Search

High-frequency (HF) ocean radar has the potential to observe sea ice, but this application has not been thoroughly investigated. In this article, we explore sea ice observation with HF radar (HFR) and develop a corresponding signal processing procedure. We derived sea ice radial velocities from raw HFR data and determined angular positions using the MUltiple SIgnal Classification (MUSIC) algorithm. Sea ice velocities were evaluated by comparison with measurements from moored acoustic Doppler current profilers (ADCPs) and a drifting buoy. The root-mean-square (rms) differences between the sea ice velocities obtained by HFR and the ADCPs decreased with increasing radar signal-to-noise ratio (SNR), implying that a high SNR was associated with accurate measurement of sea ice velocity. These results agree with previous validations of HFR ocean current observations. The sea ice pattern mapped by HFR is consistent with the simultaneous C-band microwave radar observations. In particular, the HFR captured the approach of the sea ice edge toward the coast. Sharp attenuation of the HF signal over high concentration sea ice resulted in missing and/or low-SNR data at ranges beyond the front edge of pack ice, in qualitative agreement with model estimates of attenuation.

An extensive research programme has been carried through in 2007–2009 on the ice cover geophysics in Lake Kilpisjärvi, located at 69 °03′N 20 °50′E, 473 m above sea level, and about 60 km from the shore of the Norwegian Sea. The surface area of the lake is 37.1 km2, and the maximum depth is 57 m. Data were collected of ice, snow and weather conditions with an automatic ice station in the lake. The heat budget together with ice structure, growth and melting was analysed. It was dominated by the radiation balance, down to −50 W m−2 in fall and up to 100 W m−2 in summer. Turbulent heat fluxes were significant before freeze-up in fall (absolute values up to 30 W m−2), but in the ice season they were small except for an occasional sensible heat flux which was large due to warm air advection. The evolution of ice thickness served as a very good condition for the total surface heat flux, and was found to be consistent with the ice station time series of air–ice heat fluxes.

Sea ice dynamics is examined for the frequency spectra of ice velocity using mathematical models and ice motion data. The data are from the Baltic Sea and Sea of Okhotsk. A general spectrum for linear coupled ice–ocean free drift is derived and analysed. Interior dynamics shows singularity in the Coriolis (inertial) frequency and asymptotic high frequency power law of − 2. In the presence of internal friction the spectrum is expected to evenly fall to zero with frictional resistance increasing above yield level. In the observations, the main spectral peak is wide at the synoptic time scales. The Okhotsk Sea spectra show inertial and tidal signals but they are missing from the Baltic Sea due to very weak tides and shallow depth damping inertial oscillations. Above semidiurnal frequency ice velocity spectra fall in power law of − 5/3, inherited from atmospheric and oceanic dynamics, and at very high frequencies, above 0.5 cph (cycles per hour) the fall seems to slow down to power law below − 1. The Eulerian ice velocity spectrum was higher than the Lagrangian spectrum throughout the entire obtained spectrum (0.1–5 cpd (cycles per day)). Theoretical spectra show general agreement with observations except that they have less kinetic energy at very high frequencies and much stronger inertial peak. Thus Eulerian observations are more variable than Lagrangian observations.

The air–sea ice CO 2 flux was measured over landfast sea ice in the Chukchi Sea, off Barrow, Alaska in late May 2008 with a chamber technique. The ice cover transitioned from a cold early spring to a warm late spring state, with an increase in air temperature and incipient surface melt. During melt, brine salinity and brine dissolved inorganic carbon concentration (DIC) decreased from 67.3 to 18.7 and 3977.6 to 1163.5 μmol kg −1 , respectively. In contrast, the salinity and DIC of under-ice water at depths of 3 and 5 m below the ice surface remained almost constant with average values of 32.4±0.3 (standard deviation) and 2163.1±16.8 μmol kg −1 , respectively. The air–sea ice CO 2 flux decreased from +0.7 to −1.0 mmol m −2 day −1 (where a positive value indicates CO 2 being released to the atmosphere from the ice surface). During this early to late spring transition, brought on by surface melt, sea ice shifted from a source to a sink for atmospheric CO 2 , with a rapid decrease of brine DIC likely associated with a decrease in the partial pressure of CO 2 of brine from a supersaturated to an undersaturated state compared to the atmosphere. Formation of superimposed ice coincident with melt was not sufficient to shut down ice–air gas exchange.

The air–sea-ice CO2 flux was measured in the ice-covered Saroma-ko, a lagoon on the northeastern coast of Hokkaido, Japan, using a chamber technique. The air–sea-ice CO2 flux ranged from −1.8 to +0.5 mg C m−2 h−1 (where negative values indicate a sink for atmospheric CO2). The partial pressure of CO2 (pCO2) in the brine of sea ice was substantially lower than that of the atmosphere, primarily because of the influence of the under-ice plume from the Saromabetsu river located in the southeastern part of the lagoon. This suggests that the brine had the ability to take up atmospheric CO2 into the sea ice. However, the snow deposited over the sea ice and the superimposed ice that formed from snowmelting and refreezing partially blocked CO2 diffusion, acting as an impermeable medium for CO2 transfer. Our results suggest that the air–sea-ice CO2 flux was dependent not only on the difference in pCO2 between the brine and the overlying air, but also on the status of the ice surface. These results provide the necessary evidence for evaluation of the gas exchange processes in ice-covered seas.

Field experiments were conducted during two melting periods, April 2006 and April 2007, in Lake Vendyurskoe. The observation programme included weather, ice and snow thickness and structure, water temperature and solar radiation transfer through the ice. Albedo showed a systematic decrease from 0.5–0.8 for wintertime dry ice and snow to 0.1 for wet bare ice in spring, with spatial standard deviation of about 10%. The e-folding depth of light level was 60–80 cm for congelation ice and 15 cm for snow-ice. Light transmissivity of the ice cover increased from melting but decreased from ice deterioration; it varied between 0.25–0.35 in 2006, while in 2007 there was a systematic trend from 0.1 to 0.5 in six days. The heat budget was governed by net solar radiation with daily peaks up to 400–500 W m−2 on clear days. The average daily melt was 1.2 cm at the surface, 0.5 cm at the bottom and 1–2 cm (thickness equivalent) in the interior.

Lake Paajarvi, a boreal Finnish lake, was investigated in winter for weather conditions, structure and thickness of ice and snow, solar radiation, and under-ice current and temperature. Heat budget of Lake Paajarvi in January–March was governed by terrestrial radiation losses of 20–35 W m−2 recompensed by ice growth of 0.5–1.0 cm day−1. In April, snow melted, albedo decreased from 0.8 to

Measurements of sea-ice thickness were obtained from drill holes, an ice-based electromagnetic induction instrument (IEM), and a ship-borne electromagnetic induction instrument (SEM) during the early-melt season in the southern Chukchi Sea in 2002 and 2004, and in late summer 2003 at the time of minimum ice extent in the northern Chukchi Sea. An ice roughness criterion was applied to distinguish between level and rough or ridged ice. Ice-thickness modes in the probability density functions (PDFs) derived from drill-hole and IEM measurements agreed well, with modes at 1.5–1.6 and 1.8–1.9 m for all data from level ice. The PDFs derived from SEM measurements show that the primary modes are at 0.1 and 1.1 m in 2003 and 0.7 m in 2004. In 2002 and 2004, significant fractions (between one-third and one-half) of level ice were found to consist of rafted ice segments. Snow depth varied significantly between years, with 2004 data showing more than half the snow cover on level ice to be at or below 0.05 m depth in late spring. Ice growth simulations and examination of ice drift and deformation history indicate that impacts of atmospheric and oceanic warming on level-ice thickness in the region over the past few decades are masked to a large extent by variability in snow depth and the contribution of deformation processes. In comparison with submarine sonar ice-thickness data from previous decades, a reduction in ice thickness by about 0.5–1 m is in part explained by the replacement of multi-year with first-year ice over the Chukchi and Beaufort shelves.

From December to June 2002–2003, sea-ice and oceanic data were obtained from moorings near Sakhalin in the west central Okhotsk Sea. Ice draft measured by sonar reveals distinct periods of thin and thick ice. Thin-ice periods in January–March corresponded to offshore ice movement and increasing seawater salinity. The measured change in salinity corresponds well with that derived from heat-flux calculations using the observed ice thickness. Brine rejection from ice growing in a coastal polynya off northern Sakhalin is responsible for much of the observed salinity increase. The simultaneous observation of dense shelf water ( > 26.7 σ θ ) suggests that this region is one possible source. The periods of thick-ice incursion are likely indicative of heavily deformed pack formed further north and drifting south with the current. The mean draft (1.95 m), thick-ice ratio, and keel frequency during these periods are close to values observed in the Beaufort Sea. Freshwater transport estimated from the observed ice thickness and velocity is larger than that of the Amur River discharge.

Physico-chemical properties in the brine and under-ice water were measured in Saroma-ko Lagoon on the northeastern coast of Hokkaido, Japan, which is connected to the Sea of Okhotsk, during the period from mid-February through mid-March 2006. The brine within brine channels of the sea ice was collected with a new sampling method examined in this study. Salinity, dissolved inorganic carbon (DIC), total alkalinity (TA), dissolved oxygen (DO), nutrients and oxygen isotopic ratio (δ 18 O) contained in the brine within brine channels of the sea ice and in the under-ice water varied largely in both time and space during the ice melt period, when discharge from Saromabetsu River located on the southeast of the lagoon increased markedly due to the onset of snow melting. The under-ice plume expands as far as 4.5 km from the river mouth at mid-March 2006, transporting chemical components supplied from the river into the lagoon. The under-ice river water was likely transported into the sea ice through well-developed brine channels in the sea ice due to upward flushing of water through brine channels occurred by loading of snowfalls deposited over the sea ice. These results suggest that the river water plume plays an important role in supplying chemical components into the sea ice, which may be a key process influencing the biogeochemical cycle in the seasonally ice-covered Saroma-ko Lagoon.

Electromagnetic–induction (EM) instruments can be used to estimate Sea-ice thickness because of the large contrast in the conductivities of Sea ice and Sea water, and are currently used in investigations of Sea-ice thickness. In this Study we analyze Several Snow, ice and Sea-water Samples and attempt to derive an appropriate formula to transform the apparent conductivity obtained from EM measurements to the total thickness of Snow and ice for all regions and Seasons. This was done to Simplify the EM tuning procedure. Surface EM measurement transects with the instrument at varying heights above the ice were made in the Chukchi Sea, off East Antarctica, in the Sea of Okhotsk and in Saroma-ko (lagoon). A Standardized transformation formula based on a one-dimensional multi-layer model was developed that also considers the effects of water-filled gaps between deformed ice, a Saline Snow Slush layer, and the increase in the footprint Size caused by increasing the instrument height. The overall average error in ice thickness determined with the Standardized transform was

Recent observations have revealed that dynamical thickening is dominant in the growth process of Sea ice in the Southern Sea of Okhotsk. That indicates the importance of understanding the nature of thick deformed ice in this area. The objective of the present paper is to establish a Ship-based method for observing the thickness of deformed ice with reasonable accuracy. Since February 2003, one of the authors has engaged in the core Sampling using a Small basket from the icebreaker Soya. Based on these results, we developed a new model which expressed the internal Structure of pack ice in the Southern Sea of Okhotsk, as a one-dimensional multilayered Structure. Since 2004, the electromagnetic (EM) inductive Sounding of Sea-ice thickness has been conducted on board Soya. By combining the model and theoretical calculations, a new algorithm was developed for transforming the output of the EM inductive instrument to ice + Snow thickness (total thickness). Comparison with total thickness by drillhole observations Showed fair agreement. The probability density functions of total thickness in 2004 and 2005 Showed Some difference, which reflected the difference of fractions of thick deformed ice.

The thickness of coastal landfast ice in the Sea of Okhotsk has been examined based on field data and thermodynamic modelling. The study sites were Saroma-ko Lagoon, Hokkaido and Kleye Strait, Sakhalin. The ice sheet has a two-layer structure: a granular snow–ice layer on top and a columnar ice layer below. In Saroma-ko Lagoon, the ice grows to 40–50 cm, with snow–ice portion of 10–100%. In Kleye Strait, the ice grows to about 100 cm, with a remarkable addition (on average 24 cm) during mid-March to mid-April due to snow–ice formation. A one-dimensional thermodynamic ice–snow model was calibrated with observed data and used to examine the thickness climatology; the snow component takes into account snow compaction, slush formation due to flooding, melting or rain and snow–ice growth. The model outcome showed reasonably good agreement for both sites. In Saroma-ko Lagoon, the calibration was based on four winters. The maximum annual ice thickness was in the model on average 3 cm below the observed one, 16 cm in the worst case; the model snow thickness was within 10 cm from the observed ones in February; and the date of ice breakup was on average biased late by 5 days and 11 days in the worst case. The model simulations predicted formation of slush layers and their persistency for 1–4 weeks in different winters. Climatological simulation resulted in mean maximum annual ice thickness of 32 cm, of which 15 cm was snow–ice. In Kleye Strait, the calibration was based on one ice season. The maximum annual ice thickness was 7 cm biased down, and the model snow thickness was within 10 cm from the observed level. Climatological simulation resulted in mean maximum annual ice thickness of 108 cm, of which 70 cm was congelation ice and 38 cm was snow–ice, and the ice season lasted from 5 November to 5 June. Thus, slush formation and its freezing are crucial in the study basin.

The main physical and ecological processes associated with the summer melt period in the marginal ice zone (MIZ) were investigated in a multidisciplinary research programme (ICE-BAR), which was carried out in the northern Barents Sea during June–August 1995–1996. This study provided simultaneous observations of a wide range of physical and chemical factors of importance for the melting processes of sea ice, from its southernmost margins at about 77.5°N to the consolidated Arctic pack ice at 81.5°N. This paper includes a description of the oceanographic processes, ice-density packing and structures in cores, optical properties of water masses and the ice, characteristics of the incident spectral radiation and chlorophyll — leading to primary production. Large seasonal and inter-annual variations in ice cover in the MIZ were evident from satellite images as well as ship observations. Even if the annual variation in ice extent may be large, the inter-annual variations may be even larger. The minimum observed ice extent in March, for example, can be smaller than the maximum observed ice extent in September. Oceanographic phenomena such as the semi-permanent lee polynyas found west and south-west of Kvitoya and Franz Josef Land and the bay of open water, the “Whalers Bay”, north of the Spitsbergen are structures which can change with time intervals of hours to decades. For example, the polynya south of Franz Josef Land was clearly evident in 1995 but was only seen for a short period in 1996. The observed variability in physical conditions directly affects the primary production in the MIZ. From early spring, solar radiation penetrates both leads and the ice itself, initiating algal production under the ice. Light measurements showed that the melt ponds act as windows, permitting the transmission of incoming solar radiation through to the underlying sea ice, thus, accelerating the melting process and enhancing the under-ice primary production. In June 1995, the N–S transect went through a pre-bloom area well inside the ice-covered part of the Barents Sea to a post-bloom phase in the open waters south of the ice edge. The biological conditions in the later season (August) of 1996 were considerably more variable. The longer N–S transect in August 1996 passed through areas with variable ice and oceanographic conditions, and different developmental stages of phytoplankton blooms were encountered. The previously adapted picture of a plankton bloom following the retreating ice edge northwards was not seen.

Oceanographic and sea ice cover data were taken in and below first-year landfast ice in Resolute Passage, Northwest Territories, Canada, in early May 1992. A time series of temperature profiles was measured across the ice layer and in the water below. Observations of the 3-D current regime in the frictional boundary layer under the ice were taken to describe the current regime and the turbulent heat and mass transfers adjacent to the ice-water interface. The temporal variabilities in current, temperature and salinity under the landfast ice appeared to be related to the dominant diurnal and semi-diurnal tidal components. Turbulent heat and momentum fluxes near the ice-water interface were estimated using the eddy correlation method. Over the 12 day sampling period, the mean friction velocity was 5.6 × 10 −3 ms −1 , the mean vertical eddy viscosity was 1.12 × 10 −3 m 2 s −1 , the mean oceanic heat flux from the ocean to the ice was 32 W m −2 , the mean drag coefficient was 5.4 × 10 −3 , and the mean heat transfer coefficient was 2.3 × 10 −3 .

Observations in both the ice and slush layers suggest that sea water intrudes into the snow layer following a snow storm. Ice temperature values recorded at 1 cm below the snow‐ice interface show that the upward flux of sea water is of short duration. This is followed by a period of intense brine drainage characterized by the migration of a salty brine layer, with salinities up to 42 psu. These results suggest that a snow storm induces a complete (upward) flushing of the brine channel network and major modifications of snow and ice characteristics. Melt rates and downward brine fluxes were calculated using salinity measured in a 40 cm deep box placed on the ice‐water interface, which isolated a 50 × 50 cm area of sea ice from ocean mixing processes. In this semi‐isolated environment, observed salinity changes allowed us to determine melt water fluxes and brine drainage or flushing even though ice thickness measurements did not show any significant change. Melt rates up to 21 cm/month and equivale...

Surface intrusions of freshwater from rivers and buoyancy effects caused by surface freezing or ice melt were found to play an important role in modifying turbulent processes in coastal regions. Measurements of the three dimensional current regime in the oceanic boundary layer taken under fast ice in southeast Hudson Bay during the ice melt season are discussed. Mixing processes just below the ice-water interface are affected by tidal variability, ice type and distribution, low frequency current amplitude and buoyancy characteristics in the boundary layer adjoining the sea ice. Changes in turbulence intensities are shown as functions of current speed and of depth relative to the interface, within the water column in neutral stability. Time series of the current speed and momentum flux sampled in the under-ice boundary layer are shown.

The southernmost part of the Sea of Okhotsk, along the northeastern sea coast of Hokkaido, Japan is usually covered with sea ice from early January through March. There are unstable ice packs, some of which consist of drift ice from the north, while others are locally frozen fast ice. This work reviews our knowledge on sea ice distributions and movements monitored by sea-ice radars and measurements of air-ice drag coefficients over sea ice in the coastal Sea of Okhotsk off Hokkaido. In addition, data from the marine tower standing 6.5 m deep in water and 600 off the shore have been analyzed by the eddy corrlation method to obtain wind stress and heat flux for three separate periods of pre-ice forming, ice forming and ice melt. Variations in drag coefficient and in heat flux coefficient of sea ice and water are shown as a function off wind speed, stability and the surface condition of sea ice. Moreover, surface roughness expressed by backscatter signals from a X-band radar is compared to the micrometeorological roughness, i.e., the air-ice drag coefficient.

[1] The seasonal development of the structure, salinity, and stable oxygen isotopic composition (δ18O) of landfast sea ice was studied during the winter seasons 1999–2001 in the Gulf of Finland in the Baltic Sea. The main focuses were on the seasonal and the interannual variability in ice properties and composition and on the contribution of meteoric ice to sea ice mass balance. Results provide a first statistical description of the seasonal evolution of sea ice in mild ice climate conditions. The ice has a characteristic structure with an upper granular ice layer, composed almost exclusively of superimposed ice and snow-ice, averaging at 20–35% of the total ice thickness. The remaining is composed of columnar or intermediate granular columnar ice, depending on growth conditions. While salinity shows a uniform profile through the ice, δ18O shows lower values in the surface because of meteoric ice formation. The thin ice cover is susceptible to changes in atmospheric conditions, and rapid changes in ice salinity are connected to changes in the ice thermal regime and flooding. The contribution of meteoric ice varied from 0 to 35% (by mass), depending on season and year. Superimposed ice formation is a recurring process and significantly contributed to ice growth (up to 20% by mass), especially late in the season during snowmelt-freeze cycles. Liquid precipitation also cause formation of intermittent superimposed ice layers at these latitudes. The contribution of meteoric ice to sea ice mass balance is largely dependent on the amount and timing of snow accumulation and timing of snowmelt-freeze processes, which all showed large year-to-year variation. The conditions presented here may start to occur at higher latitudes if global warming continues.

[1] Samples of land-fast sea ice collected along the Finnish coast of the Baltic Sea, between latitudes 60.2°N and 65.7°N, in January to April 2000 were analyzed for physical, biological, and chemical parameters. Both spatial and temporal variability were investigated. Snow-ice contributed in average a third of the total ice thickness, while the snow fraction (by mass) of the ice was 20% on average. Snow-ice formation increased the nitrogen concentrations substantially, mainly in the upper parts of the ice cover. Phosphorus on the other hand was controlled by biological uptake, with distinct maxima in the bottommost parts of the ice cover. The chlorophyll-a concentrations were dependent on the physical properties of the ice to some extent. In more saline waters the chlorophyll-a concentrations in the ice were variable (1–17 μg l−1). However, in the less saline waters of the Bothnian Bay the concentrations were generally considerably lower (

Downward material fluxes under seasonal sea ice were measured using a time-series sediment trap installed at an offshore site in the Okhotsk Sea north of Hokkaido, Japan, from 13 January to 23 March 2005. The maximum fluxes of lithogenic material (753 mg m−2 day−1) and organic matter (mainly detritus; 333 mg m−2 day−1) were recorded during the period in which sea ice drifted ashore and increased in extent, from 13 January to 9 February. Organic matter as fecal pellets (81–93 mg m−2 day−1) and opal as biosilica (51–67 mg m−2 day−1), representing diatom fluxes, were abundant in sediment trap samples obtained during the period of full sea ice coverage from 10 February to 9 March. Microscopic observations revealed that fecal pellets were largely diatom frustules, suggesting that zooplankton actively grazed on ice algae during the period of full sea ice coverage. During the period of retreating sea ice, from 10 to 23 March, the phytoplankton flux showed a rapid increase (from 9.5 to 22.5 × 106 cells m−2 day−1), reflecting their release into the water column as the sea ice melted. Our results demonstrate that the quantity and quality of sinking biogenic and lithogenic materials vary with the seasonal extent of sea ice in mid-winter.