TY - JOUR

T1 - Pivotal role of snow depth, local atmospheric conditions, and large-scale climate signals on ice thinning in Finnish lakes

AU - Naderian, Danial

AU - Noori, Roohollah

AU - Bateni, Sayed M.

AU - Jun, Changhyun

AU - Kim, Dongkyun

AU - Shahmohammad, Mohsen

AU - Alizadeh, Farzaneh

AU - Kianmehr, Peiman

AU - Woolway, R. Iestyn

PY - 2025/2/25

Y1 - 2025/2/25

N2 - Lake ice thickness is a critical indicator of climate change, with profound implications for aquatic ecosystems. Despite its importance, our understanding of changes in lake ice thickness is relatively limited. Additionally, the influence of snow depth on ice thickness at a regional scale is not well understood, primarily due to the lack of long-term, high-resolution in-situ data. In this study, we leveraged in-situ data from 35 stations across 32 deep Finnish lakes, collected from 1995 to 2023, to investigate and simulate changes in maximum ice thickness using traditional statistical tools and artificial intelligent models. Our results indicate a decrease in maximum ice thickness at nearly all (34 out of 35) studied sites, with an average annual decrease of ∼3.4 mm/yr. The rate of decline was slower in lakes situated at higher latitude and at higher elevation. In the southern lakes of Finland, local atmospheric conditions were more effective in explaining the variance in ice thickness. However, at higher latitudes, the influence of these local factors decreased. Winter air temperature and snow depth were identified as the main drivers of maximum ice thickness. Snow depth was positively correlated with ice thickness (R = 0.63), largely due to its insulating effect, which supports faster ice growth in deeper lakes. Among the large-scale climate signals tested, the North Atlantic Oscillation and Scandinavia Pattern, as well as global CO2 concentrations, were key drivers of the developed ice thickness models. Our findings suggest that readily accessible large-scale climate signals can be used to inform modelling efforts of lake ice thickness.

AB - Lake ice thickness is a critical indicator of climate change, with profound implications for aquatic ecosystems. Despite its importance, our understanding of changes in lake ice thickness is relatively limited. Additionally, the influence of snow depth on ice thickness at a regional scale is not well understood, primarily due to the lack of long-term, high-resolution in-situ data. In this study, we leveraged in-situ data from 35 stations across 32 deep Finnish lakes, collected from 1995 to 2023, to investigate and simulate changes in maximum ice thickness using traditional statistical tools and artificial intelligent models. Our results indicate a decrease in maximum ice thickness at nearly all (34 out of 35) studied sites, with an average annual decrease of ∼3.4 mm/yr. The rate of decline was slower in lakes situated at higher latitude and at higher elevation. In the southern lakes of Finland, local atmospheric conditions were more effective in explaining the variance in ice thickness. However, at higher latitudes, the influence of these local factors decreased. Winter air temperature and snow depth were identified as the main drivers of maximum ice thickness. Snow depth was positively correlated with ice thickness (R = 0.63), largely due to its insulating effect, which supports faster ice growth in deeper lakes. Among the large-scale climate signals tested, the North Atlantic Oscillation and Scandinavia Pattern, as well as global CO2 concentrations, were key drivers of the developed ice thickness models. Our findings suggest that readily accessible large-scale climate signals can be used to inform modelling efforts of lake ice thickness.

KW - Large-scale climate signals

KW - Finnish lakes

KW - Ice thickness

KW - Random forest

U2 - 10.1016/j.scitotenv.2025.178715

DO - 10.1016/j.scitotenv.2025.178715

M3 - Article

VL - 966

JO - Science of the Total Environment

JF - Science of the Total Environment

SN - 0048-9697

ER -