Relationships between air temperature and ice conditions on the southern Baltic coastal lakes in the context of climate change

Shallow, lagoonal coastal lakes of the Southern Baltic are exceptionally susceptible to atmospheric factors. This work examines the influence of winter air temperatures on ice parameters (dates of first ice occurrence and last ice disappearance, ice season duration, number of days with ice, and annual maximum ice thickness) on Southern Baltic coastal lakes (Jamno, Bukowo, Gardno, Lebsko), and trends displayed by changes in these parameters over the period spanning 1960/61-2019/20. The research data was retrieved from the archives of Institute of Meteorology and Water Management - National Research Institute. As a first step of the analysis, we investigated the statistical relationships in spatial and temporal variations in winter air temperature and coastal lake ice parameters. Correlation and regression method was employed to determine the influence of air temperature on coastal lake ice conditions. Correlation coefficients were calculated, and linear regression equations were determined. The statistical significance of the observed relationships was assessed using Fisher-Snedecor test. Additionally, linear trend models were constructed. Our analysis indicates that from 1960/61 to 2019/20, the average rate of increase in winter temperature (December-March) equalled 0.04 degrees C-year-1. The correlation coefficients for air temperature versus ice parameters were highly statistically significant (alpha<0.001). The strongest relationships (with correlation coefficients below-0.90) occurred between air temperature and number of days with ice. Ice season duration and number of days with ice are both closely linked with last ice disappearance date and ice thickness. Our analysis shows that a 1 degrees C increase in average seasonal air temperature will result in the number of days with ice on the studied coastal lakes becoming reduced by 16-17 days. All trends in ice parameters indicate a mildening of ice conditions. The assessed trends are statistically significant, except for L (date of last ice disappearance) for Lake Jamno and H (annual maximum ice thickness) on lakes Jamno and Bukowo, at least at alpha=0.05 level. Their correlation coefficients are usually in the range from-0.35 to-0.50. The strongest trends were observed for ice season duration. Ice season duration shortens in an eastward direction by 0.66 day center dot year-1 (Lake Jamno) to 1.16 day center dot year-1 (Lake Lebsko). All correlation coefficients for ice trends were found to increase eastward, which could be explained by an increasing influence of the warming climate over the Southern Baltic in this direction. The strong relationships between air temperature and Southern Baltic coastal lake ice parameters, and the determined rate of changes may have a significance for forecasting, as the shifting dates of ice formation and disappearance on lakes are highly important for the lake hydrodynamics, and the functioning of aquatic ecosystems.