Climatology and spatial distribution of Northern Hemisphere cryo-cover
Date
2022
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
This manuscript examines sea ice and snow cover (here defined as “cryo- cover”) across the Northern Hemisphere through a series of three studies. Cryo-cover is unique in that it is both a responder to climate (sensitive to temperature) but is also a climate forcer (via impacting the radiation balance), therefore understanding its complex interactions and how they vary temporally and spatially are imperative for many applications within the atmospheric, environmental, and hydrologic sciences. ☐ The first study builds a climatology of cryo-cover using three datasets: the National Snow and Ice Data Center (NSIDC) Sea Ice Extent (1979 – 2020), the Rutgers Global Snow Lab (GSL) Snow Cover Extent (1979 – 2020), and the MeaSuREs State of the Cryosphere (1979 – 2012) products. Temporal trends of cryo-cover, both as a combined and as separate components are analyzed, and a derived cryo-cover concentration dataset is created using the MeaSuREs values. From these values cryo-cover is classified to be either stable (covered 75% of the time or more, which has decreased in most months over time) or transient (covered less than 75% of the time, which has increased in many months over time). Cryo-cover has predominantly decreased since 1979, with sea ice decreasing in all months and snow cover increasing in some areas in the autumn. ☐ The second study implements the derived cryo-cover concentration dataset (1979 – 2012) to investigate three key transitional cryo-cover months (March, June, and October). Self-Organizing Maps (SOMs) are used to define specific regions of cover transition across the Northern Hemisphere. Results suggest a transition from stable cover to transient cover in March and June as SOM identified stable cover patterns lose spatial area in favor of neighboring patterns with lower average concentration values gaining area over time. October results differ as there were regions experiencing a decrease in stability (especially along the sea ice domain edges) but also regions that experienced increasing cryo-cover trends centered around 50N latitude. Although all three months exhibited areas with both increasing and decreasing trends in concentration, the implementation of the SOMs made visualizing specific areas undergoing the most rapid transformation clearer. ☐ The third study also utilizes the derived cryo-cover concentration dataset and the SOM defined cryo-cover regions from the second study for March, June, and October (1979 – 2012). To better understand the transitions taking place in the regional patterns, a confusion matrix-based analysis is used to investigate how pixels move between the SOM nodes (regions). In general, regional cryo-cover patterns are not exchanging pixels (representative of area) with longitudinal neighboring patterns but instead exchange them with higher latitude patterns. In March, Central and Western Eurasia patterns gain the most area from the shrinking stable pattern, in June the stable pattern loses the most area in favor of the Central Eurasia and Alaska-Russia patterns, and in October things are less clear. ☐ Combined, the three articles build a unique understanding of cryo-cover fluctuations temporally, derive specific regions of cryo-cover transformation not previously documented, and calculate the amount of area being lost from patterns with higher concentration to those with lower concentration for the months of March, June, and October, with cumulative results suggesting a decrease in cryo-cover stability as regimes preferentially shift toward a more transient nature.
Description
Keywords
Climate change, Cryosphere, Sea ice, Snow cover, SOMs, Surface cover