Reconstructing Great Lakes air temperature and ice dynamics data back to 1897.

Credible modeling, tools, and guidance, regarding the changing Laurentian Great Lakes and the climatic impacts, are needed by local decision-makers to inform their management and planning. The present study addresses this need through a model evaluation study of the representation of lake–atmosphere interactions and resulting lake-effect snowfall in the Great Lakes region. Analysis focuses on an extensive ensemble of 74 historical simulations generated by 23 high-resolution global climate models (GCMs) from the High-Resolution Model Intercomparison Project (HighResMIP). The model assessment addresses the modeling treatment of the Great Lakes, the spatial distribution and seasonality of climatological snowfall, the seasonal cycle of lake-surface temperatures and overlake turbulent fluxes, and the lake-effect ratio between upwind and downwind precipitation. A deeper understanding of model performance and biases is achieved by partitioning results between HighResMIP GCMs that are 1) coupled to 1D lake models versus GCMs that exclude lake models, 2) between prescribed-ocean model configurations versus fully coupled configurations, and 3) between deep Lake Superior versus relatively shallow Lake Erie. While the HighResMIP GCMs represent the Great Lakes by a spectrum of approaches that include land grid cells, ocean grid cells (with lake surface temperature and ice cover boundary conditions provided by the Met Office Hadley Center Sea Ice and Sea Surface Temperature Dataset), and 1D lake models, the current investigation demonstrates that none of these rudimentary approaches adequately represent the complex nature of seasonal lake temperature and ice cover evolution and its impact on lake–atmosphere interactions and lake-effect precipitation in the Great Lakes region. Significance Statement The purpose of this study is to evaluate the capability of high-resolution global climate models to simulate lake–atmosphere interactions and lake-effect snowfall in the Great Lakes region, given the critical influence of the lakes on regional climate and vast societal and environmental impacts of lake-effect snowfall. It is determined that the models inadequately represent lake temperatures and ice cover, often leading to insufficient annual snowfall in the lake-effect zones. More advanced, three-dimensional lake models need to be coupled to climate models to support greater credibility in regional lake and climate simulations and future climate projections.

Seasonal ice cover is one of the important attributes of lakes at middle and high latitude regions. The annual freeze-up and break-up dates and the durations of ice cover (i.e., lake ice phenology) are sensitive to the weather and climate, and hence can be used as an indicator of climate variability and change. The Calibrated Enhanced Resolution Brightness Temperature (CETB) dataset available from the National Snow and Ice Data Center (NSIDC) provides an alternate source of passive microwave brightness temperature (TB) measurements for the determination of lake ice phenology on a 3.125 km grid. This study used Scanning Multi-channel Microwave Radiometer (SMMR), Special Sensor Microwave Image (SSM/I) and Special Sensor Microwave Imager/Sounder (SSMIS) data from the CETB dataset to extract the ice phenology for 56 lakes across the Northern Hemisphere from 1979 to 2019. According to the differences in TB between lake ice and open water, a threshold algorithm based on the moving t test method was applied to determine the lake ice status for grids located at least 6.25 km away from the lake shore, and the ice phenology dates for each lake were then extracted. When ice phenology could be extracted from more than one satellite over overlapping periods, results from the satellite offering the largest number of observations were prioritized. The lake ice phenology results showed strong agreement with an existing product derived from Advanced Microwave Scanning Radiometer for EOS (AMSR-E) and Advanced Microwave Scanning Radiometer 2 (AMSR-2) data (2002 to 2015), with mean absolute errors of ice dates ranging from 2 to 4 days. Compared to near-shore in-situ observations, the lake ice results, while different in terms of spatial coverage, still showed overall consistencies. The produced lake ice record also displayed significant consistencies when compared to a historical record of annual maximum ice cover of the Laurentian Great Lakes of North America. From 1979 to 2019, the average complete freezing duration and ice cover duration for lakes forming a complete ice cover on an annual basis were 153 and 161 days, respectively. The lake ice phenology dataset – a new climate data record (CDR) – will provide valuable information to the user community about the changing ice cover of lakes in the last four decades. The dataset is available at https://www.pangaea.de/tok/c8fc0eab3d30777fc38979ad514217b6b7e86a65 (Cai et al., 2021).

It has been found that large lakes, a common component of the North American (NA) landscape, can affect the water cycle and modulate temperatures in the surrounding regions. The purpose of this study is to evaluate different lake representations in regional climate model simulations. We use the Weather Research and Forecasting (WRF) model, a widely-used regional climate model, forced by the ERA5 reanalysis product. The study is performed for a 40 year historic period (1979-2019) at a resolution of 12 km. The lakes of concern include the Laurentian Great Lakes, which straddle the US-Canada border; the Great Slave and Great Bear Lakes of the Northwest Territories; and the Lakes Winnipeg and Winnipegosis. Alongside the default lake model, two new column lake models are employed: FLake, a more widely used model, and GL25, a recent, physics-based model. These models have been somewhat successful at alleviating inadequacies of the default model by introducing additional process representations, such as a more realistic surface albedo formulation and a better parameterization for vertical overturning. Additionally, we consider the effect of vertical eddy diffusivity and lake stratification on the model performance. While no column lake model is expected to perform perfectly, our goal here is to identify when (seasonally) and where (geographically) a model produces reliable results, including ice cover distribution and near-surface temperature. The effect of the lake representation on the surrounding regions (e.g., lake-effect precipitation) is also evaluated. We find that the two new lake models perform reasonably well, but there are significant differences in seasonal biases with GL25 performing better in summer and FLake performing better in winter; in fact, FLake reproduces the ice cover of the Great Lakes very well. Furthermore, the biases do not seem to be affected by surface wind induced circulation, and hence 3D modelling may not be a requirement for lake modelling. This study can help with the selection of lake models for regional climate modelling in the NA region and inform the interpretation of the predictions.

The impacts of North Atlantic Oscillation (NAO) and El Niño–Southern Oscillation (ENSO) on Great Lakes ice cover were investigated using lake ice observations for winters 1963–2010 and National Centers for Environmental Prediction reanalysis data. It is found that both NAO and ENSO have impacts on Great Lakes ice cover. The Great Lakes tend to have lower (higher) ice cover during the positive (negative) NAO. El Niño events are often associated with lower ice cover. The influence of La Niña on Great Lakes ice cover is intensity‐dependent: strong (weak ) La Niña events are often associated with lower (higher) ice cover. The interference of impacts of ENSO and NAO complicates the relationship between ice cover and either of them. The nonlinear effects of ENSO on Great Lakes ice cover are important in addition to NAO effects. The correlation coefficient between the quadratic Nino3.4 index and ice cover (−0.48) becomes significant at the 99% confidence level. The nonlinear response of Great Lakes ice cover to ENSO is mainly due to the phase shift of the teleconnection patterns during the opposite phases of ENSO. Multiple‐variable nonlinear regression models were developed for ice coverage. Using the quadratic Nino3.4 index instead of the index itself can significantly improve the prediction of Great Lakes ice cover (the correlation between the modeled and observed increases from 0.35 to 0.51). Including the interactive term NAO·Nino3.42 further improves the prediction skill (the correlation increases from 0.51 to 0.59). The analysis is also applied to individual lakes. The model for Lake Michigan has the highest prediction skill, while Lake Erie has the smallest skill.

For centuries the Great Lakes have been treated callously These five magnificent lakes—Superior, Michigan, Huron, Erie, and Ontario—located along the eastern half of the Canadian–U.S. border have served as a virtual sewer catching waste from industry, agriculture, commercial shipping, and households. Their natural barriers to other water systems have been breached, exposing indigenous ecosystems to aggressive invaders. They’ve been used as a highway for colossal ships that require deepened and broadened channels to crisscross the lakes, and that import exotic species along with their intended cargo. At times it could seem that this long-suffering water system will see no end of indignities. But recent renewed focus on the unique and tremendous value of this resource by governments and communities surrounding the lakes may turn the tide of neglect and abuse. According to the U.S. Environmental Protection Agency (EPA), the Great Lakes contain 21% of the Earth’s and about 84% of United States’ surface freshwater. That’s about 22,000 cubic kilometers of water spread over 94,250 square miles. Each year the lakes provide more than 6.7 million cubic meters of water to municipalities and quadruple that to industry. They support a commercial fishery worth about $13 million as of 2002, according to the U.S. Geological Survey (USGS), and a sport-fishing industry of nearly $1.3 billion as of 2001, according to the U.S. Fish and Wildlife Service. Today about 25% of Canadians and 10% of Americans—a total of more than 33 million people—live in the Great Lakes watershed. “The whole industrial expansion that took place during the ‘robber baron’ era [of the late nineteenth century] expanded along the shores of the Great Lakes,” says Deborah Swackhamer, a University of Minnesota professor of environmental chemistry. Soon ships were carrying iron ore, coal, and limestone from mines and quarries to steel mills and later steel to factories and products to markets. In addition to serving as a transportation system, the lakes provided a place to discharge manufacturing by-products. Unlike a sewer, however, whatever enters this lake system stays awhile. On average, less than 1% of the five lakes’ water turns over each year, which means that many pollutants stay in place. They settle in sediments, adhere to other surfaces, become suspended in water, and bioaccumulate in organisms. Similarly, with the exception of migratory birds, most wildlife in the basin spend their entire life cycle in or near the lakes. As a result of all these stressors, the lakes now house fish that are dangerous to eat, water that can be unsafe to drink, anoxic “dead zones”—areas in which virtually no plants or animals can survive—that appear each summer like clockwork, and an ever-growing population of unwanted species from other parts of the world.

A review: Building the resilience of Great Lakes beneficial uses to climate change

A 20-km regional climate model, the Abdus Salam International Centre for Theoretical Physics Regional Climate Model version 4 (ICTP RegCM4), is employed to investigate heavy lake-effect snowfall (HLES) over the Great Lakes Basin and the role of ice cover in regulating these events. When coupled to a lake model and driven with atmospheric reanalysis data between 1976 and 2002, RegCM4 reproduces the major characteristics of HLES. The influence of lake ice cover on HLES is investigated through 10 case studies (2 per Great Lake), in which a simulated heavy lake-effect event is compared with a companion simulation having 100% ice cover imposed on one or all of the Great Lakes. These experiments quantify the impact of ice cover on downstream snowfall and demonstrate that Lake Superior has the strongest, most widespread influence on heavy snowfall and Lake Ontario the least. Ice cover strongly affects a wide range of atmospheric variables above and downstream of lakes during HLES, including snowfall, surface energy fluxes, wind speed, temperature, moisture, clouds, and air pressure. Averaged among the 10 events, complete ice coverage causes major reductions in lake-effect snowfall (>80%) and turbulent heat fluxes over the lakes (>90%), less low cloudiness, lower temperatures, and higher air pressure. Another important consequence is a consistent weakening (30%–40%) of lower-tropospheric winds over the lakes when completely frozen. This momentum reduction further decreases over-lake evaporation and weakens downstream wind convergence, thus mitigating lake-effect snowfall. This finding suggests a secondary, dynamical mechanism by which ice cover affects downstream snowfall during HLES events, in addition to the more widely recognized thermodynamic influence.

Arp, C.D., B.M. Jones, M. Whitman, A. Larsen, and F.E. Urban, 2010. Lake Temperature and Ice Cover Regimes in the Alaskan Subarctic and Arctic: Integrated Monitoring, Remote Sensing, and Modeling. Journal of the American Water Resources Association (JAWRA) 46(4): 777‐791. DOI: 10.1111/j.1752‐1688.2010.00451.xAbstract: Lake surface regimes are fundamental attributes of lake ecosystems and their interaction with the land and atmosphere. High latitudes may be particularly sensitive to climate change, however, adequate baselines for these lakes are often lacking. In this study, we couple monitoring, remote sensing, and modeling techniques to generate baseline datasets of lake surface temperature and ice cover in the Alaskan Subarctic and Arctic. No detectable trends were observed during this study period, but a number of interesting patterns were noted among lakes and between regions. The largest Arctic lake was relatively unresponsive to air temperature, while the largest Subarctic lake was very responsive likely because it is fed by glacial runoff. Mean late summer water temperatures were higher than air temperatures with differences ranging from 1.7 to 5.4°C in Subarctic lakes and from 2.4 to 3.2°C in Arctic lakes. The warmest mean summer water temperature in both regions was in 2004, with the exception of Subarctic glacially fed lake that was highest in 2005. Ice‐out timing had high coherence within regions and years, typically occurring in late May in Subarctic and in early‐July in Arctic lakes. Ice‐on timing was more dependent on lake size and depth, often varying among lakes within a region. Such analyses provide an important baseline of lake surface regimes at a time when there is increasing interest in high‐latitude water ecosystems and resources during an uncertain climate future.

Abstract. Warming trends in the Laurentian Great Lakes and surrounding areas have been observed in recent decades, and concerns continue to rise about the pace and pattern of future climate change over the world's largest freshwater system. To date, most regional climate models used for Great Lakes projections either neglected the lake-atmosphere interactions or are only coupled with a 1-D column lake model to represent the lake hydrodynamics. This study presents a Great Lakes climate change projection that has employed the two-way coupling of a regional climate model with a 3-D lake model (GLARM) to resolve 3-D hydrodynamics essential for large lakes. Using the three carefully selected Coupled Model Intercomparison Project Phase 5 (CMIP5) general circulation models (GCMs), we show that the GLARM ensemble average substantially reduces surface air temperature and precipitation biases of the driving GCM ensemble average in present-day climate simulations. The improvements are not only displayed from an atmospheric perspective but are also evident in the accurate simulations of lake temperature and ice coverage. We further present the GLARM projected climate change for the mid-21st century (2030–2049) and the late 21st century (2080–2099) in the Representative Concentration Pathway (RCP) 4.5 and RCP 8.5 scenarios. Under RCP 8.5, the Great Lakes basin is projected to warm by 1.3–2.1 ∘C by the mid-21st century and 4.1–5.0 ∘C by the end of the century relative to the early century (2000–2019). Moderate mitigation (RCP 4.5) reduces the mid-century warming to 0.8–1.8 ∘C and late-century warming to 1.8–2.7 ∘C. Annual precipitation in GLARM is projected to increase for the entire basin, varying from 0 % to 13 % during the mid-century and from 9 % to 32 % during the late century in different scenarios and simulations. The most significant increases are projected in spring and fall when current precipitation is highest and a minimal increase in winter when it is lowest. Lake surface temperatures (LSTs) are also projected to increase across the five lakes in all of the simulations, but with strong seasonal and spatial variability. The most significant LST increases occur in Lakes Superior and Ontario. The strongest warming is projected in spring that persists into the summer, resulting from earlier and more intense stratification in the future. In addition, diminishing winter stratification in the future suggests the transition from dimictic lakes to monomictic lakes by the end of the century. In contrast, a relatively smaller increase in LSTs during fall and winter is projected with heat transfer to the deep water due to the strong mixing and energy required for ice melting. Correspondingly, the highest monthly mean ice cover is projected to reduce to 3 %–15 % and 10 %–40 % across the lakes by the end of the century in RCP 8.5 and RCP 4.5, respectively. In the coastal regions, ice duration is projected to decrease by up to 60 d.

Climate warming is especially pronounced in winter and at high latitudes. Warming winters are leading to the loss of lake ice and changing snow cover on lakes. Historically, lake scientists have paid less attention to the ice cover period, leading to data and theory gaps about the role of winter conditions in lake ecosystem function and the consequences of changing winters. Here we use simple models to show that the latitudinal interaction between ice cover duration and light flux seasonality has profound and underappreciated implications for lakes. Our models focus on light and temperature, two key drivers of ecosystem processes. We show that the relative amount of light arriving in lakes during ice cover increases non-linearly with latitude and that the light climate of high latitude lakes is much more sensitive to changing winter conditions than that of lower latitude lakes. We also demonstrate that the synchronicity between high light and warm temperatures may decrease with latitude, with implications for primary and secondary production. Our results suggest that ice loss may lead to greater relative change to productivity and biotic interactions in higher latitude lakes and also offer several testable predictions for understanding the consequences of climate-induced changes across latitudinal gradients.

It is generally understood that extensive regions of significant lake ice cover impact lake-effect (LE) snow storms by decreasing the upward heat and moisture fluxes from the lake surface; however, it is only recently that studies have been conducted to more thoroughly examine this relationship. This study provides the first examination of Great Lakes LE snow storms that developed in association with an extensively ice-covered lake. The LE snow events that occurred downwind of Lake Erie on 12–14 February 2003 and 28–31 January 2004 produced maximum snowfall totals of 43 and 64 cm in western New York state, respectively. The presence of widespread ice cover led these snows to be less anticipated than snowfalls from Lake Ontario, which had limited ice cover. For both events, a variety of ice-cover conditions and meso- and synoptic-scale factors (i) helped support LE snow storm development, (ii) lead to the transitions in LE convective type, and (iii) resulted in noteworthy snowfalls near Lake Erie. Thinner ice cover along with favorable fetch directions during the 2004 event likely aided the development of more significant snowband time periods and the resulting greater snowfall. Although Lake Erie had regions with lower ice concentration during the 2003 event, thicker ice cover was present across a greater area of the lake, fetch directions during lake-effect time periods were positioned over higher ice concentration regions, and snowbands had a shorter duration and impacted the same region to a lesser degree than the 2004 case.

The Weather Research and Forecasting (WRF) Model is used to dynamically downscale ERA-Interim global reanalysis data to test its performance as a regional climate model (RCM) for the Great Lakes region (GLR). Four cumulus parameterizations and three spectral nudging techniques applied to moisture are evaluated based on 2-m temperature and precipitation accumulation in the Great Lakes drainage basin (GLDB). Results are compared to a control simulation without spectral nudging, and additional analysis is presented showing the contribution of each nudged variable to temperature, moisture, and precipitation. All but one of the RCM test simulations have a dry precipitation bias in the warm months, and the only simulation with a wet bias also has the least precipitation error. It is found that the inclusion of spectral nudging of temperature dramatically improves a cold-season cold bias, and while the nudging of moisture improves simulated annual and diurnal temperature ranges, its impact on precipitation is complicated. Significance Statement Global climate models are vital to understanding our changing climate. While many include a coarse representation of the Great Lakes, they lack the resolution to represent effects like lake effect precipitation, lake breeze, and surface air temperature modification. Therefore, using a regional climate model to downscale global data is imperative to correctly simulate the land–lake–atmosphere feedbacks that contribute to regional climate. Modeling precipitation is particularly important because it plays a direct role in the Great Lakes’ water cycle. The purpose of this study is to identify the configuration of the Weather Research and Forecasting Model that best simulates precipitation and temperature in the Great Lakes region by testing cumulus parameterizations and methods of nudging the regional model toward the global model.

Thermodynamic processes of ice in three lakes and landfast ice around Zhongshan Station, Antarc- tica, were observed in 2006. The mass balance of lake ice was compared with that of landfast ice. The responses of lake ice and sea ice temperatures to the local surface air temperature are explored. Vertical conductive heat fluxes at varying depths of lake ice and sea ice were derived from vertical temperature profiles. The freeze up of lake ice and landfast ice occurred from late February to early March. Maximum lake ice thicknesses occurred from late September to early October, with values of 156–177 cm. The maximum sea ice thicknesses of 167–174 cm occurred relatively later, from late October to late November. Temporal variations of lake ice and landfast ice internal temperatures lagged those of air temperatures. High-frequency variations of air temperature were evidently attenuated by ice cover. The temporal lag and the high-frequency attenuation were greater for sea ice than for lake ice, and more distinct for the deeper ice layer than for the upper ice layer. This induced a smaller conductive heat flux through sea ice than lake ice, at the same depth and under the same atmospheric forcing, and a smoother fluctuation in the conductive heat flux for the deeper ice layer than for the upper ice layer. Enhanced desalination during the melt season increased the melting point temperature within sea ice, making it different from fresh lake ice.

Timely information on lake ice cover conditions is critical in support of commercial shipping, winter-road transportation, and winter leisure activities such as snowmobiling and ice fishing. Monitoring of lake ice extent and ice phenology (i.e. dates associated with freeze-up/break-up and ice duration) is also valuable for improving numerical weather prediction (NWP) and for climate monitoring. The availability of free synthetic aperture radar (SAR) data from the European Space Agency’s Sentinel-1 A/B constellation provides an unprecedented opportunity to develop operational algorithms for lake ice cover mapping at a temporal frequency not available until now from SAR missions (ca. every 1–5 days depending on latitude). For NWP, rapid and accurate mapping of ice cover and open water areas in lake-rich regions is required. However, the classification of SAR imagery using traditional machine learning (ML) approaches is challenging due to the large data volume generated by imaging systems such as Sentinel-1 SAR as well as the complexity of radar signatures as a function of sensor and target characteristics. For lakes specifically, radar backscatter can vary greatly with incidence angle, polarization and surface properties (e.g. calm and wind-roughed open water, new thin ice, surface roughness due to the presence of pressure ridges, ice type, melting of ice and on-ice snow). To address the challenge of large-scale lake ice mapping, we investigate the GPU-boosted deep neural network approach that is efficient at handling big complex data. In this paper, we design a novel maximum a posterior (MAP) approach combining a convolutional neural network (CNN) and conditional random field (CRF) for better addressing the challenges of operational lake ice mapping. The proposed approach is tested on 17 Sentinel-1 dual-polarization (VV and VH) SAR images, where eight are used for training and validation and nine used for out-of-sample testing. To identify the optimal network architecture, an independent validation set is used to evaluate the performance of a series of six CNNs with increasing model complexity. The best model overcomes the noise effect in Sentinel-1 SAR imagery and the lake ice signature ambiguity issue; it achieves average classification accuracies of 97.10% and 97.14% for open water and ice, respectively, on the validation set. Moreover, the best model outperforms the last-ranked model by about 2% in terms of mean overall accuracy (OA), demonstrating the improvement and importance of model selection. The use of CRF can consistently improve the CNN by reducing the artefacts and noise effect in ice maps, outperforming the CNN model when used alone by about 3% in terms of mean OA on the validation set. The proposed CNN-CRF approach also achieves high accuracy on the nine test scenes, achieving a mean OA of 93.10%, demonstrating strong generalization capability that is important for SAR lake ice mapping.

Coherence between atmospheric teleconnections, Great Lakes water levels, and regional climate