• Bergeron C. J. Jr., , , Brusstar J. R. , , Yi N. , , Wu Y. , , and Ioup J. W. , 1999: A new vertical continuation procedure for airborne electromagnetic field data from the modified image method. Geophysics, 64 , 13641368.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eicken, H., , Grenfell T. C. , , Perovich D. K. , , Richter-Menge J. A. , , and Frey K. , 2004: Hydraulic controls of summer Arctic pack ice albedo. J. Geophys. Res., 109 , C08007. doi:10.1029/2003JC001989.

    • Search Google Scholar
    • Export Citation
  • Haas, C., 1998: Evaluation of ship-based electromagnetic-inductive thickness measurements of summer sea-ice in the Bellingshausen and Amundsen Seas, Antarctica. Cold Regions Sci. Technol., 27 , 116.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heil, P., 2006: Atmospheric conditions and fast ice at Davis, East Antarctica: A case study. J. Geophys. Res., 111 , C05009. doi:10.1029/2005JC002904.

    • Search Google Scholar
    • Export Citation
  • Heil, P., , Allison I. , , and Lytle V. I. , 1996: Seasonal and interannual variations of the oceanic heat flux under a landfast Antarctic sea ice cover. J. Geophys. Res., 101 , (C11). 2574125752.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hristoforou, E., , and Chiriac H. , 2002: Position measuring system for applications in field sports. J. Magn. Magn. Mater., 249 , 407410.

  • Hristoforou, E., , Chiriac H. , , Neagu M. , , and Karayannis V. , 1997: On the calibration of position sensors based on magnetic delay lines. Sens. Actuators, 59A , 8993.

    • Search Google Scholar
    • Export Citation
  • Hristoforou, E., , Dimitropoulos P. D. , , and Petrou J. , 2006: A new position sensor based on the MDL technique. Sens. Actuators, 132A , 112121.

    • Search Google Scholar
    • Export Citation
  • Hudson, R., 1990: Annual measurement of sea ice thickness using an upward-looking sonar. Nature, 344 , 135137.

  • Karagiannis, V., , Manassis C. , , and Bargiotas D. , 2003: Position sensors based on the delay line principle. Sens. Actuators, 106A , 183186.

    • Search Google Scholar
    • Export Citation
  • Kawamura, T., , Ohshima K. I. , , Takizawa T. , , and Ushio S. , 1997: Physical, structural, and isotopic characteristics and growth processes of fast sea ice in Lützow-Holm Bay, Antarctica. J. Geophys. Res., 102 , (C2). 33453355.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Laxon, S., , Peacock N. , , and Smith D. , 2003: High interannual variability of sea ice thickness in the Arctic region. Nature, 425 , 947950.

  • Liu, J., , Schmidt G. A. , , Martinson D. G. , , Rind D. , , Russell G. , , and Yuan X. , 2003: Sensitivity of sea ice to physical parameterizations in the GISS global climate model. J. Geophys. Res., 108 , 3053. doi:10.1029/2001JC001167.

    • Search Google Scholar
    • Export Citation
  • Mélia, D. S., 2002: A global coupled sea ice–ocean model. Ocean Modell., 4 , 137172.

  • Perovich, D. K., , and Elder B. C. , 2001: Temporal evolution of Arctic sea-ice temperature. Ann. Glaciol., 33 , 207211.

  • Perovich, D. K., , and Elder B. C. , 2002: Estimates of ocean heat flux at SHEBA. Geophys. Res. Lett., 29 , 1344. doi:10.1029/2001GL014171.

  • Perovich, D. K., , and Richter-Menge J. A. , 2006: From points to poles: Extrapolating point measurements of sea-ice mass balance. Ann. Glaciol., 44 , 188192.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perovich, D. K., , Elder B. C. , , and Richter-Menge J. A. , 1997: Observation of the annual cycle of sea ice temperature and mass balance. Geophys. Res. Lett., 24 , 555558.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perovich, D. K., , Grenfell T. C. , , Richter-Menge J. A. , , Light B. , , Tucker W. B. III, , and Eicken H. , 2003: Thin and thinner: Sea ice mass balance measurements during SHEBA. J. Geophys. Res., 108 , 8050. doi:10.1029/2001JC001079.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perovich, D. K., , Elder B. C. , , Claffey K. J. , , Stammerjohn S. , , Smith R. , , Ackley S. F. , , Krouse H. R. , , and Gow A. J. , 2004: Winter sea-ice properties in Marguerite Bay, Antarctica. Deep-Sea Res. II, 51 , 20232039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Preunkert, S., , and Wagenbach D. , 1998: An automatic recorder for air/firn transfer studies of chemical aerosol species at remote glacier sites. Atmos. Environ., 32 , 40214030.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pringle, D. J., , Eicken H. , , Trodahl H. J. , , and Backstrom L. G. E. , 2007: Thermal conductivity of landfast Antarctic and Arctic sea ice. J. Geophys. Res., 112 , C04017. doi:10.1029/2006JC003641.

    • Search Google Scholar
    • Export Citation
  • Purdie, C. R., , Langhorne P. J. , , Leonard G. H. , , and Haskell T. G. , 2006: Growth of first-year landfast Antarctic sea ice determined from winter temperature measurements. Ann. Glaciol., 44 , 170176.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Richter-Menge, J. A., , Perovich D. K. , , Elder B. C. , , Claffey K. , , Rigor I. , , and Ortmeyer M. , 2006: Ice mass-balance buoys: A tool for measuring and attributing changes in the thickness of the Arctic sea-ice cover. Ann. Glaciol., 44 , 205210.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rothrock, D. A., , Yu Y. , , and Maykut G. A. , 1999: Thinning of the Arctic sea-ice cover. Geophys. Res. Lett., 26 , 34693472.

  • Saloranta, T. M., 2000: Modeling the evolution of snow, snow ice and ice in the Baltic Sea. Tellus, 52A , 93108.

  • Smedsrud, L. H., , Budgell W. P. , , Jenkins A. D. , , and Ådlandsvik B. , 2006: Fine-scale sea-ice modelling of the Storfjorden polynya, Svalbard. Ann. Glaciol., 44 , 17.

    • Search Google Scholar
    • Export Citation
  • Strass, V. H., 1998: Measuring sea ice draft and coverage with moored upward looking sonars. Deep-Sea Res. I, 45 , 795818.

  • Sun, B., , Wen J. , , He M. , , Kand J. , , Luo Y. , , and Li Y. , 2003: Sea ice thickness measurement and its underside morphology analysis using radar penetration in the Arctic Ocean. Sci. China, 46D , 11511160.

    • Search Google Scholar
    • Export Citation
  • Tamura, T., , Ohshima K. I. , , Markus T. , , Cavalieri D. J. , , Nihashi S. , , and Hirasawa N. , 2007: Estimation of thin ice thickness and detection of fast ice from SSM/I data in the Antarctic Ocean. J. Atmos. Oceanic Technol., 24 , 17571772.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Untersteiner, N., 1961: On the mass and heat budget of Arctic sea ice. Arch. Meteor. Geophys. Bioklimatol., 12A , 151182.

  • Vavrus, S., , and Harrison S. P. , 2003: The impact of sea-ice dynamics on the Arctic climate system. Climate Dyn., 20 , 741757.

  • Vinnikov, K. Y., and Coauthors, 1999: Global warming and Northern Hemisphere sea ice extent. Science, 286 , 19341937.

  • View in gallery

    The MDL positioning sensor.

  • View in gallery

    The monitoring apparatus for sea ice thickness.

  • View in gallery

    The monitoring apparatus (denoted by MDL) and the thermal-wire thickness gauges (denoted by TWTG).

  • View in gallery

    Variations in sea ice thickness and growth rate in May and August 2006: (a),(b) data records derived from the apparatus and (c),(d) average data records derived from the thermal-wire gauges. The arrow in (a) denotes the record on 20 May (YD 139).

  • View in gallery

    Time series of (a) wind speed and (b) snow depth, derived from 29 Jul to 3 Aug 2006 (YDs 209–214).

  • View in gallery

    Sea ice thickness records obtained from drill holes and the monitoring apparatus during the experiment.

  • View in gallery

    Positive error, negative error, and the linear-regression datum line of records obtained from the monitoring apparatus during the experiment (arrows denotes an error on 20 May, YD 139).

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 42 42 8
PDF Downloads 20 20 4

A New Apparatus for Monitoring Sea Ice Thickness Based on the Magnetostrictive-Delay-Line Principle

View More View Less
  • 1 State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian, China
  • | 2 Polar Research Institute of China, Shanghai, China
  • | 3 State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian, China
  • | 4 Wu Xi Fengrun Science and Technology Ltd., Wuxi, China
© Get Permissions
Full access

Abstract

High-precision ice thickness observations are required to gain a better understanding of ocean–ice–atmosphere interactions and to validate numerical sea ice models. A new apparatus for monitoring sea ice and snow thickness has been developed, based on the magnetostrictive-delay-line (MDL) principle for positioning sensors. This system is suited for monitoring fixed measurement sites on undeformed ice. The apparatus presented herein has been tested on landfast ice near Zhongshan Station, East Antarctica, for about 6 months during the austral autumn and winter of 2006; valid data records from the deployment are available for more than 90% of the deployment’s duration. The apparatus’s precision has been estimated to be ±0.002 m for the deployment. Therefore, it is possible that this apparatus may become a standard for sea ice/snow thickness monitoring.

Corresponding author address: Ruibo Lei, State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian 116024, China. Email: leiruibo@hotmail.com

Abstract

High-precision ice thickness observations are required to gain a better understanding of ocean–ice–atmosphere interactions and to validate numerical sea ice models. A new apparatus for monitoring sea ice and snow thickness has been developed, based on the magnetostrictive-delay-line (MDL) principle for positioning sensors. This system is suited for monitoring fixed measurement sites on undeformed ice. The apparatus presented herein has been tested on landfast ice near Zhongshan Station, East Antarctica, for about 6 months during the austral autumn and winter of 2006; valid data records from the deployment are available for more than 90% of the deployment’s duration. The apparatus’s precision has been estimated to be ±0.002 m for the deployment. Therefore, it is possible that this apparatus may become a standard for sea ice/snow thickness monitoring.

Corresponding author address: Ruibo Lei, State Key Laboratory of Coastal and Offshore Engineering, Dalian University of Technology, Dalian 116024, China. Email: leiruibo@hotmail.com

1. Introduction

Sea ice plays an important role in the global climate system (Vavrus and Harrison 2003) and also is the most sensitive indicator of local and global climate change (Vinnikov et al. 1999; Heil 2006). Sea ice thickness is the most fundamentally integrative and crucially important parameter for describing ice conditions. The uncertainty of ice thickness measurements is the major difficulty in setting the ocean heat flux from ice mass and temperature measurements (Heil et al. 1996; Perovich et al. 1997; Perovich and Elder 2002). The snow cover depth is crucial for parameterizing surface albedo, evaluating snow–ice energy equilibrium, and investigating flooding on ice surfaces (Kawamura et al. 1997; Saloranta 2000; Eicken et al. 2004). Although for an area as large as the Arctic and Antarctica, and even a global scale, the final expression of the ocean–ice–atmosphere interactions is achieved through numerical modeling (Mélia 2002; Liu et al. 2003). Measuring sea ice and snow thickness at a fixed site can be utilized, in conjunction with satellite remote sensing and numerical models, to estimate regional ice growth and melt (Perovich et al. 2003; Perovich and Richter-Menge 2006; Richter-Menge et al. 2006). Thus, field programs focus ice thickness measurement in general, for example, during the Surface Heat Budget of the Arctic Ocean (SHEBA) experiment from 1997 to 1998 (Perovich et al. 2003). Meanwhile, since 2000, more and more ice mass–balance buoys (IMBs) have been deployed in the Arctic to measure the ice cover mass during an annual cycle (Richter-Menge et al. 2006).

At present, although the numerical model has been developed well (Smedsrud et al. 2006), the measurement precision of the ice/snow thickness has not improved significantly, and so it cannot provide the evidence necessary to validate the modeling results. Drill-hole measurements and thermal-wire measurements are recognized as reliable methods for measuring ice thickness (Perovich et al. 2003; Heil 2006), but their low efficiency cannot satisfy the need for monitoring ice thickness with high temporal resolution. Automatic monitoring techniques including satellite remote sensing, submarine sonar profiling, the combination of laser and electromagnetic sounding devices, and radar penetration can provide large-scale ice thickness distribution, as well as information on the onset of melt and freeze-up (Haas 1998; Bergeron et al. 1999; Rothrock et al. 1999; Laxon et al. 2003; Sun et al. 2003; Tamura et al. 2007). Nevertheless, the response of these methods to the sea ice physical properties and ambient conditions, such as the topographical property of ice surface, the volume fraction of brine within the ice matrix, and ocean-water properties, result in lower resolution for resolving the air–ice or ice–water interfaces. For much of the year, the temperature profiles derived from the thermistor string were used to infer the air–ice and ice–water interfaces (Perovich et al. 1997). Purdie et al. (2006) came to the conclusion that since the thermodynamic properties near the ice bottom are considerably unstable, the ice temperature profile is insufficient to infer the ice–water interfaces. In particular, this method is severely limited during the summer melt season when there is no significant thermal contrast between the lower layer of ice and the upper part of the water column (Perovich et al. 1997). Moored upward-looking sonar and acoustic sensors equipped in the IMBs are recognized as the most effective techniques for recording the variations in sea ice thickness automatically (Hudson 1990; Strass 1998; Perovich and Elder 2001; Perovich et al. 2004; Richter-Menge et al. 2006). When compared with drill-hole measurements, the accuracy of the moored upward-looking sonars is typically 0.20 m, and the relative error is 11.5%, owing to the uncertainty involved in the conversion process from sonar records to the reliable measurements of the ice thickness (Strass 1998). In addition, moored upward-looking sonars cannot record the time series of the snow depth and cannot be utilized to monitor the sea ice thickness in shallow alongshore areas. The accuracy of the acoustic sensor is ±0.005 m when it was not utilized for measurements, in other words, when not concerned with the measurement conditions (Richter-Menge et al. 2006). However, it has been demonstrated that the accuracy of the below-ice acoustic sensor depends on the resolution of the acoustic velocity in ocean water (specifically, the physical properties such as density, salinity, and composition). In such cases, the practical accuracy of this acoustic sensor would behave somewhat poorly, possibly ascending to more than 0.01 m, although it is deployed slightly below the ice bottom, ideally around 1–2 m.

With these methods we have found that it is difficult to measure the ice thickness to a millimeter level of precision. If longer time intervals are considered, the accuracy of some traditional methods would be enhanced (Perovich et al. 1997; Perovich and Elder 2002), but we would lose some short-term information. Thus, the error might be significant, when based on these traditional methods, in attempting to measure sea ice thickness with a growth rate of only a few millimeters per day—for instance, landfast sea ice around the Antarctic continent in winter (Purdie et al. 2006) and sea ice in the Arctic in autumn or spring (Perovich et al. 2003). Therefore, it has been necessary to develop a new automatic apparatus for fixed-site monitoring of the sea ice thickness with a high accuracy and a high temporal resolution. This objective has now been achieved with the development of a new apparatus. This apparatus operates most effectively for undeformed ice and is based on the magnetostrictive-delay-line (MDL) principle. The application of the apparatus is very good for thickness monitoring (with several sites) and for conducting basic research into the physics of sea ice growth and melting. This apparatus was utilized to monitor landfast sea ice thicknesses around Zhongshan Station, East Antarctica, for about 6 months during the austral autumn and winter of 2006.

This paper discusses the principles of the design, the configuration of this apparatus, and the results from its first field applications in Antarctica. The apparatus’s precision in the field is also described.

2. Design principles

According to the MDL technique, a positioning sensor that is able to detect the positions of a variety of moving permanent magnets set at various distances from the sensing elements has been developed; its operating principles were reported in detail by Hristoforou et al. (1997) and Karagiannis et al. (2003). The sensor is illustrated in Fig. 1. Responding to the stimulation of the pulse current, the circular magnetic field is shaped by the magnetic loop. Subsequently, the annular torsional distortion would come into being and be spread to both ends of the MDL in the form of a torsional wave. The echoes of the torsional waves can be detected by a checking machine. The position of the magnetic loop 1, L1, can be expressed as L1 = 0.5υ t1; here, υ is the propagation velocity of the torsional wave and t1 is the period from when the stimulating pulse appeared up to the time the echo reached the checking machine. Since this propagation velocity of the torsional wave is steady in different circumstances, this sensor has the characteristic of providing high-precision position detection. In addition, multiple positions can be detected synchronously when the moving magnetic loops are set in different positions. This positioning sensor has been used widely for measurements of, for example, the positions of pneumatic pistons, liquid levels, and absolute ground velocity (Hristoforou and Chiriac 2002; Hristoforou et al. 2006). Based on the principles of the MDL positioning sensor and the special requirements of environmental suitability, automatic measurement, and low-power dissipation for field applications, a new apparatus for monitoring sea ice thickness and its corresponding software has been designed.

This monitoring apparatus is composed of an apparatus box and two measuring poles, as shown in Fig. 2. The outsides of the poles are constructed of white polyethylene hollow rods, which also have been used as the deployment housing of the thermistor strings for deployment in sea ice or glacier ice (Preunkert and Wagenbach 1998; Pringle et al. 2007). This device avoids preferential freezing or melting, as its thermal conductance is very small. The diameter (0.02 m) of the poles is so small that the disturbances of the poles to the heat flux exchange at the snow surface and the ice bottom, and the wind shielding at the snow surface, are at the same levels as the housed rod of the acoustic sensor or the thermistor string. There were two stainless steel tubes fixed in the polyethylene hollow rods, which strengthen the measuring poles consumedly and prevent the measuring poles from bending under external forcing, especially during strong winds. The horizontal space between these two poles is 0.15 m. The apparatus box and the measuring poles are connected by an air pipe and a cable. The air pipe is used to connect the cylinder, which is fixed in the apparatus box, with the buoy, which is equipped in the lower magnetic loop; the cable is used to connect a battery pack with a miniwinch and to transmit data records to the datalogger, which is fixed in the apparatus box. The air pipe leads to one stainless steel tube in the measuring pole, and the MDL is set in one stainless steel tube in the other measuring pole. Two moving magnetic loops are fixed in the movement machines along the measuring poles, and one settled magnetic loop is fixed at the top of the measuring pole. The measuring operation is completed by controlling the movement of the magnetic loops. When the measuring operation is performed, the upper magnetic loop, which is jointed with a heavy hammer with a total weight of 0.15 kg, moves down under the control of the miniwinch at a speed of 0.02 m s−1. When the underpan of the upper magnetic loop, with a size of 0.005 m2, is lowered to the surface of the ice/snow surface, the miniwinch stops immediately, in order to avoid depressing the snow surface or causing any change in the snow structure. At the same time the buoy will be filled with air supplied by the cylinder through the air pipe. Afterward, the lower magnetic loop will move up under the resultant force between gravitation and buoyancy, and then its overpan, also with a size of 0.005 m2, is finally set under the bottom of the ice. The underpan of the upper magnetic loop and the overpan of the lower magnetic loop could ensure that the magnetic loops are parallel to the ice–snow surface or the ice bottom and minimize the effects of the unevenness of ice–snow surface or ice bottom on the resolution of the measurement records. The distance from both the upper and lower magnetic loop to the settled magnetic loop is detected by the MDL positioning sensor. The distance between the upper and lower magnetic loop is the total thickness, including sea ice and snow cover. The variations of the ice/snow surface and the ice bottom are calculated by comparing the positions of the upper and lower magnetic loop with their original values, respectively. The measurement data then will be transmitted to the datalogger.

When the detection operation is completed, the upper magnetic loop will move up under the draft of the miniwinch, which is assembled at the top of the measuring poles, through a steel-wire string; meanwhile, the lower magnetic loop will move down after the air in the buoy has been sent back to the cylinder through the air pipe. This process, after detection, not only prevents the upper magnetic loop from being covered by snow and the lower magnetic loop from being frosted under ice, but also minimizes the possibility of the disturbance to the heat flux exchange at the snow surface and the ice bottom. The miniwinch, and the air pump that is assembled in the cylinder, are driven by the power of the battery packs. The PC software used includes functions for restoring and cleaning up the data in the datalogger, displaying and adjusting a real-time clock, displaying the temperature of the battery packs, controlling the charge, and updating the program.

The low-temperature operating environment has been taken into account adequately. Thus, the apparatus box is characterized by heat insulation and corrosion prevention, and it is waterproof. A special latex material was adopted for the buoy, cylinder seal, and air pipe to prevent problems arising from hardening and other changes brought about by low temperatures. The battery packs can withstand a temperature of −30°C for more than 10 days when the measurement interval is set for 10 min, in one power-supply cycle.

Based on capability testing in the laboratory, the principal technical criteria for this apparatus are as follows: 1) the resolution is 0.0001 m, 2) the design precision of the MLD positioning sensor assembled in the apparatus is ±0.001 m, 3) the temperature range is from −55° to 50°C, and 4) the measuring interval is 10–180 min.

3. In situ experiments

a. Instrument disposition

The monitoring apparatus was utilized to measure landfast sea ice thickness around Zhongshan Station, East Antarctica, during the austral autumn and winter of 2006. The apparatus was disposed at the site (69°22′9″S, 76°21′45″E) with a water depth of 9 m on 27 March 2006 [yearday (YD) 85]. The deployment work is similar to that involved in deploying the array of thermistors in sea ice (Pringle et al. 2007). When the apparatus was installed, there were 0.01 m of snow, 0.25 m of ice, and 0.025 m of freeboard. After a vertical ice hole of 0.25 m diameter had been drilled through the thickness of the ice cover, and the ice brash in the hole had been cleaned out, the measuring poles were fixed in the hole. The measurement ranges for detecting the ice/snow surface and the ice bottom were 0.75 and 2.25 m, respectively. The apparatus box was deposited on sea ice, with a 2-m distance from the measuring pole, perpendicular to the dominant wind direction, to avoid wind shielding. The deployment work, including ice-hole drilling, apparatus fixing, and apparatus testing, was supported by three winter-over crew members from Zhongshan Station, and was finished in 1 h. Figure 3 presents a field photograph of the apparatus, which is denoted by MDL.

Three days later the offset between the record obtained by the apparatus and the record derived from the drill hole at a distance of 1 m from the apparatus was less than 0.03 m. It was proven that the ice hole was naturally refreezing. Subsequently, the portion of the hole above the water level was artificially frozen in by infilling water. The disturbance of the deployment hole in the measure records can be limited in the first week after deployment. From April onward, the in situ measurements were carried until 21 September 2006 (YD 263).

In addition, for comparison, 16 thermal-wire thickness gauges consisting of a stainless steel wire with a steel rod attached as a crossbar on the bottom end and a wooden handle on the top end, spaced every 15 m along a section, were installed near our apparatus in late March 2006. The principles of this method for monitoring the sea ice mass balance have been described in detail by Untersteiner (1961) and Perovich et al. (2003). To make a measurement, first, the stainless steel wire has to be loaded a 36-V dc and then its electrical resistance is melted free. Finally, it can be pulled up and read against a scale on a gauge anchored into the ice, when the crossbar on its bottom end moves up against the ice bottom. The gauges also were used as snow-depth stakes. The uncertainty levels of the stake and gauge readings were typically less than ±0.005 m. These thermal-wire thickness gauges are denoted by TWTG in Fig. 3.

The deployment site was periodically visited to retrieve data, change the battery pack, and check the operational status of the apparatus for ensuring the acquisition of uninterrupted data, every 7–10 days according to meteorological conditions, although the battery pack could usually function for more than 20 days. Sea ice at the deployment site was undeformed, no snow ice formation occurred, and the freeboard was positive during the experiment. The operation of the apparatus was hampered by the extremely harsh field conditions, which were characterized by low air temperature, frequent snowstorms, periodic variations in water levels and tide currents, the unintentional disturbance of animal life under the ice, etc. The field conditions during the experiment are summarized in Table 1.

With a diameter of 0.02 m, the measuring poles sound rather flimsy, but the degree of its flexure during the strongest wind with a speed reach to 25.9 m s−1 was less than 5°, and no perpetual distortion occurred after any gale event for the measuring poles. By the end of our field experiment, however, the measuring poles remained vertical to the ice surface. Thus, we can consider the intensity of the measuring poles to be sufficient to resist the external forcings in the field.

Some mechanical malfunctions of the apparatus occurred during the experiment; for example, the lower magnetic loop was obstructed by an unidentified object under the sea ice, and the upper magnetic loop was obstructed by snow adhering along the measuring poles when measurements were taken. These malfunctions would eventually cease by themselves, but a few records were lost. The steel-wire string, which is used to connect the lower magnetic loop with the miniwinch, was destroyed during the snowstorm event. This malfunction was dealt with by changing the steel string when the field support crew visited the deployment site. Fortunately, this malfunction occurred only 2 times during the experiment. These mechanical malfunctions can be solved without any questions in the future by augmenting the space between the moving magnetic loops and the measuring poles; choosing a new material for the string; and optimizing the connection among the miniwinch, the steel-wire string, and the lower magnetic loop.

b. Experimental results

During the experiment the measurement interval was set for 3 h except for a few days when it was set for 2 h. In all, 1460 records for the ice/snow surface and 1460 records for the ice bottom were obtained. Based on field observations, the growth rate was not more than 0.05 m day−1, and the intraday variations of the ice/snow surface were not more than 0.2 m during the experiment; thus, the record for the ice bottom is rejected where the offset is more than 0.05 m and the record for the ice/snow surface is rejected where the offset is more than 0.2 m when compared with their daily average values, respectively. After these rejections, 1352 records for the ice/snow surface and 1368 effective records for the ice bottom remained. The effective fractions of the data records for detecting the ice/snow surface and the ice bottom were 92.6% and 93.7%, respectively. All the obviously erroneous records were caused by mechanical malfunctions of the instrument as described in the above paragraphs. When the upper magnetic loop was obstructed and could not reach the snow surface, the snow depth was recorded as being so much greater than could be believed. Likewise, when the lower magnetic loop was obstructed and could not reach the ice bottom, the data were also beyond belief. Thus, this small amount of obviously erroneous data was rejected easily and did not interfere with the process of identifying and analyzing the ice growth and snow accumulation processes. Therefore, this apparatus can be considered to be highly efficient and reliable when operating under extreme polar field conditions. Sea ice growth rates were calculated by computing the first derivatives of the curves of the time series of ice thickness using
i1520-0426-26-4-818-e1
where Hi is the ice thickness, tj is the time of one measurement, and tj+1 is the time of the next measurement (Perovich et al. 2003). The sea ice growth rate was in the range of 0–0.0173 m day−1 during the experiment. The monthly average sea ice growth rate is greatest in May (from YD 120 to 150) with a value of 0.0106 m day−1; the monthly average sea ice growth rate is smallest in August (from YD 212 to 242) with a value of 0.0058 m day−1, as shown in Figs. 4a and 4b. From Fig. 4, it can be seen that the ice growth rate was in the range of 0.0019–0.0173 m day−1 in May and in the range of 0.0017–0.0093 m day−1 in August. These results indicate that the ice growth rate was only a few millimeters both in the autumn months and in the winter months; furthermore, the ice growth rate was markedly unstable because of the fluctuation of the air temperature and oceanic heat flux, as well as the seasonal variations of solar radiation, etc. During the experiment from March to September, the mass balance was manually measured every 1–2 days using the thermal-wire gauges, and 132 sets of records including ice thickness and snow depth were obtained. The variations of the average ice thickness and average ice growth rate according to these 16 thermal-wire gauges in May and August are shown in Figs. 4c and 4d. Although the monthly average ice growth rate derived from the thermal-wire gauges is close to that derived from the apparatus, there still is some artificial information in the records derived from the thermal-wire gauges when taking into account the daily ice growth rate. For instance, the data records from the thermal-wire gauges show that an ice-melting event occurred on 29 August 2006 (YD 240), and that the stability and continuity of the ice growth rate was artificially weakened remarkably. Thus, the natural properties of ice growth would be distorted by the traditional measurement methods owing to their low precision and temporal resolution. However, considering its high precision and temporal resolution, our apparatus can help us to gain better accuracy on an hourly basis when estimating the equivalent latent heat flux from ice thickness measurements and also when estimating the ocean heat flux under sea ice.

Solid precipitation was observed rarely during the experiment. There was a gale event, with wind speeds higher than 17 m s−1 following a snowfall event. Thus, the depth of the snow cover that accumulated on the sea ice surface around the apparatus was very small throughout the experiment. Nevertheless, there was a short but strong snowfall event that lasted from 30 to 31 July. The time series of the wind speed, from 29 July (YD 209) to 3 August (YD 214) in 2006, is plotted in Fig. 5a. The time series of the snow depth derived from the apparatus, and the thermal-wire gauge that was installed nearest to the apparatus at a distance of 5 m, during the same period, is shown in Fig. 5b. As the plot in Fig. 5a shows, the wind speed increased distinctly after 1700 local time (LT) on 31 July (YD 211). Although the snow depth increased before 1700 LT on 31 July, owing to the solid precipitation, and decreased subsequently at a good pace owing to wind drift, the rapid change in snow depth was captured accurately by our apparatus. However, the rapid change in snow depth would be missed by the thermal-wire gauge measurements because of its lower temporal resolution.

c. Precision estimation

As the snow depth was small throughout the experiment, only the precision estimation for records of the ice bottom is discussed here. Because there was no ice surface ablation or snow ice formation during the experiment, this precision estimation of the ice bottom can be recognized as that of ice thickness.

In an attempt to investigate the precision of the ice thickness records, it must be made certain that the data are valid primarily. For this investigation, a total of 24 drill-hole records were collected intermittently over a distance of 2 m around our apparatus. The ice thickness holes were drilled by a slim drill with a stainless steel auger 0.05 m in diameter. The drill-hole records, with a precision of ±0.005 m, ranged from 0.335 to 1.570 m. These drill-hole records and the apparatus records, obtained at the same time, are shown in Fig. 6. The overall absolute dispersion between the drill-hole records and the apparatus records is 0.0015 ± 0.001 m. It is proven that the apparatus records are valid, based on the precision of drill-hole records and the natural spatial variations in both the ice surface and ice bottom.

As the precision and the time resolution of both the drill-hole records and the thermal-wire thickness gauges records were not very high, and there were some spatial variations in both the ice surface and ice bottom, the precision of the apparatuses in the field cannot be estimated by comparison with the drill-hole records or thermal-wire thickness gauges records. To estimate the precision of our apparatuses subtlety, an approach whereby the growth rate of the sea ice on a certain day is constant has been assumed. Based on this approach, a linear regression according to the least squares method is first applied for data records derived from a particular day, then the slope of the linear-regression trend is recognized as the growth rate of the sea ice on that day, and finally the precision of the monitoring apparatuses in the field on that day can be estimated as follows:
i1520-0426-26-4-818-e2
i1520-0426-26-4-818-e3
i1520-0426-26-4-818-e4
where Eri, Erj, and Er are the positive error, the negative error, and the general error, respectively; Thi, , and n are the ice thickness records, which are larger than their estimated linear-regression values, the estimated linear-regression values according to Thi, and the number of Thi, respectively; and Thj, , and m are the ice thickness records, which are smaller than their estimated linear-regression values, the estimated linear-regression values according to Thj, and the number of Thj, respectively.

For example, the positive error, the negative error, and the general error on 20 May 2006 (YD 139, denoted by the arrow in Fig. 7) are 0.0004, −0.0005, and 0.0005 m, respectively. The maximal value among the general errors on all field days can be recognized as the precision of the apparatuses during the experiment.

The errors from all field days have been estimated, as shown in Fig. 7. The solid lines denote the positive error and the negative error, and the dashed line denotes the linear-regression datum line, which expresses a link among the linear-regression lines on all field days, and has been transferred from the original values to the zeros. The maximal absolute value of both the positive error and the negative error is 0.002 m. This estimation approach leads to a slightly conservative estimation of the precision of the apparatuses, because some slight variations in the rate of ice growth are evident even in the intraday period, and the estimating strategy according to Eqs. (2)(4) seems to be slightly rigorous. In other words, the actual precision of the apparatuses may be somewhat better than the estimated values, and this estimated value can be considered to be the upper limit of the actual precision of the apparatuses in the field. When applying the moving average value of every eight records in every day instead of the estimated linear-regression values in Eqs. (2) and (3), the positive error and the negative error are 0.0018 and 0.0017 m, respectively.

The estimated precision of the apparatus is slightly worse than the designed precision of ±0.001 m for the MDL positioning sensor assembled in our apparatus, owing to the extremely harsh field conditions in Antarctica. However, there is no doubt that this apparatus has the capacity to detect variations in ice/snow thickness with high precision at the millimeter level and high temporal resolution on an hourly basis.

4. Concluding remarks

Our new apparatus for monitoring ice/snow thickness based on MDL principles, which can be used under harsh environmental conditions in polar regions and in shallow seas, addresses the low efficiency of the traditional manual methods effectively. The effects of the sea ice physical properties and circumstantial conditions on the measure resolution, which are difficult to avoid when using the techniques of satellite remote sensing, submarine sonar profiling, moored upward-looking sonars, electromagnetic sounding, radar penetration, and acoustic sensors, can be minimized by using this apparatus. Practice in the field proves that the precision of this apparatus is ±0.002 m, which is more accurate than any records derived from other techniques; the effective fractions of the records for detecting the ice/snow surface and the ice bottom are 92.6% and 93.7%, respectively.

Based on current techniques, this apparatus cannot replace the acoustic sensor equipped in the IMB because the equipping battery pack can only operate for 10–20 days, but this technical problem can be resolved by improving the driving mechanism of the moving magnetic loops and strengthening the capability of the battery pack, which then could be installed in the IMBs, enhancing the quality of future data records.

This apparatus also has some limits for field applications: 1) it can be utilized only for fixed-site undeformed ice, but cannot measure ice thickness along a line or in a large area, or be used to estimate profiles of ice ridges; and 2) although this apparatus has the ability to detecting ice/snow surface melting, it cannot be utilized to detect how snow contributes to the sea ice mass balance.

In conclusion, this monitoring apparatus could have a significant role to play in the family of ice/snow thickness monitoring techniques. The measured data have the potential to express, at the basic-research level, the ice growth and decay rate; to estimate the ocean heat flux coupled with sea ice temperature measurements obtained from other instruments; and to supply evidence for the need to modify certain sea ice numerical models.

Acknowledgments

This study was supported by the National Natural Science Foundation of China under Contracts 40676001 and 40233032, and by the National Key Technology Research and Development Program under Contract 2006BAB18B03. We are grateful to the Chinese Arctic and Antarctic Administration for its logistic support during our experiment; to Mr. Li Dong and Mr. Xiufeng Liu, who were part of the 2006 winter-over crew at Zhongshan Station in the 22nd CHINARE, for their field-work support; to the Meteorological Office of Zhongshan Station for providing meteorological data; and to the National Tidal Centre, Australian Bureau of Meteorology, for providing tide-prediction data. Three anonymous reviewers are thanked for their comments, which considerably improved this work.

REFERENCES

  • Bergeron C. J. Jr., , , Brusstar J. R. , , Yi N. , , Wu Y. , , and Ioup J. W. , 1999: A new vertical continuation procedure for airborne electromagnetic field data from the modified image method. Geophysics, 64 , 13641368.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Eicken, H., , Grenfell T. C. , , Perovich D. K. , , Richter-Menge J. A. , , and Frey K. , 2004: Hydraulic controls of summer Arctic pack ice albedo. J. Geophys. Res., 109 , C08007. doi:10.1029/2003JC001989.

    • Search Google Scholar
    • Export Citation
  • Haas, C., 1998: Evaluation of ship-based electromagnetic-inductive thickness measurements of summer sea-ice in the Bellingshausen and Amundsen Seas, Antarctica. Cold Regions Sci. Technol., 27 , 116.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Heil, P., 2006: Atmospheric conditions and fast ice at Davis, East Antarctica: A case study. J. Geophys. Res., 111 , C05009. doi:10.1029/2005JC002904.

    • Search Google Scholar
    • Export Citation
  • Heil, P., , Allison I. , , and Lytle V. I. , 1996: Seasonal and interannual variations of the oceanic heat flux under a landfast Antarctic sea ice cover. J. Geophys. Res., 101 , (C11). 2574125752.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hristoforou, E., , and Chiriac H. , 2002: Position measuring system for applications in field sports. J. Magn. Magn. Mater., 249 , 407410.

  • Hristoforou, E., , Chiriac H. , , Neagu M. , , and Karayannis V. , 1997: On the calibration of position sensors based on magnetic delay lines. Sens. Actuators, 59A , 8993.

    • Search Google Scholar
    • Export Citation
  • Hristoforou, E., , Dimitropoulos P. D. , , and Petrou J. , 2006: A new position sensor based on the MDL technique. Sens. Actuators, 132A , 112121.

    • Search Google Scholar
    • Export Citation
  • Hudson, R., 1990: Annual measurement of sea ice thickness using an upward-looking sonar. Nature, 344 , 135137.

  • Karagiannis, V., , Manassis C. , , and Bargiotas D. , 2003: Position sensors based on the delay line principle. Sens. Actuators, 106A , 183186.

    • Search Google Scholar
    • Export Citation
  • Kawamura, T., , Ohshima K. I. , , Takizawa T. , , and Ushio S. , 1997: Physical, structural, and isotopic characteristics and growth processes of fast sea ice in Lützow-Holm Bay, Antarctica. J. Geophys. Res., 102 , (C2). 33453355.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Laxon, S., , Peacock N. , , and Smith D. , 2003: High interannual variability of sea ice thickness in the Arctic region. Nature, 425 , 947950.

  • Liu, J., , Schmidt G. A. , , Martinson D. G. , , Rind D. , , Russell G. , , and Yuan X. , 2003: Sensitivity of sea ice to physical parameterizations in the GISS global climate model. J. Geophys. Res., 108 , 3053. doi:10.1029/2001JC001167.

    • Search Google Scholar
    • Export Citation
  • Mélia, D. S., 2002: A global coupled sea ice–ocean model. Ocean Modell., 4 , 137172.

  • Perovich, D. K., , and Elder B. C. , 2001: Temporal evolution of Arctic sea-ice temperature. Ann. Glaciol., 33 , 207211.

  • Perovich, D. K., , and Elder B. C. , 2002: Estimates of ocean heat flux at SHEBA. Geophys. Res. Lett., 29 , 1344. doi:10.1029/2001GL014171.

  • Perovich, D. K., , and Richter-Menge J. A. , 2006: From points to poles: Extrapolating point measurements of sea-ice mass balance. Ann. Glaciol., 44 , 188192.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perovich, D. K., , Elder B. C. , , and Richter-Menge J. A. , 1997: Observation of the annual cycle of sea ice temperature and mass balance. Geophys. Res. Lett., 24 , 555558.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perovich, D. K., , Grenfell T. C. , , Richter-Menge J. A. , , Light B. , , Tucker W. B. III, , and Eicken H. , 2003: Thin and thinner: Sea ice mass balance measurements during SHEBA. J. Geophys. Res., 108 , 8050. doi:10.1029/2001JC001079.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Perovich, D. K., , Elder B. C. , , Claffey K. J. , , Stammerjohn S. , , Smith R. , , Ackley S. F. , , Krouse H. R. , , and Gow A. J. , 2004: Winter sea-ice properties in Marguerite Bay, Antarctica. Deep-Sea Res. II, 51 , 20232039.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Preunkert, S., , and Wagenbach D. , 1998: An automatic recorder for air/firn transfer studies of chemical aerosol species at remote glacier sites. Atmos. Environ., 32 , 40214030.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pringle, D. J., , Eicken H. , , Trodahl H. J. , , and Backstrom L. G. E. , 2007: Thermal conductivity of landfast Antarctic and Arctic sea ice. J. Geophys. Res., 112 , C04017. doi:10.1029/2006JC003641.

    • Search Google Scholar
    • Export Citation
  • Purdie, C. R., , Langhorne P. J. , , Leonard G. H. , , and Haskell T. G. , 2006: Growth of first-year landfast Antarctic sea ice determined from winter temperature measurements. Ann. Glaciol., 44 , 170176.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Richter-Menge, J. A., , Perovich D. K. , , Elder B. C. , , Claffey K. , , Rigor I. , , and Ortmeyer M. , 2006: Ice mass-balance buoys: A tool for measuring and attributing changes in the thickness of the Arctic sea-ice cover. Ann. Glaciol., 44 , 205210.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rothrock, D. A., , Yu Y. , , and Maykut G. A. , 1999: Thinning of the Arctic sea-ice cover. Geophys. Res. Lett., 26 , 34693472.

  • Saloranta, T. M., 2000: Modeling the evolution of snow, snow ice and ice in the Baltic Sea. Tellus, 52A , 93108.

  • Smedsrud, L. H., , Budgell W. P. , , Jenkins A. D. , , and Ådlandsvik B. , 2006: Fine-scale sea-ice modelling of the Storfjorden polynya, Svalbard. Ann. Glaciol., 44 , 17.

    • Search Google Scholar
    • Export Citation
  • Strass, V. H., 1998: Measuring sea ice draft and coverage with moored upward looking sonars. Deep-Sea Res. I, 45 , 795818.

  • Sun, B., , Wen J. , , He M. , , Kand J. , , Luo Y. , , and Li Y. , 2003: Sea ice thickness measurement and its underside morphology analysis using radar penetration in the Arctic Ocean. Sci. China, 46D , 11511160.

    • Search Google Scholar
    • Export Citation
  • Tamura, T., , Ohshima K. I. , , Markus T. , , Cavalieri D. J. , , Nihashi S. , , and Hirasawa N. , 2007: Estimation of thin ice thickness and detection of fast ice from SSM/I data in the Antarctic Ocean. J. Atmos. Oceanic Technol., 24 , 17571772.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Untersteiner, N., 1961: On the mass and heat budget of Arctic sea ice. Arch. Meteor. Geophys. Bioklimatol., 12A , 151182.

  • Vavrus, S., , and Harrison S. P. , 2003: The impact of sea-ice dynamics on the Arctic climate system. Climate Dyn., 20 , 741757.

  • Vinnikov, K. Y., and Coauthors, 1999: Global warming and Northern Hemisphere sea ice extent. Science, 286 , 19341937.

Fig. 1.
Fig. 1.

The MDL positioning sensor.

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Fig. 2.
Fig. 2.

The monitoring apparatus for sea ice thickness.

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Fig. 3.
Fig. 3.

The monitoring apparatus (denoted by MDL) and the thermal-wire thickness gauges (denoted by TWTG).

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Fig. 4.
Fig. 4.

Variations in sea ice thickness and growth rate in May and August 2006: (a),(b) data records derived from the apparatus and (c),(d) average data records derived from the thermal-wire gauges. The arrow in (a) denotes the record on 20 May (YD 139).

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Fig. 5.
Fig. 5.

Time series of (a) wind speed and (b) snow depth, derived from 29 Jul to 3 Aug 2006 (YDs 209–214).

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Fig. 6.
Fig. 6.

Sea ice thickness records obtained from drill holes and the monitoring apparatus during the experiment.

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Fig. 7.
Fig. 7.

Positive error, negative error, and the linear-regression datum line of records obtained from the monitoring apparatus during the experiment (arrows denotes an error on 20 May, YD 139).

Citation: Journal of Atmospheric and Oceanic Technology 26, 4; 10.1175/2008JTECHO613.1

Table 1.

Field conditions during the experiment.

Table 1.
Save