COS 53-7 - Temporal dynamics in lake-ice breakup dates around the Northern Hemisphere from 1903 to 2003

Tuesday, August 7, 2012: 3:40 PM
B115, Oregon Convention Center

ABSTRACT WITHDRAWN

Sapna Sharma, Loyola University Chicago; John J. Magnuson, University of Wisconsin-Madison; Stephen R. Carpenter, University of Wisconsin - Madison

Background/Question/Methods

Ice break-up dates and other changes in lake-ice dynamics have occurred around the Northern Hemisphere in response to changing climatic conditions over the past 150 years.  We investigated the temporal patterns in annual ice break-up dates with 42 lakes around the Northern Hemisphere between 1903 and 2003. Our objectives were two-fold.  First, to disentangle the relative contribution of linear trends and cyclical patterns, and second, to identify the time-scales on which these processes influenced ice conditions.  We analyzed the contributions of long-term directional climate change and cyclical global climate dynamics, such as the North Atlantic Oscillation (NAO) and Southern Ocean Index (SOI).  A traditional and non-traditional approach to time-series analyses were used to model temporal variability in ice break-up dates: i) an autoregressive moving average (ARMA) model to predict the behaviour of ice break-up dates; and ii) a variation partitioning method in combination with a novel spectral analysis approach called Moran Eigenvector Maps (MEM) to disentangle the relative contribution of weather and climatic drivers on ice break-up dates and to identify the time-scales on which these processes are operating on ice conditions. 

Results/Conclusions

Over the past 100 years, lakes in the Northern Hemisphere are melting one to 24 days earlier.  Linear decreases in ice break-up dates attributed to longer-term climatic change, explain some variation in ice break-up dates through time.  A combination of global climate and weather patterns acting at a variety of scales (including the 2-3, 5, 11, and 34 year cyclic dynamics) corresponding to teleconnection patterns, such as the quasi-biennial, SOI, sunspot, and AMO cycles are explaining variation in ice dynamics. In addition, local weather patterns explain over 40% of variation in ice break-up dates through time. Therefore, we found that longer-term climatic change, in conjunction with local weather and global climate drivers acting at a variety of temporal scales explain ice dynamics over the past century.  Changes in regional and global meteorological patterns could result in continued earlier ice break-up dates leading to changes in aquatic ecosystems with socio-economic implications.