Interannual Variations in Spring Snowmelt Timing of Alaskan Black Spruce Forests Using a Bulk‐Surface Energy Balance Approach.

Spring snowmelt occurs for a short duration on an annual time scale, but their timings considerably affect the carbon and hydrological cycle in high‐latitude ecosystems. Here, we developed a simple snowmelt model, treating the ecosystem surface as a bulk‐surface layer. Energy fluxes across this bulk surface and the snow‐soil boundary determine snow temperature and the energy utilized for snowmelt. Parameterizing the bulk surface using decade‐long eddy covariance site data from two Alaskan open black spruce forests offered an opportunity to quantitatively evaluate meteorological drivers affecting snowmelt timings without the needs for detailed canopy information. The sensitivity analysis suggested that the total snowfall on the forest floor, ranging from 0.35 m in 2016 to about 1 m in 2018 and 2020, was the most crucial driver for snowmelt timing. This factor accounted for a 10‐day difference in the interannual variations in snow disappearance dates. The importance of the snowfall varied from year to year, and in 2013, the late snowmelt was characterized by low air temperatures, which increased sensible heat loss from the snowpack. The importance of atmospheric radiation was revealed in relatively warm years, such as 2016 and 2019. Our modeling approach necessitates adjusting one empirical parameter that reflects the heat conductivity from the bulk surface to the snow, based on observations. Nevertheless, despite this need for adjustment, the bulk‐surface approach helps identify important meteorological drivers underlying observed snowmelt within a simple theoretical framework. Plain Language Summary: It is essential to understand why and how spring snowmelt occurs in boreal forests because of its impact on both the biological and hydrological processes in the ecosystem. However, meteorological drivers related to snowmelt often change simultaneously, which makes it challenging to determine which single driver is vital for snowmelt. To address this challenge, we constructed a simple snowmelt model that treats the ecosystem surface as a single bulk layer. The simple model bypasses the needs for detailed canopy information and is easily constrained by observed data. Applying the model to decadal observation data from Alaskan boreal forests, we found that snowfall, air temperature, and atmospheric radiation were important meteorological drivers for explaining interannual variations in snowmelt timing. Each of them explained about 1–2 weeks of snowmelt disappearance dates. We also demonstrated that the model helped explain why the snowmelt was late in 2013 and early in 2016 and 2019 and why snow disappearance dates differed between study sites. While a simple model has its limitations, it can be beneficial for understanding snowmelt characteristics within a simple theoretical framework. Key Points: A model‐based approach was introduced to delineate the roles of meteorological drivers on interannual variations in snowmelt timingSnowfall was the most important driver of the interannual variation in snowmelt timing followed by air temperature and atmospheric radiationLate snowmelt in 2013 was attributed to low air temperature, increasing sensible heat loss from the snowpack [ABSTRACT FROM AUTHOR]