Premium
The Role of Climate and Lake Size in Regulating the Ice Phenology of Boreal Lakes
Author(s) -
Higgins S. N.,
Desjardins C. M.,
Drouin H.,
Hrenchuk L. E.,
van der Sanden J. J.
Publication year - 2021
Publication title -
journal of geophysical research: biogeosciences
Language(s) - English
Resource type - Journals
eISSN - 2169-8961
pISSN - 2169-8953
DOI - 10.1029/2020jg005898
Subject(s) - climatology , phenology , physical geography , cryosphere , snow , boreal , environmental science , climate change , arctic ice pack , atmospheric sciences , sea ice , geology , geography , ecology , oceanography , meteorology , biology , paleontology
We quantified the relationships between seasonal air temperatures and ice phenology for a 0.54 km 2 boreal lake in Northwestern Ontario, Canada using a 50‐year time series (1970–2019). Significant temporal trends in the duration of ice cover (−4.0 days decade −1 ) occurred over the study period and both ice‐on and ice‐off dates were highly predictable from seasonal air temperatures. While temporal trends in ice‐off dates were not significant, ice‐on dates trended later by 2.2 days decade −1 , and both ice‐off dates and the duration of ice‐cover became increasingly variable over the study period. For three consecutive winter seasons, we also evaluated regional variations of ice‐phenology and snow and ice‐thickness from 9 to 30 boreal lakes across a lake size gradient (0.02–26 km 2 ) using ground based and satellite observations. Regional variation of ice‐on dates (30–45 days), ice‐off dates (10–21 days), and total duration of ice‐cover (22–38 days) between lakes displayed significant nonlinear relationships to lake size across all years. Regional variation of ice‐off dates was structured according to lake size, and appeared driven by relationships between lake size, snow thickness, and ice thickness. Multiple linear regression and generalized additive modeling approaches to predict ice‐on and ice‐off dates at the regional scale by integrating the effects of air temperature and lake size had similar performance, explaining >90% of the variance of observed.