Freeze-up Period Research Articles

Freeze/Thaw Detection and Validation Using Aquarius’ L-Band Backscattering Data

The seasonal cycle of landscape freeze/thaw (FT) state across mid- to high latitudes influences critical processes such as the land surface energy balance, carbon cycle dynamics related to vegetation growth, and hydrological partitioning between surface runoff and infiltration. In this paper, we produce the first daily FT classification for the 2011–2014 period based on L-band radar measurements from Aquarius. The radar FT algorithm used in this paper is based on a seasonal threshold approach, which is also the baseline algorithm applied to higher-resolution (3 km) radar measurements from NASA’s Soil Moisture Active/Passive (SMAP) mission (Launched January 31, 2015). The lower frequency (L-band) radar backscatter measurements from Aquarius provide enhanced sensitivity to FT conditions in vegetation canopy, snow and surface soil layers, although the relative radar penetration depth and sensitivity of the FT signal to these landscape elements will vary according to surface moisture and vegetation biomass conditions, and underlying land cover and terrain heterogeneity [1] , [2] . Evaluation of the seasonal threshold FT algorithm using Aquarius was performed using surface air and soil temperatures from selected stations in the Snow Telemetry (SnoTel) network. Analysis identified good agreement during the fall freeze-up period with flag agreement exceeding the 80% SMAP accuracy target when summarized on a monthly basis. Disagreement was greater during the spring thaw transition due in part to uncertainty in characterizing thaw from in situ measurements. Unlike the fall season, stronger agreement in the spring was identified when the reference state was characterized with air temperature compared to soil temperature.

Temporal evolution of Hudson Bay Sea Ice (1971–2011)

Previous work has found Hudson Bay seasonal sea ice particularly sensitive to climate change with a strong signal of earlier breakup dates. This work extends the previous analysis by including eight additional years of recent sea ice data. The expanded sea ice record, 1971 to 2011, revealed stronger and more statistically significant trends than the earlier work, most strikingly for the later freeze up. The average magnitude of the temporal trend for all 36 locations studied is 0.50 days/year for earlier breakup, 0.46 days/year for the later freeze-up, and 0.91 days/year for longer ice-free period. Of the 36 points, 12 points for the breakup period, 30 points for the freeze-up period, and 22 points for the ice-free season have accelerating temporal trends during the past decade.

Monitoring the Variation in Ice-Cover Characteristics of the Slave River, Canada Using RADARSAT-2 Data—A Case Study

The winter regime of river-ice covers in high northern latitude regions is often a determining factor in the management of water resources, conservation of aquatic ecosystems and preservation of traditional and cultural lifestyles of local peoples. As ground-based monitoring of river-ice regimes in high northern latitudes is expensive and restricted to a few locations due to limited accessibility to most places along rivers from shorelines, remote sensing techniques are a suitable approach for monitoring. This study developed a RADARSAT-2 based method to monitor the spatio-temporal variation of ice covers, as well as ice types during the freeze-up period, along the main channel of the Slave River Delta in the Northwest Territories of Canada. The spatio-temporal variation of ice covers along the river was analyzed using the backscatter-based coefficient of variation (CV) in the 2013–2014 and 2014–2015 winters. As a consequence of weather and flow conditions, the ice cover in the 2013–2014 winter had the higher variation than the 2014–2015 winter, particularly in the potential areas of flooded/cracked ice covers. The river sections near active channels (e.g., Middle Channel and Nagle Channel), Big Eddy, and Great Slave Lake also yielded higher intra-annual variation of ice cover characteristics during the winters. With the inclusion of backscatter and texture analysis from RADARSAT-2 data, four water and ice cover classes consisting of open water, thermal ice, juxtaposed ice, and consolidated ice, were discriminated in the images acquired between November and March in both the studied winters. In addition to river geomorphology and climatic conditions such as river width, sinuosity or air temperature, the fluctuation of water flows during the winter has a significant impact on the variation of ice cover as well as the formation of different ice types in the Slave River. The RADARSAT-2 based monitoring algorithm can also be applied to other river systems in high latitude ecosystems to annually monitor their river-ice variation and formation during the freeze-up and ice cover progression period.

Thin Ice Detection in the Barents and Kara Seas With AMSR-E and SSMIS Radiometer Data

We have studied thin ice detection using Advanced Microwave Scanning Radiometer-Earth Observing System (AMSR-E) and Special Sensor Microwave Imager/Sounder radiometer data acquired over the Barents and Kara Seas during three winters (November-April) in 2008-2011. Moderate Resolution Imaging Spectroradiometer-based ice thickness charts were used as reference data. Thin ice detection was studied using polarization and spectral gradient ratios (PR and GR) calculated from the 36/37 and 89/91 GHz radiometer data. Thresholds for thin ice detection and maximum thicknesses for the detected thin ice (hT) were determined, as were error rates for misdetections. The results for different 1-D PR and GR parameters led to the conclusion that the AMSR-E PR36 and H-polarized GR8936 would be the best parameters for a 2-D classifier. We adopted the linear discrimination analysis (LDA) as a statistical tool. Thin ice areas with hT of 30 cm could be separated from thicker ice fields with approximately 20% error level. In our large data set, the estimation of thin ice thickness was not possible with reasonable accuracy due to the large scatter between ice thickness and the PR and GR signatures. This is likely due to a large data set, besides thin ice in polynyas also thin ice in the marginal ice zone and thin ice from freeze-up period. The optimal LDA parameters in the classifier and hT depended on the daily mean air temperature ((T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">am</sub> )). We could not yet parameterize the classifier optimally according to (T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">am</sub> ), but the constructed classifier worked rather robustly as indicated by the relative small error rate variation between the three analyzed winters.

Monitoring the freeze-up and ice cover progression of the Slave River

River ice is an important component to maintain traditional and cultural lifestyles for the peoples along the Slave River in the Northwest Territories. During the winter a stable ice cover provides a vital transportation link to hunting, trapping, and fishing areas along the river. However, little was known about the Slave River ice cover characteristics and behaviour during the freeze-up and ice cover progression period. RADARSAT-2 satellite and time-lapse camera imagery were used in this study to document the different types of ice and understand the mechanisms of ice cover formation progression along the river during the course of winter. Time-lapse images were analyzed to observe the frazil ice generation and patterns of stable ice cover formation of the Slave River near Fort Smith during freeze-up. RADARSAT-2 images acquired from the Slave River Delta areas captured ice cover flooding due to higher river flows in mid-winter. Field surveys along the river provided insights about the ice cover growth in various sections along the river. Air pockets and layers under the ice cover were also detected during the ice surveys. The variation of water flows during the winter has a great impact on the Slave River ice regime. Increases in discharge cause the ice cover to crack or dislodge from the river banks leading to water seeping onto the ice and flooding it, which has implications for muskrat and beaver populations.

The distribution and relationship of isotope and nutrient during freeze-up period in the Lake Dalinuoer:A quantitative approach

The distribution and relationship of isotope and nutrient during freeze-up period in the Lake Dalinuoer:A quantitative approach

Winter Flows in the Mackenzie Drainage System

Winter low flow of northern rivers refers to the diminished discharge between the time of rapid flow reduction in the freeze-up period and the arrival of spring freshet, when the flow makes a quick rise. For the Mackenzie River in Canada, the duration of the winter low-flow season so defined varies considerably within the river’s large basin (1.8 million km2); therefore, to give a common time frame that enables between-basin comparison we consider 1 November to 31 March as the winter flow season. Several hydroclimatic conditions influence winter flows to varying degrees. Lengthy periods of sub-freezing temper­atures inhibit rain events and prevent snowmelt, while the formation of river ice increases channel storage at the expense of discharge. Groundwater sustains baseflow, and the flow amount at most stations is related to autumnal discharge, which reflects groundwater storage status in the pre-winter season. Large reservoirs and lakes provide substantially higher winter flows than their neighboring non-lake areas. Winter flow increases downstream as more water is gathered from the expanded drainage network, but flow contribution varies: larger baseflow is delivered from uplands than from lowlands, and discharge from the Williston Lake reservoir, regulated for hydropower production, provides about half of the total winter flow of the Mackenzie. Monotonic linear trends in winter flow are detected statistically for some tributaries, but the effect of short-term flow variability and the confounding influence of managed flow should be evaluated when considering long-term tendencies and their causative factors.

Ice regime of Transbaikalian rivers under changing climate

The results of studying the ice regime in Transbaikalian rivers and its dependence on climate change are given. The magnitudes of many-year changes in the dates begin and end of freeze-up period, its duration, and maximal ice thickness are determined. The dependence of major characteristics of ice regime on air temperature and river runoff is evaluated.

Empirical sea ice thickness retrieval during the freeze-up period from SMOS high incident angle observations

Abstract. Sea ice thickness information is important for sea ice modelling and ship operations. Here a method to detect the thickness of sea ice up to 50 cm during the freeze-up season based on high incidence angle observations of the Soil Moisture and Ocean Salinity (SMOS) satellite working at 1.4 GHz is suggested. By comparison of thermodynamic ice growth data with SMOS brightness temperatures, a high correlation to intensity and an anticorrelation to the difference between vertically and horizontally polarised brightness temperatures at incidence angles between 40 and 50° are found and used to develop an empirical retrieval algorithm sensitive to thin sea ice up to 50 cm thickness. The algorithm shows high correlation with ice thickness data from airborne measurements and reasonable ice thickness patterns for the Arctic freeze-up period.

The role of ice dynamics in shaping vegetation in flowing waters

Ice dynamics is an important factor affecting vegetation in high-altitude and high-latitude streams and rivers. During the last few decades, knowledge about ice in streams and rivers has increased significantly and a respectable body of literature is now available. Here we review the literature on how ice dynamics influence riparian and aquatic vegetation. Traditionally, plant ecologists have focused their studies on the summer period, largely ignoring the fact that processes during winter also impact vegetation dynamics. For example, the freeze-up period in early winter may result in extensive formation of underwater ice that can restructure the channel, obstruct flow, and cause flooding and thus formation of more ice. In midwinter, slow-flowing reaches develop a surface-ice cover that accumulates snow, protecting habitats under the ice from formation of underwater ice but also reducing underwater light, thus suppressing photosynthesis. Towards the end of winter, ice breaks up and moves downstream. During this transport, ice floes can jam up and cause floods and major erosion. The magnitudes of the floods and their erosive power mainly depend on the size of the watercourse, also resulting in different degrees of disturbance to the vegetation. Vegetation responds both physically and physiologically to ice dynamics. Physical action involves the erosive force of moving ice and damage caused by ground frost, whereas physiological effects - mostly cell damage - happen as a result of plants freezing into the ice. On a community level, large magnitudes of ice dynamics seem to favour species richness, but can be detrimental for individual plants. Human impacts, such as flow regulation, channelisation, agriculturalisation and water pollution have modified ice dynamics; further changes are expected as a result of current and predicted future climate change. Human impacts and climate change can both favour and disfavour riverine vegetation dynamics. Restoration of streams and rivers may mitigate some effects of anticipated climate change on ice and vegetation dynamics by, for example, slowing down flows and increasing water depth, thus reducing the potential for massive formation of underwater ice.

All-Sky Surface Radiation and Clear-Sky Surface Energy Budgets: Summer to Freeze-Up in the Western Maritime Arctic

The 2009 ArcticNet expedition was a field campaign in the Amundsen Gulf–eastern Beaufort Sea region from mid-July to the beginning of November aboard the CCGS Amundsen that provided an opportunity to describe the all-sky surface radiation and the clear-sky surface energy budgets from summer to freeze-up in the data sparse western maritime Arctic. Because the fractional area of open water was generally larger than the fractional area of ice floes, the net radiation at the water surface controlled the radiation budget. Because the water albedo is much less than the albedo of the ice floes, the extent and duration of open water in summer is an important albedo feedback mechanism. From summer to freeze-up, the net all-sky shortwave radiation declined steadily as the solar angle lowered, while coincidently the net all-sky longwave radiation became increasingly negative. The all-sky net surface radiation switched from positive in summer to negative during the freeze-up period. From summer to freeze-up, both upward and downward turbulent heat fluxes occurred. In summer, a positive surface energy budget residual contributed to the melting of ice floes and/or to the warming of the Arctic Ocean's mixed layer. During the freeze-up period, with temperatures below approximately −5°C, the residuals were mainly negative suggesting that heat loss from the ocean's mixed layer and heat released by the phase change of water were significant components of the energy budget's residual.

Quantifying steep channel freezeup processes

Abstract Freezeup processes in steep channels are particularly dynamic and generate important water level, water temperature, and discharge variations. This work presents unique field data quantifying steep channel ice processes during the freezeup period. Water temperature variability and ice dam buildup and breaching rates measured in the Montmorency watershed, Quebec, Canada are reported. The conditions leading to the development and melting of ice dams are discussed. The rate at which water is stored in the drainage system at different channel Strahler orders and the consequent freezeup flooding rates and discharge depressions are estimated. A major contribution of this paper is the formulation and application of an original heat budget model designed to specifically address the particularities of steep channel heat fluxes. The modeled net heat flux is compared with measured water temperature and with the timing of ice dam buildup and breaching events for a second order channel. This work can be of interest for cold region hydraulic and transportation engineers who are responsible for the safe design of bridges and culverts, for cold region hydrologists, for hydroelectric producers and for water intake operators in need of evaluating and forecasting low freezeup discharge conditions. Freshwater and riparian zone biologists can also refer to this work to evaluate the conditions that affect the fauna and flora at freezeup and public emergency services can use this work to quantify the risk of freezeup flooding along steep channels.

Monthly water quality forecasting and uncertainty assessment via bootstrapped wavelet neural networks under missing data for Harbin, China

In this paper, bootstrapped wavelet neural network (BWNN) was developed for predicting monthly ammonia nitrogen (NH(4+)-N) and dissolved oxygen (DO) in Harbin region, northeast of China. The Morlet wavelet basis function (WBF) was employed as a nonlinear activation function of traditional three-layer artificial neural network (ANN) structure. Prediction intervals (PI) were constructed according to the calculated uncertainties from the model structure and data noise. Performance of BWNN model was also compared with four different models: traditional ANN, WNN, bootstrapped ANN, and autoregressive integrated moving average model. The results showed that BWNN could handle the severely fluctuating and non-seasonal time series data of water quality, and it produced better performance than the other four models. The uncertainty from data noise was smaller than that from the model structure for NH(4+)-N; conversely, the uncertainty from data noise was larger for DO series. Besides, total uncertainties in the low-flow period were the biggest due to complicated processes during the freeze-up period of the Songhua River. Further, a data missing-refilling scheme was designed, and better performances of BWNNs for structural data missing (SD) were observed than incidental data missing (ID). For both ID and SD, temporal method was satisfactory for filling NH(4+)-N series, whereas spatial imputation was fit for DO series. This filling BWNN forecasting method was applied to other areas suffering "real" data missing, and the results demonstrated its efficiency. Thus, the methods introduced here will help managers to obtain informed decisions.

Continuous monitoring of river surface ice during freeze-up using upward looking sonar

This paper describes an algorithm that has been developed to process field data from two upward looking sonars, one high (546 kHz) and one low (235 kHz) frequency, a 2 MHz current profiler and a monitoring station, to measure surface ice characteristics on the North Saskatchewan River in the north east of Edmonton, AB, Canada, during the 2009/2010 freeze-up season. The algorithm used to calculate pans/rafts drafts and lengths, and the surface ice concentration is described. The validity and the accuracy of this technique were tested and computed surface ice conditions are presented. For the purpose of examining the effects of hydro-meteorological conditions on measured pan characteristics, the freeze-up period was divided into three distinct stages depending on the measured surface ice concentration. Pan/raft drafts were found to range from 0.1 to 1.0 m and pan/raft lengths from 0.6 to 8.0 m. The sonar proved to be very accurate in detecting the exact surface ice conditions locally above the sonar beam. However, interpretation of the results can sometimes be challenging, especially when physical processes such as bridging affect local ice conditions. Therefore, additional visual observations (e.g. time lapse photography) of surface ice conditions are recommended to aid in the interpretation of sonar measurements. ► A computer algorithm was developed to process data gathered using upward looking sonars during river freeze-up. ► the algorithm computes pans/rafts drafts and lengths, and the surface ice concentration. ► upward looking sonars proved to be very accurate for measuring local surface river ice conditions.

Oxygen regime of rivers in Volga, Ob, and Lena basins

A close negative correlation was found to exist between oxygen concentration in winter and the latitude in rivers in the basins of the Ob and Volga. This can be explained by an increase in the duration of the freeze-up period (during which oxygen is consumed for oxidation of organic matter (OM) in water and bottom sediments) in the direction from the south to the north, while the input from oxygen from the atmosphere and photosynthesis is practically zero. No inverse relationship was found to exist between dissolved oxygen and the latitude, because in rivers of mountain and semimountain types, typical of this basin, organic matter (OM) does not accumulate in bottom sediments and the conditions of atmospheric circulation are better than in lowland rivers. The magnitude of seasonal variations in oxygen in lowland rivers is greater than in mountain ones because of an abrupt drop in oxygen concentration in slowly flowing rivers in winter and the saturation of their water by oxygen from photosynthesis in summer.

Atmospheric forcing of sea ice in Hudson Bay during the fall period, 1980–2005

The principal objective of this study is to describe the autumn sea ice regime of Hudson Bay in the context of atmospheric forcing from 1980 to 2005. Both gridded Canadian Ice Service (CIS) data and Passive Microwave (PMW) data are used to examine the freezeup period for weeks of year (WOY) 43–52. Sea ice concentration (SIC) anomalies reveal statistically significant trends, ranging from −23.3% to −26.9% per decade, during WOY 43–48 using the CIS data and trends ranging from −12.7% to −16.8% per decade during WOY 45–50 using the PMW data. Surface air temperature (SAT) anomalies are highly correlated with SIC anomalies (r2 = 0.52–0.72) and with sea ice extents (r2 = 0.53–0.72). CIS data show that mean sea ice extents based on SICs ≥80% (consolidated ice) have decreased by 1.05 × 105 to 1.17 × 105 km2 for every 1°C increase in temperature in late November; PMW data show similar results. Regression analysis between SAT and standardized climate indices over the 1951–2005 period show that the East Pacific/North Pacific index is highly predictive of interannual SATs followed by the North Atlantic Oscillation and Arctic Oscillation indices. The data show that the Hudson Bay area has recently undergone a climate regime shift, in the mid 1990s, which has resulted in a significant reduction in sea ice during the freezeup period and that these changes appear to be related to atmospheric indices.

Comparison study of the modern ostracod associations in the Kara and Laptev seas: Ecological aspects

Recent ostracod assemblages were investigated from coretop sediment samples collected in the eastern Kara Sea from water depths down to 300 m. A total of 45 species were identified, 27 of them were reported for the Kara Sea for the first time. The Kara Sea data were compared with our results on the distribution of ostracods in the eastern Laptev Sea. The spatial distribution of recent taxa and the ecological groupings demonstrate a clear relation to dominant environmental factors which range from estuarine to full-marine conditions. Four assemblages related to average summer bottom water salinities were established: (1) a freshwater assemblage from the inner estuaries of the Ob' and Yenisei rivers with salinities less than 2 and from thermokarst lagoons of the southern Laptev Sea coast with strong salinization in winter; (2) a brackishwater assemblage of the outer estuaries of the Ob' and Yenisei rivers with salinities up to 26; (3) a mixed euryhaline–marine assemblage dominated by euryhaline species Paracyprideis pseudopunctillata and Heterocyprideis sorbyana from the inner shelf river-affected zone of the Kara and Laptev seas, where salinities range between 26 and 32; (4) a taxonomically diverse marine assemblage dominated by shallow-water marine taxa from the northern parts of the Kara and Laptev shelves and upper continental slope with stable bottom environments and a salinity higher than 32. Abundant euryhaline species found at greater water depths are identified as part of an ice-rafted assemblage. They are possibly entrained into the newly formed fast ice during autumn storms and freeze-up period and then transported to the distal open-sea areas during summer.

Pixel-scale evaluation of SSM/I sea-ice algorithms in the marginal ice zone during early fall freeze-up

Observed reduction in recent sea ice areal extent and thickness has focused attention on the fact that the Arctic marine system appears to be responding to global-scale climate variability and change. Passive microwave remote-sensing data are the primary source underpinning these reports, yet problems remain in geophysical inversion of information on ice type and concentration. Uncertainty in sea-ice concentration (SIC) retrievals is highest in the summer and fall, when water occurs in liquid phase within the snow–sea-ice system. Of particular scientific interest is the timing and rate of new ice formation due to the control that this form of sea ice has on mass, energy and gas fluxes across the ocean–sea-ice–atmosphere interface. In this paper we examine the critical fall freeze-up period using in situ data from a ship-based and aerial survey programme known as the Canadian Arctic Shelf Exchange study combined with microwave and optical Earth observations data. Results show that: (1) the overall physical conditions observed from aerial survey photography were well matched with coincident moderate-resolution imaging spectroradiometer data and Radarsat ScanSAR imagery; (2) the shortwave albedo was linearly related to old ice concentration derived from survey photography; (3) the three SSM/I SIC algorithms (NASA Team (NT), NASA Team 2 (NT2), and Bootstrap (BT)) showed considerable discrepancies in pixel-scale comparison with the Radarsat ScanSAR SICs well calibrated by the aerial survey data. The major causes of the discrepancies are attributed to (1) the inherent inability to detect the new thin ice in the NT and BT algorithms, (2) mismatches of the thin-ice tie point of the NT2 algorithm, and (3) sub-pixel ambiguity between the thin ice and the mixture of open water and sea ice. These results suggest the need for finer resolution of passive microwave sensors, such as AMSR-E, to improve the precision of the SSM/I SIC algorithms in the marginal ice zone during early fall freeze-up. Copyright © 2006 John Wiley & Sons, Ltd.

Surface energy budget over the central Arctic Ocean during late summer and early freeze‐up

During the Arctic Ocean Expedition 1996 we measured radiative, turbulent, and conductive heat fluxes from late summer to early freeze‐up, over the Norwegian Sea into the Arctic Ocean to latitude 88°N, mainly in air that had spent several days over the pack ice. In summer the net radiation was positive, on average +26 W m−2, where the leads contributed more than the ice floes. Clouds had a cooling net effect. During the freeze‐up period the net radiation was dominated by the ice floes, on average −8 W m−2, with a warming cloud net effect. The turbulent fluxes were on average small and unable to balance the net radiation, except during cyclone activity, when the total turbulent heat fluxes averaged to +11 W m−2. During the freeze‐up period the conductive heat flux averaged to −4 W m−2. In summer, melting of ice must largely have caused the average positive residual (+21±5 W m−2). During the freeze‐up period the average negative residual (−7±3 W m−2) must result from formation of new ice and oceanic turbulent/convective fluxes in the leads. There was a diurnal cycle in net radiation both in summer (17 W m−2 average amplitude) and the freeze‐up period (8 W m−2). The turbulent sensible heat flux (average amplitude ∼5 W m−2), as well as the turbulent latent heat flux (1 W m−2 in summer) and conductive flux in the ice (2.5 W m−2 during the freeze‐up) responded to this. These results emphasize the high sensitivity of the heat and radiation budget balances to changes in cyclone activity, leads, and clouds.

Composition and distribution of the pelagic and sympagic algal assemblages in the Laptev Sea during autumnal freeze-up

The phytoplankton and ice algal assemblages in the Siberian Laptev Sea during the autumnal freeze-up period of 1995 are described. The spatial distribution of algal taxa (diatoms, dinoflagellates, chrysophytes, chlorophytes) in the newly formed ice and waters at the surface and at 5 m depth differed considerably between regions. This was also true for algal biomass measured by in situ fluorescence, chlorophyll (Chl) a and taxon-specific carbon content. Highest in situ fluorescence and Chl a concentrations (ranging from 0.1 to 3.2 μg l–1) occurred in surface waters with maxima in Buor Khaya Bay east of Lena Delta. The algal standing stock on the shelf consisted mainly of diatoms, dinoflagellates, chrysophytes and chlorophytes with a total abundance (excluding unidentified flagellates <10 μm) in surface waters of 351–33 660 cells l–1. Highest algal abundance occurred close to the Lena Delta. Phytoplankton biomass (phytoplankton carbon; PPC) ranged from 0.1 to 5.3 μg C l–1 in surface waters and from 0.3 to 2.1 μg C l–1 at 5 m depth, and followed the distribution pattern of abundances. However, the distribution of Chl a differed considerably from the distribution pattern shown by PPC. The algal assemblage in the sea ice, which could not be quantified due to high sediment load, was dominated by diatom species, accompanied by dinoflagellates. Thus, already during the early stage of autumnal freeze-up, incorporation processes, selective enrichment and subsequent growth lead to differences between surface water and sea ice algal assemblages.