Search

Great Lakes ice cover has been regularly documented and reported only since the 1970s, while inconsistent records of ice cover began in the 1960s. These ice charts from the 1960s and early 1970s had yet to be digitized from physical prints and preliminary scans into a computer- and web-friendly format. Aerial surveys and satellite imagery of the ice cover started in the 1960s; there were very few records of Great Lakes ice cover before these surveys and ice charts began. Collection of weather data (temperature, wind, precipitation, etc.) began in the 1800s in the region, but ice cover reports were sparse and difficult to estimate. Using the surface air temperature data from 1897 to 1983, ice cover can be estimated and hindcasted back to the start of the weather record for the Great Lakes. Another atmospheric component that influences ice cover on the lakes is that of atmospheric teleconnections, such as the ENSO (El Niño Southern Oscillation), NAO (North Atlantic Oscillation), and newer ABNA (Asian-Bering-North American). While larger, more well-known teleconnections such as the ENSO and NAO have been analyzed next to Great Lakes ice cover, ABNA had yet to be compared to the lakes’ annual ice cover. These gaps in the collection of Great Lakes ice research were filled through this collaborative project between the University of Michigan School for Environment and Sustainability (SEAS) and NOAA’s Cooperative Institute for Great Lakes Research (CIGLR). The historical ice charts from 1963 to 1972 are now digitally available through the University of Michigan’s Deep Blue repository for this project; the Great Lakes’ ice cover has been hindcasted back to the winter of 1898, available here in Appendix II; and the ABNA has been recreated and statistically compared to the Great Lakes’ Annual Maximum Ice Cover (AMIC) values, and has proven to be a strong contender in forecasting and hindcasting ice cover on the Great Lakes. Great Lakes ice cover provides various ecosystem services in the Great Lakes, from tourism, to ice caves, to supporting the spawning of fish by protecting their eggs from wave action. At the same time, ice cover is a significant obstacle for winter navigation in the Great Lakes. Federal and commercial icebreaking operations are thus a vital aspect of wintertime shipping within the Great Lakes. Knowing ice conditions ahead of the winter season, or even with a longer lead time, is critical in planning in all these sectors. However, seasonal or longer forecasting of Great Lakes ice conditions has been challenging because of the strong year-to-year fluctuations in AMIC and little knowledge in what teleconnection pattern(s) influence the weather across the Great Lakes region. The deliverables of our project – the extended time series of AMIC, and identification of the ABNA as an important teleconnection pattern – well address these needs and provide a strong foundation to improve seasonal Great Lakes ice forecasting at NOAA Great Lakes Environmental Research Laboratory, the client of this project.

Abstract: Lake and sea ice and their snow covers are major components of Earth’s cryosphere and act to strongly modify climatic and biological systems. Both ice types serve as an important habitat for micro-fauna and support macro-fauna and strongly modify the exchange of energy, gases, and momentum between the atmosphere and the underlying water. Observations of areal properties of ice and snow have been widely available for the past 40 years, allowing for monitoring of the long term decreases in e.g. ice extent, duration and age. Unfortunately, observations of other ice properties remain limited, due to reductions in monitoring programs, operational and logistical limitations such as cost and remoteness and technological limitations such as the absorption of EM radiation by water. Laser and radar altimeters can provide measurements of the surface properties of the ice cover and snow covers such as roughness and thickness but have traditionally suffered from the same issues of poor spatial resolution or coverage. In 2010, The European Space Agency launched CryoSat-2 carrying SIRAL, a pulse-limited Delay-Doppler Synthetic Aperture and Interferometric radar altimeter with improved spatial coverage and along-track resolution and across-track position information using dual antennae. We provide several analyses of CryoSat-2 radar altimeter data over sea ice and lake ice using on-ice manual and autonomous measurements, airborne laser and radar altimetry and electromagnetic sounding and satellite radar imagery. First, a method to produce airborne roughness from a laser scanner without complicated inertial navigation data is presented and shown to better represent the roughness of sea ice and to better highlight differences in roughness between different types of ice, smooth first-year ice (FYI) north of Svalbard, rough drifting multi-year ice (MYI) and FYI in Fram Strait and very rough landfast ice off the eastern coast of Greenland than a single-beam laser altimeter. Second, the first retrievals of lake ice thickness using CryoSat-2, or any radar altimeter, are presented. The retrievals are able to reproduce the seasonal development in ice thickness observed by drill holes and model data (r > 0.65) and with low uncertainty (RMSE < 0.33 m) given that the minimum thickness that can be sensed is 0.26 m. Three studies into the response of CryoSat-2 over sea ice are presented using comparisons with ground and airborne data in the Arctic and Antarctica. The first of these studies highlighted errors with the early versions of CryoSat-2 data. The modal laser and radar freeboards and their latitudinal gradients were found to be the same despite the fact that the CryoSat-2 radar freeboard is assumed to come from the ice surface and the laser freeboard is representative of the snow surface. The agreement between the two freeboard measures may indicate incomplete radar penetration into the snow and the complicating issue of surface roughness and differences in spatial coverage. The second CryoSat-2 study highlighted the influence of snow depth on the retracking of CryoSat-2 radar echoes and the resulting freeboard results over sea ice with thick and thin snow covers. Lastly, the third study compares all available coincident CryoSat-2 and airborne laser and radar altimeter, and EM sounding data collected between 2010 and 2014. Results show low correlations (

The Arctic Coastal Plain of northern Alaska is covered with thousands of lakes developed atop continuous permafrost. Lake ice meltout patterns near Barrow were evaluated across the historical record using available cloud-free Landsat imagery. An Iterative Self-Organizing Data clustering algorithm was used to differentiate ice from water during meltout, and was overlain on lake boundary vectors to calculate the percentage of ice covering each lake in the ~11,000 km2 study area near Barrow. This processing technique was carried out for seven available scenes in early July across the ~35-year Landsat record. Analysis of these scenes reveals that lakes exhibit interannual variability in the general timing of lake ice meltout. A consistent regional spatial pattern is apparent in all years, with a general decrease in the percentage of ice coverage on lakes further south reflecting the regional climatic pattern. The ice cover on many lakes near the coast persists for a longer period due to the influence of onshore winds and cooler temperatures, possibly enhanced by coastal fog. Some lakes consistently experience earlier or delayed meltout compared to lakes nearby owing to lake-specific factors that may include lake depth, basin geometry, or water salinity. This study illustrates the utility of satellite-derived lake ice observations as a regional climatic indicator.

Research on lake ice permeability in the McMurdo Dry Valleys, Antarctica required the design and construction of a novel multi-support, field based, gas permeameter. The permeameters simple design allows rapid, precise, and non-destructive permeability measurements in a harsh polar environment. The research goals using the multi-support gas permeameter (MSGP) are to (1) determine lake ice physical hydrology through the investigation of primary permeability, (2) how permeability affects the mechanics and physics of fluid and gas transport through lake ice, (3) describe any variable nature of lake ice permeability, (4) scale up permeability to different sections of the lakes, and (5) document the change in permeability during the austral spring. The multi-support gas permeameter is of original design, however, a tip seal system utilizing the developments of a laboratory-based permeameter by Tidwell and Wilson (1997) was incorporated. The tip seal design has five interchangeable components, with inner seal radii of 0.15cm, 0.31cm, 0.63cm, 1.27cm, and 2.54cm, and outer radii twice the inner, which allows multiple gas diffusion depths, thus permitting permeability measurements over a wide variety of ice thickness and formation. The system delivers a positive pressure of 0 to 100kPa of compressed laboratory grade nitrogen (N2) gas into the tip seal interface that is in direct contact with the ice being tested. The volumetric flow rate, absolute injection pressure, barometric pressure, and temperature of the N2 gas penetrating the ice are recorded manually from a digital readout. The permeability is then calculated using a modified version of Darcys Law. Spot primary permeability measurements were used to scale permeability readings up to the entire lakes surface areas. The two lakes are influenced by different environmental conditions including source of wind (continental or marine influence), amount of sunlight (ablation rate), and ice conditions (annual and perennial ice lasting approximately 7 to 10 years). Measurements were made on a small scale (nine sample points per 81cm2 area) and were repeated weekly over a four-week period from early to late austral spring. By understanding the heterogeneity in smallest of scales at the different sampling sites (and through any temporal changes), the lakes’ permeability heterogeneity could be factored out and scaled up to reflect the primary permeability of the entire lake as a single, homogenous system. Data analysis revealed little difference in surficial primary permeability over a large spatial area, regardless of ice type sampled or the environmental conditions encountered. Although slight trends of permeability variability were noticed, the ice is intrinsically semi-permeable on the primary level. Average primary permeability of void spaces between ice crystals was at approximately 10-13 to 10-14 m2 for all in situ measurements. Observation revealed that although there was little difference in primary permeability of lake ice in general, it is very permeable due to secondary structures (fractures, bubbles, and lenses) contained within the ice column. This study dealt exclusively with the primary permeability and could not quantitatively constrain the secondary, and therefore, the overall permeability of the lakes.