A spatio-temporal analysis of trends in Northern Hemisphere snow-dominated area and duration, 1971–2014

Abstract

ABSTRACT. Seasonal snow-cover modulates water and energy budgets across large areas of the Northern Hemisphere. Previous research, based on satellite imagery interpreted and curated by the Rutgers University Snow Laboratory, has identified significant negative and positive trends in annual snow-covered duration and area at hemispheric and continental scales between 1971 and 2014. This study uses the same dataset to generate more detailed descriptions of spatial variations in these trends, maps intraannual variations in sign, statistical significance and strength, and quantifies associations with latitude and elevation. It also considers the limitations and uncertainties associated with a binary classification of this type, and the implications for trend magnitudes of adopting alternatives to the conventional assumption of 100% (0%) actual fractional snow-covered area in 'snow-covered' ('snow-free') spatial units at different stages of the snow-season. This prompts adoption of alternative terminology, referring to 'snow-dominated' area and duration. In response to questions about the dataset's veracity raised by some prior studies, it discusses climatological factors of potential relevance in explaining spatio-temporal trend patterns, and considers how biases might possibly have been introduced as a result of extraneous influences.

Regional influence of ocean–atmosphere teleconnections on the timing and duration of MODIS-derived snow cover in British Columbia, Canada

“…This allowed us to not only detect SD DUR from the time series, but also SD ON and SD OFF . To extend this work further back in time, one could use lower-spatial-resolution Advanced Very High Resolution Radiometer (AVHRR) data (Allchin and Déry, 2017) or investigate in situ measurements (McClung, 2013;Barton, 2017). Our methods could be used anywhere that the MODIS snow cover product is present; however, the definition of the hydrological year, the LOWESS bandwidth, and the NDSI thresholds may need to be optimized.…”

“…BC has undergone significant landscape changes over the last 20 years. These include widespread forest insect outbreaks (e.g., Stahl et al, 2006b), severe wildfire seasons (e.g., Crowley et al, 2019), summer drought (e.g., Barnett et al, 2005), and extensive harvesting of forests (e.g., Hansen et al, 2013). All of these land cover changes have important effects on snow hydrology (Parajka et al, 2012) and may act to either enhance or reduce the observed interannual SD response to both the ONI and PDO.…”

“…The timing and duration of seasonal snow cover have been shown to influence a number of landscape processes, namely soil moisture and drought (e.g., Fang and Pomeroy, 2008), wildfire activity and intensity (e.g., Westerling et al, 2006), freshwater availability (e.g., Barnett et al, 2005), flooding (e.g., Barnett et al, 2005), plant ecology (e.g., Jonas et al, 2008), and wildlife behavior (e.g., Boelman et al, 2019). The western Canadian province of British Columbia (BC) is roughly 1 million km 2 in size, has multiple mountain chains, and shares its border with one province, two territories, and four states (Fig.…”

“…Generally mild and dry winters in BC have been identified as important factors in forest insect outbreaks (Stahl et al, 2006b), wildfire activity (Crowley et al, 2019), glacier change (e.g., Menounos et al, 2018), and overall drought conditions (Trishchenko et al, 2016). The winter snowpack differs in timing and duration throughout the province and is influenced by, among other processes, climate change (e.g., Allchin and Déry, 2017;Foord, 2016;Najafi et al, 2017), local geography (e.g., Moore and McKendry, 1996;Schnorbus et al, 2014;Stahl et al, 2006a), land cover change (e.g., Boon, 2012), and ocean-atmosphere teleconnections (e.g., Fleming and Whitfield, 2010;Hsieh and Tang, 2001;Shabbar, 2006;Winkler and Moore, 2006).…”

A spatio-temporal analysis of trends in Northern Hemisphere seasonal snow-cover, 1971-2017

“…The annual transformation of terrestrial physical properties and conditions by seasonal Administration's climate data record of NH snow-cover, compiled and curated by the Rutgers University Snow Laboratory (hereafter the 'NOAA-Rutgers Snow Archive', NRSA), offers the longest period of record of all remotely-sensed environmental datasets (Robinson et al 2012;Estilow et al 2015). It has therefore formed the foundation for a series of SSC-related investigations since the mid-1990s (Robinson et al 1993;Groisman et al 1994;Frei et al 1999;Dye 2002;Rikiishi et al 2004;Déry and Brown 2007;Choi et al 2010;Cohen et al 2012;Liu et al 2012;Hernández-Henríquez et al 2015;Allchin and Déry 2017).…”