsets regression. Daily snowmelt losses and streamflow were analyzed using cross-correlation. Air temperature was strongly related to snow and streamflow. Snow fraction (the fraction of precipitation falling as snow) was negatively related to air temperature during the snow season. Runoff ratios (unit-area streamflow divided by precipitation) were negatively related to air temperature in winter and spring in the transient snow zone and in spring in the seasonal snow zone. Summer streamflow (on July 1 and during minimum summer streamflow) also was negatively related to spring air temperature and positively related to spring precipitation in both transient and seasonal snow zones. Controlling for air temperature, winter runoff ratios declined over the study period in transient snow zone watersheds, including watersheds with maturing forest plantations. Relationships of streamflow to measures of snowpack differed by watershed elevation. Runoff ratios were positively related to snow fraction in winter and spring in the transient snow zone, in spring in the seasonal snow zone, and in summer in the highest elevation watershed (757 to 1620 m). Snow disappearance date and snow fraction also explained the timing of minimum summer streamflow in most study watersheds. Several snow variables were highly correlated with one another. Daily snowmelt was significantly positively cross-correlated with streamflow for 120 days at watersheds in the transient snow zone and for 150 days at watersheds in the seasonal snow zone. These findings indicate that precipitation falling as snow contributes to streamflow during both winter and spring in the transient snow zone but only in spring in the seasonal snow zone. Findings also imply that in these conifer-forest dominated watersheds, when air temperature is higher, more precipitation falls as rain, and evapotranspiration is greater, which in turn decreases streamflow. These effects vary with elevation, occurring during winter and spring in watersheds below 800 m, and in spring and summer in watersheds above 800 m. Also, based on daily data, the positive effect of snowmelt on streamflow persists for five months in the seasonal snow zone compared to four months in the transient snow zone. Long-term declines in runoff ratios in winter, after removing the effects of air temperature, also imply that winter evapotranspiration has increased over the 23-year study period, perhaps as a result of forest growth. These findings highlight the combined influence of vegetation, snow regime, air temperature and precipitation on the sensitivity of streamflow to snowpacks. These findings underscore the need for water resource planning to account for forest age, forest management, and snow regime effects on streamflow, particularly where rising temperatures may increase evapotranspiration and lessen snowmelt contributions to streamflow.