Trends, variability, and extremes in lake ice dates since 1855

Poster Disciplines/Format:
Poster Number: 
305
Presenter/Primary Author: 
Barbara Benson
Co-Authors: 
John J. Magnuson
Co-Authors: 
Olaf P. Jensen

Lake ice seasonality (freeze day, breakup day, and ice duration) is closely related to climate change and variability. Trends and changes in variability and extreme events are apparent. The shared variance between mean anomalies of ice duration and Northern Hemisphere land air temperature over the last 150 years is r2 = 0.47. Mean trends are in the direction of warming. Ice duration has the steepest decline with mean trends (days/decade) of 1.7 over the last 150 years, 1.1 over the last 100 years, and 6 over the last 30 years. Over all three periods, mean freeze dates are later (1.1 days/decade in 150 years, 0.3 in 100, 1.6 in 30) and breakup dates earlier (0.9 in 150, 0.5 in 100, 1.9 in 30). Differences in trends for ice dates exist among lakes, regions, and years analyzed. Numbers of lakes included in the analyses ranged from 8 (150-year duration) to 71 (100-year breakup).

For variability, trends were analyzed over 100- and 150-year periods in 10-year and 50-year windows. Variability in 10-year windows increased in the 150-year period only for freeze dates. For the 100-year period variability in 10-year windows decreased for the freeze and breakup dates. Variability in 50-year windows increased over the 150 years only for freeze dates between the first two 50 year windows. In the 100-year data set, mean variability in 50-year windows decreased for the Northern Hemisphere, and differences were apparent for specific regions. In Europe mean variability significantly decreased for freeze and breakup and in northeastern North America for duration, while in north central North America variability significantly increased for breakup.

The most extreme ice event would be a year without complete ice cover; these events have increased significantly over the last 100 years. In the past 150 years the odds of an extreme early freeze decreased, the odds of an extreme early ice breakup increased, and the odds of an extremely late breakup decreased across the Northern Hemisphere. In the past 100 years changes in odds are less consistent.

Trends in ice dates are strong and largely consistent around the Northern Hemisphere. In contrast, changes in variability and extremes are less consistent among ice-date measures and among regions, suggesting that more regional influences such as those of large-scale climate drivers affect patterns of variability and extremes.