5. Archived Data Processing and Data Format

5.1 Data Processing of J-CAD
  Each sensor mounted on the J-CAD measures oceanographic, meteorological, and buoyﾅ@status data every hour. The data obtained from each sensor are processed in the J-CAD and transmitted to the laboratory via the ORBCOMM and/or Argos satellite communications. Table 5.1 shows resolutions, the minimum, and maximum value of each data, which are used for the data conversion between physical value and bit value for transmission via satellite communication.

Table 5.1. Resolution, minimum and maximum value of J-CAD data
Description
Unit
Resolution
Min
Max
Regard
Air temperatureºC0.25-5013.75*1
Barometric pressureHPa0.19501052.3 
Wind directiondeg1.4060358.53 
Wind speedm/sec0.25063.75 
Water temperature in the hullºC0.13588-259.6494*1
Platform headingdeg1.4060358.53 
Platform tiltdeg0.5031.5 
LatitudeºN0.00044092.4284 
LongitudeºE0.00035-180187.00125 
GPS drift speedcm/sec0.50127.5 
GPS drift directiondeg1.4060358.53 
Conductivity of CT sensor
 		mmho/cm
 		0.00037
 		2.5
2.0		4.0
2.5		J-CAD 1,2
J-CAD 3 & later
Temperature of CT sensorºC0.00061-23.49877*1
CTD1 depthm0.078105124.89 
CTD2 depthm0.078235254.89 
ADCP bin velocitycm/sec0.39-10099.29 
ADCP intensity(none)8.22580254.9998 
ADCP percent good(none)2040100 
ADCP headingdeg0.3520360.096 
ADCP pitch, rolldeg0.156863-2020.000065 
ADCP temperatureºC0.156-27.828*1
Watson compass headingdeg0.3520360.096J-CAD 4
Watson compass pitch, rolldeg0.156863-2020.000065J-CAD 4
Watson compass X, Y-magnetometermGauss2-250260J-CAD 4
Watson compass Z-magnetometermGauss2250760J-CAD 4
*1 These data fields roll over once reaching the minimum/maximum value.

  Oceanographic, meteorological, and buoy status data observed by the J-CADs are sent to JAMSTEC via the ORBCOMM and/or Argos satellite system. The data received at JAMSTEC are automatically processed and complied into each component of data in real time. The processing system runs in the following order: data acquisition, data format conversion, correction and estimation of time and position data, abnormal data correction, updating of monthly files made from raw data, and monthly component data file creation. In addition, graphs of the J-CAD data are created to release on the Internet
(http://www.jamstec.go.jp/arctic/J-CAD_e/jcadindex_e.htm).
  In the following sections, we provide monthly component data from each J-CAD and data format information. Monthly component data include meteorological (jX_yyyymm_met.txt), buoy status (jX_yyyymm_stat.txt), CTD (jX_yyyymm_ctd.txt), and ADCP (jXyyyymm_adcp1.txt or jXyyyymm_adcp2.txt) data. All text data are CSV format and the files can be read using Microsoft Excel.

5.2 Weather Information data (jX_yyyymm_met.txt)
  Weather information data consist of observational time, J-CAD number, latitude, longitude, air temperature, air pressure, wind direction, and wind speed. Below is an example of the weather information data for 3 hours.

1line
Year, Month, Day, Hour, JCAD Number, Latitude, Longitude, Air Temperature, Air Pressure, Wind Direction, Wind Speed, Flg(position),
2line(1st hour)
2001, 11, 30, 5, 4, 44.6960, -63.62500, 4.75, 1027.0, 144.898, 7.00, 0,
3line(2nd hour)
2001, 11, 30, 6, 4, 44.6960, -63.62500, 5.25, 1026.5, 135.056, 5.25, 0,
4line(3rd hour)
2001, 11, 30, 7, 4, 44.6960, -63.62500, 6.00, 1025.1, 136.462, 4.75, 0,
  The first line is the header, and from the second line, each line shows an hourly data of every item.
Year, Month, Day, Hour;    Observational Time
J-CAD Number; Deployment number of J-CAD
Latitude, Longitude; Observational position measured by GPS or Argos position data services. Refer to the flag for buoy position.
Air Temperature; Air temperature measured by the thermistor inside the sensor mast.
Air Pressure; Air pressure measured by the barometer installed inside the J-CAD platform.
Wind Direction, Speed; Wind direction and speed measured by Young anemometer Wind direction is defined as true azimuth
Flag (position); “0” indicates the data as measured by GPS and "2" asﾅ@calculated from ARGOS position data service of class 2 or 3. The flag "4" is used for only J-CAD 1 and this flag indicates the data calculated from Argos positional data of CRREL ice buoy which was deployed close to J-CAD 1.


5.3 Buoy status data (jX_yyyymm_stat.txt)
  Buoy status data consist of time, J-CAD number, latitude, longitude, temperature of water surrounding the J-CAD hull (sea surface temperature), platform heading, platform tilt, battery voltage of main electronics and ORBCOMM system, time to first fix, GPS signal strength, GPS drift speed, GPS drift direction (true azimuth). Below is an example of the buoy status information data for 2 hours.

1line
Year, Month, Day, Hour, JCAD Number, Latitude, Longitude, Sea surface Temperature, Platform Heading, Platform Tilt, Battery Voltage, ORBCOMM Voltage, Time to first fix, GPS signal strength, GPS Drift Speed, GPS Drift Direction, Flg(position), Heading(nocorr),
2line(1st hour)
2001, 11, 30, 6, 4, 44.6960, -63.62500, 6.79592, 29.606, 2.0, 17.10, -999, 40, 49, 0.5, 73.112, 0, 49.210,
3line(2nd hour)
2001, 11, 30, 7, 4, 44.6960, -63.62500, 6.79592, 29.606, 2.5, 17.10, -999, 30, 47, 3.5, 160.284, 0, 49.210,
  The first line is the header, and from the second line, each line shows an hourly data of each item.
Year, Month, Day, Hour; Observational Time
J-CAD Number; Deployment number of J-CAD
Latitude, Longitude; Observational position measured by GPS or Argos position data services. Refer to the flag for buoy position.
Sea surface temperature; Temperature of water surrounding the J-CAD hull
Platform heading, tilt; Platform heading and vertical tilt measured by a three-axis magnetometer on the electric support channel inside the J-CAD hull. Heading is defined as true azimuth.
Battery voltage; Battery voltage for J-CAD system
ORBCOMM voltage; Battery voltage for ORBCOMM satellite communication system
GPS drift speed, direction;    Buoy drift speed and direction measured by GPS
Flag (position); "0" indicates the data as measured by GPS and "2" as calculated from ARGOS position data service of class 2 or 3.
Heading (nocorr); Platform heading without any correction. Heading data indicates a buoy direction to magnetic north.


5.4 CTD data (jX_yyyymm_ctd.txt)
  CTD data consist of time, location, weather information, data sampling depth, pressure, water temperature, and salinity. Below is a sample CTD data for 2 hours.
1line
Year, Month, Day, Hour, JCAD Number, Latitude, Longitude, Air Temperature, Air Pressure, Flg(est-prs), Flg(position), datano,
2line
Depth, Pressure, Temperature, Conductivity, Salinity, Potential Temp, Density,
3line(1st hour)
2001, 11, 30, 6, 4, 44.6960, -63.62500, 5.25, 1026.5, 01, 0, 6,
4line
5.476, 5.521, 7.04447, 3.12582, 30.5573, 7.04569, 1023.915,
5line
10.952, 11.042, 7.15915, 3.14448, 30.6546, 7.15991, 1023.977,
6line
17.523, 17.667, 7.00482, 3.15839, 30.9398, 7.00498, 1024.221,
7line
26.284, 26.500, 6.76814, 3.15473, 31.1096, 6.76752, 1024.385,
8line
39.425, 39.750, 6.57355, 3.14595, 31.1844, 6.57182, 1024.469,
9line
54.754, 55.208, 6.17888, 3.12838, 31.3454, 6.17592, 1024.644,
10line(2nd hour)
2001, 11, 30, 7, 4, 44.6960, -63.62500, 6.00, 1025.1, 01, 0, 6, 5.476, 5.521, 7.01214, 3.12033, 30.5266, 7.01335, 1023.895,
10.952, 11.042, 7.07619, 3.13606, 30.6374, 7.07694, 1023.974,
17.523, 17.667, 6.98774, 3.15985, 30.9711, 6.98790, 1024.248,
26.284, 26.500, 6.70531, 3.15144, 31.1307, 6.70469, 1024.410,
39.425, 39.750, 6.55342, 3.14595, 31.2027, 6.55169, 1024.486,
54.754, 55.208, 6.33199, 3.13533, 31.2814, 6.32901, 1024.575,
  The first and second lines are the header of the file. The first line of an hourly data consists of year, month, day, hour, J-CAD number, latitude, longitude, air temperature, air pressure, flag for estimated pressure, flag for positional information and the total line number of an hourly data. The rest lines show the CT/CTD data. The output order is from the shallow depth. The CT/CTD data consist of depth, pressure, temperature, conductivity, salinity, potential temperature, and density. Only when either conductivity or water temperature is measured (and also that is not dummy data), the CT/CTD data are outputted.
Year, Month, Day, Hour;    Observational Time
J-CAD Number; Deployment number of J-CAD
Latitude, Longitude; Observational position measured by GPS or Argos position data services. Refer to the flag for buoy position.
Air Temperature; Air temperature measured by the thermistor inside the sensor mast.
Air Pressure; Air pressure measured by the barometer installed inside the J-CAD platform.
Flag (est-pre); Flag for water depth estimation. Details of the estimation are described in the following subsection.
Flag (position); “0” indicates the data as measured by GPS and "2" as calculated from ARGOS position data service of class 2 or 3.
Data number; Number of hourly CT/CTD data shown in the following
Depth, pressure; (*) please refer to the following subsection.
Temperature; In situ temperature measured by CT/CTD sensor (SBE37IM)
Conductivity; Conductivity measured by CT/CTD sensor (SBE37IM)
Salinity; Salinity calculated from measured conductivity, temperature, and water pressure
Potential temperature; Potential temperature referred to the surface
Density; Density calculated from temperature, salinity, and depth


5.4.1 Flag for water depth estimation
  Water Pressure is measured by the CTD at 120m and 250m. The pressures where the CTs locate are estimated using the pressure measured by the CTD. It is called "the estimated pressure". Using the pressure measured by the CTD, the estimated pressure is calculated by liner interpolation. For instance, when the pressures at 120m and 250m measured by the CTDs are 120P and 250P, the pressure at 180m, P180 can be estimated as follows:
P180 = (P250-P120)/(250-120)*(180-120)(3)
  When the pressure only at 250m is measured, the pressure at 180m can be estimated as follows:
P180 = P250/250*180(4)
  The formula to be used and the CT pressure can be estimated depend on the CTD locations and the type of the J-CADs. For the details, refer to Table 5.2 and 5.3. In the Table, "CT which cannot be estimated", means that CT pressure estimate is unobtainable and the water pressure data is covered by the set up water depth. Flags indicate the information on the water pressure estimate. The CT pressure at 180m of the J-CAD2ﾅ`5 (flag11) is calculated using the formula (3) mentioned previously. The other CT pressures are calculated using the formula (4) (only parameters are different). Set up depth is applied to "the estimated depth" where the CT pressure cannot be estimated.

Table 5.2. Flag for water pressure estimation for J-CAD 1
Set up water depth (m)
Flag
CTD which can
measure pressure		CT (Pressure can be
estimated)		CT (Pressure cannot be
estimated)
X
x
20,130,175,250
00
250
20,130,175
X
10

Table 5.3. Flag for water pressure estimation for J-CAD 2~5
Set up water depth (m)
Flag
CTD which can
measure pressure		CT (Pressure can be
estimated)		CT (Pressure cannot be
estimated)
X
x
25,50,80,120,180,250
00
120
25,50,80,180,250
x
01
250
25,50,80,120,180
x
10
120,250
25,50,80,180
x
11

  The water depth is calculated using the constant and formula based on UNESCO 1983. The water depth determined from water pressure is called “the estimated depth”. For the estimations, latitude data are required. If the data have no positional information, set up depth is applied to the estimated depth.


5.5 ADCP data (jX_yyyymm_adcp1.txt or jX_yyyymm_adcp2.txt)
  Current data measured by each ADCP and other data related to currents are outputted. The data file named "jX_yyyymm_adcp1.txt" has the ADCP data mounted at 20 m or 25m in depth and the data file named “jX_yyyymm_adcp2.txt” has the data at 260 m. Below is a sample of J-CAD 4 ADCP data for 2 hours.

1line
Year, Month, Day, Hour, JCAD Number, Latitude, Longitude, Air Temperature, Air Pressure, Wind Direction, Wind Speed, Platform Heading, Platform Tilt, GPS Drift Speed, GPS Drift Direction, Flg(est-prs), Flg(position), Heading(nocorr),
2line
WC-pitch, WC-roll, WC-heading, WC-X, WC-Y, WC-Z, wc-heading(nocorr), Heading-flg,
3line
ADCP1pitch, ADCP1roll, ADCP1heading, ADCP1pitch-std, ADCP1roll-std, ADCP1heading-std, CT1Cond, CT1Temp, CT1pres, Heading(nocorr), datano,
4line
Depth, ADCP_x velocity, ADCP_y velocity, ADCP_z velocity, Error, NS velocity, EW velocity, UD velocity, NS velocity(GPS drift correct), EW velocity(GPS drift correct), Flg (sound-spd), mapped bin,
5line(1st hour)
2001, 11, 30, 6, 4, 44.6960, -63.62500, 5.25, 1026.5, 135.056, 5.25, 29.606, 2.0, 0.5, 73.112, 01, 0, 49.210,
6line
-0.078781, 0.078079, 106.060, -122 ,-168, 560, 125.664, 1,
7line
-0.392125, -0.392125, 103.244, 0.1, 0.0, 2, 3.125816, 7.04447, 5.521, 122.848, 8,
8line
24, 2.32, 2.32, -0.59, 0.00, -2.79, 1.72, -0.59, -2.64, 2.20, 1, 1111,
9line
34, -2.32, 0.00, -1.44, 0.00, 2.25, 0.52, -1.45, 2.39, 1.00, 1, 2222,
10line
44, -2.90, 1.16, -1.33, 0.10, 2.55, 1.78, -1.36, 2.69, 2.26, 1, 3333,
11line
54, 0.00, 1.16, -1.23, -0.20, -0.27, 1.12, -1.24, -0.13, 1.60, 1, 4444,
12line
64, -1.16, 2.32, -0.81, -0.20, 0.59, 2.51, -0.83, 0.74, 2.99, 1, 5555,
13line
74, 1.16, 0.58, -0.91, -0.10, -1.27, 0.29, -0.91, -1.12, 0.77, 1, 6666,
14line
104, -1.74, 1.74, -1.86, -0.20, 1.28, 2.07, -1.88, 1.43, 2.55, 1, 9999,
15line
114, -3.48, 3.48, -1.65, 0.20, 2.58, 4.17, -1.70, 2.72, 4.64, 1, aaaa,
16line(2nd hour)
2001, 11, 30, 7, 4, 44.6960, -63.62500, 6.00, 1025.1, 136.462, 4.75, 29.606, 2.5, 3.5, 160.284, 01, 0, 49.210,
-0.235641, -0.078781, 108.524, -130 ,-164, 560, 128.128, 1,
-0.392125, -0.392125, 103.948, 0.1, 0.1, 1, 3.120326, 7.01214, 5.521, 123.552, 8,
24, 2.32, 0.58, -0.07, -0.10, -2.39, 0.00, -0.06, -5.68, 1.18, 1, 1111,
34, -2.90, -1.16, -0.28, -0.10, 3.09, -0.43, -0.29, -0.21, 0.75, 1, 2222,
44, -0.58, -1.16, -0.70, 0.69, 0.84, -0.99, -0.70, -2.46, 0.19, 1, 3333,
54, 0.00, -0.58, -0.28, 0.10, 0.14, -0.56, -0.27, -3.16, 0.62, 1, 4444,
64, -0.58, 0.00, 0.14, 0.30, 0.56, 0.14, 0.14, -2.73, 1.32, 1, 5555,
74, 0.00, 1.74, -0.07, -0.10, -0.42, 1.69, -0.08, -3.71, 2.87, 1, 6666,
104, -1.16, 1.16, -0.38, -0.59, 0.84, 1.40, -0.40, -2.45, 2.58, 1, 9999,
114, -1.16, 1.74, -0.70, -0.30, 0.70, 1.96, -0.72, -2.59, 3.14, 1, aaaa,
  From the first to forth lines are the headers. The first line of an hourly data consists of time, J-CAD number, latitude, longitude, air temperature, wind direction, wind speed, platform heading (true azimuth), platform tilt, GPS drift speed, GPS drift direction, flag for estimated pressure, flag for position, and platform heading without any correction.
  The second line of an hourly data consists of WC (Watson Compass) pitch, WC roll, WC heading (true azimuth), three components of magneto metric field measured by the Watson Compass, WC heading without any correction, and flag for the compass data. Note that this line is removed from the ADCP data of the J-CAD 1, 2, 3, and 5, because the Watson Compass was attached to only the J-CAD 4.
  The third line consists of ADCP pitch, ADCP roll, ADCP heading (true azimuth), standard deviations of ADCP pitch, roll, and heading, CT1 conductivity, CT1 temperature, CT1 pressure, ADCP heading without any correction, and total data number. Salinity and water depth are necessary for the correction of sound speed. If any of salinity, water depth, and water temperature data is lacking, sound speed cannot be obtained, consequently, the correction will not be performed.
  The rest of lines consist of depth, (x, y, z, error) velocity on the ADCP coordinates, (North-South, East-West, Up-Down) velocity on the earth coordinates without ice-drift correction, (North-South, East-West) velocity on the earth coordinates after GPS drift correction, flag for correction on sound speed, and mapped bin of the ADCP data. The drift velocity measured by GPS is used for the drift correction as ice-drift velocity.
  Note that, when the ADCP percent good is below 40, or the ADCP intensity is below 32.0, the current speed and strength are replaced with dummy data.

Year, Month, Day, Hour; Observational Time
J-CAD Number; Deployment number of J-CAD
Latitude, Longitude; Observational position measured by GPS or Argos position data services. Refer to the flag for buoy position.
Air Temperature; Air temperature measured by the thermistor inside the sensor mast.
Air Pressure; Air pressure measured by the barometer installed inside the J-CAD platform.
Wind Direction, Speed; Wind direction and speed measured by Young anemometer. The wind direction is defined as true azimuth.
Platform heading, tilt; Platform heading and vertical tilt measured by a three-axis magnetometer on the electric support channel inside the J-CAD hull. Heading is defined as true azimuth.
GPS drift speed, direction; Buoy drift speed and direction measured by GPS
Flag (est-pre); Flag for water depth estimation. The flag of estimated pressure is the same as mentioned previously.
Flag (position); “0” indicates the data as measured by GPS and “2” as calculated from ARGOS position data service of class 2 or 3.
Heading (nocorr); Platform heading without any correction. The heading data indicates a buoy direction to magnetic north.
WC -pitch, roll, heading; Pitch, roll, and heading data by Watson compass
WC -X, Y, Z; Three components of magneto metric field measured by Watson Compass
WC -heading (nocorr); Watson compass heading without any correction. Heading data indicates a direction to magnetic north.
Heading-flg; Flag for compass heading data. It shows "1" as the Watson compass and "0" as the ADCP compass. Heading data by the Watson compass are given priority over the ADCP heading data. The compass data is required to calculate the earth coordinate velocities from the data based on the ADCP coordinates.
ADCP pitch, roll, heading; Pitch, roll, and heading data measured by the Watson compass
ADCP pitch, roll, and heading -std;      Standard deviation of the above data
CT1 Cond, Temp, pres; Conductivity, temperature, and pressure data measured by the CT sensor at the minimum depth.
Heading (nocorr); ADCP heading without any correction. Heading data indicates a direction to magnetic north.
Data number; Number of hourly CT/CTD data
Depth; Depth of the calculated velocity
ADCP _x, y, z; Velocity data converted to forward, sideway, and up component relative to the ADCP.
Error; Velocity data calculated from the difference between the two estimates of vertical velocity. Error velocity allows you to evaluate whether the horizontal homogeneity is appropriate.
NS velocity, EW velocity, UD velocity;
Velocity data converted into north, east, and up components based on the earth.
NS velocity (GPS drift correct), EW velocity (GPS drift correct);
Horizontal absolute velocity, which is subtracted the ice drift from the above earth coordinate data.
Flg (sound-spd); “1” indicates sound speed correction has been conducted and "-1" indicates having not been conducted.
Mapped bin; Cell data for the velocity calculation of each depth



5.5.1 Conversion to the earth-referenced absolute velocity
  The ADCP mounted on the J-CAD measures the currents relative to the J-CAD. The J-CAD is self-oriented and moves relative to the earth. The J-CAD sends velocity, intensity, and percent good of each beam and each depth cell as ADCP velocity data. Therefore, it is usually necessary to calculate the current velocity from beam velocity data, and to correct the data for the ADCP attitude and motion. In this subsection, we describe the methods for the conversion of the earth-referenced absolute velocity from the ADCP data.
  The calculation and correction of the absolute current velocity is; sound speed correction, depth cell mapping, calculation of the current velocity relative to the ADCP, correction to the earth coordinate, and buoy drift correction. Sound speed is initially set to constant values. We adopted 1450 m/s as a constant value and recorded on the ADCP. To calculate the actual sound velocity from the constant value above, CT data and an equation are used as follows;
          C = 1449.2 + 4.6*T - 0.055*T2 + 0.00029*T3 + (1.34 - 0.01*T)*(S - 35) + 0.016*D,
  T, S, and D are the temperature, salinity and depth respectively, measured by the nearest CT sensor from the ADCP. Then, the velocity obtained from the ADCP can be corrected for the sound speed by using the following equation:
          Vcorrect = Vuncorrect * (C/CADCP).
  Depth cell mapping is conducted to ensure horizontal homogeneity for the calculation of the current velocity from the beam velocity at the same depth. For example, if pitch or roll angle of the ADCP is large, the depth cells move up and down relative to one another. Depth cell mapping matches these cells together at the same depth.
  Each beam measures only a single velocity component that is parallel to the beam. To obtain three components of velocities, we use pairs of beam velocity. If the ADCP is facing downward, these three components of velocities can be calculated as follows;
          VXA = (+V1-V2)/(2.0*sin (20ﾟ)), VYA = (-V3+V4)/(2.0*sin (20ﾟ)),
          VZA = (V1+V2+V3+V4)/(4.0*sin (20ﾟ))
  VXA, VYA, VZA are the calculated current velocities of the X, Y and Z components based on the ADCP coordinates, V1, V2, V3, and V4 are the measured beam velocities, respectively. Also, error velocity is estimated from the difference between the two estimates of vertical velocity by the following equation:
          VEA = (V1+V2-V3-V4)/4.0.
  The ADCP rotates velocity components into true north and east, so the conversion to the earth coordinates is necessary. This conversion requires heading, pitch, and roll of the ADCP. The equations are as follows;
          VXE = VXA*(CH*CR+SH*SR*SP)+VYA*SH*CP+VZA*(CH*SR+SH*CR*SP),
          VYE = -VXA*(SH*CR-CH*SR*SP)+VYA*CH*CP+VZA*(SH*SR+CH*CR*SP),
          VZE = -VXA*SR*CP+VYA*SP+VZA*CR*CP,
          CH = cos (heading), SH = sin (heading),
          CR = cos (roll), SR = sin (roll), CP = cos (pitch), SP = sin (pitch),
  VXE, VYE, and VZE are zonal, meridional, and vertical components of the current velocities respectively, based on the earth coordinate before the buoy drift correction.
  Given the absolute earth-coordinate velocity, the buoy drift correction should be conducted. This correction requires the J-CAD drift velocity obtained from GPS. The following equations are used to subtract the drift effect:
          VX = VXE + VEDrift, VY = VYE + VNDrift,
  VEDrift and VNDrift are the zonal and meridional velocities of the J-CAD drift.
  The above nine velocity components derived from the beam velocity data, i.e., VXA, VYA, VZA, VEA, VXE, VYE, VZE, VX, and VY, correspond to ADCP _x, y, z, Error, EW velocity, NS velocity, UD velocity, EW velocity (GPS drift correct), and NS velocity (GPS drift correct) written in the ADCP data file, respectively.
  For more detailed information on the ADCP current measurement, please refer to “Acoustic Doppler Current Profilers Principles of Operation: A Practical Primer” by RD Instruments.

5.5.2 Heading error estimation of J-CAD ADCP data
  As described in the previous sections, RD Instruments WH-ADCP 300 kHz has been used for the direct measurements of upper-ocean currents. After the deactivation of Selective Availability to GPS on May 2nd 2000, GPS position improved its accuracy (less than 10 meters). We can obtain the buoy drift velocities and the consequently absolute velocities more precisely than ever. The estimated accuracy is 1.0 cm/s in velocity, which is calculated from the sum of the statistical uncertainty of the measurement, the remaining of statistic error after averaging, error from the calculation of the GPS drifting velocities, and other elements of the error.
  According to RD Instruments, the compass mounted with the WH-ADCP guarantees the measurements in a place with more than 100 mGauss in horizontal intensity. However, since horizontal magnetic intensity (less than 30 mGauss) exsits around the North Pole, we need to estimate heading errors for the magnetic compass mounted with the WH-ADCP. In this subsection, we describe the heading error estimation for the ADCP data.
  The main idea of the heading error estimation is based on the following concept. The WH-ADCP mounted on J-CAD measures the ice drift relative to the motion of a buoy. The velocity measured by the WH-ADCP is the sum of both ocean current and buoy drifting velocities, and this velocity has the opposite direction of buoy drift. Assuming that there was no ocean flow and that the data had no error other than heading errors, the current direction and speed measured by the ADCP would show the vector which is completely opposite to the direction and speed of the buoy drift. If there were heading errors, the vector would have an angle gap. This gap should be extracted from the data. However, there are in fact ocean currents, and the data are likely to have various error factors besides heading errors. The flow near sea ices has a certain speed, so the current direction and speed measured at the maximum depth possible should be used for the estimates. Moreover, we use the data with the maximum velocity because the influence of the ocean currents and the other effects become relatively small. Sea ice drift data are calculated from GPS, and the data for 5 hours are averaged to eliminate high frequency noise. ADCP velocity data are also calculated from each bin velocity data using compass data and then the data for 5 hours are averaged.
  We compared the ADCP velocity data with the sea ice drift data observed by J-CAD 1 from April 26th to November 1st 2000. When the sea ice drift speed was more than 15.0 cm/s, the angle difference between the ADCP velocity and the sea ice drift was -1.13 ± 9.99 degrees. After changing the threshold value of the sea ice drift speed to 12.0, 15.0, 18.0, and 20.0 cm/s, each gap was -0.23 ± 12.85, -1.13 ± 9.99, -0.94 ± 9.41, -0.71 ± 8.98 degrees, respectively. When the value was over 15.0 cm/s, the angle gap and deviation remained nearly unchanged. As a result, the angle error of the ADCP data of the J-CAD 1 is estimated as within ~ 10.0 degrees deviation. As mentioned above, the data are likely to have other elements of errors, therefore, the estimates show the maximum value of angle error range considering other error factors. The actual heading error caused by the compass should be less than the angle differences shown above.
 Also, we estimated the angle error for the ADCP data of the J-CAD 4 by the same method. The Watson compass is co-located with WH-ADCP and measures independent data of heading, pitch, roll, and 3-axis magneto-metric field. We estimated the heading error using both the ADCP compass and the Watson compass. Using the ADCP compass data of November 2002, the heading error was estimated as -13.0 ± 15.5 degrees. In particular, when the latitude was over 88.0N, the gap and deviation of the heading error both became higher, more than 30 degrees and 20 degrees, respectively. On the other hand, the heading error using the Watson compass data was 0.8 ± 10.4 degrees. In comparing the Watson Compass and the ADCP Compass, the results show that the Watson compass generates better calculations for the absolute velocity in high latitude areas. As to the ADCP data from the J-CAD 2 and 3, we could not estimate the error range due to the short duration of the data-sampling periods.

  From the results, we concluded that the WH-ADCP compass functioned successfully for J-CAD 1 but not for J-CAD 4. On the other hand, the Watson compass performed well for J-CAD 4 to calculate the absolute velocity direction. There is a conflict of performance between them. We use the WH-ADCP under the extremely severe conditions for the compass. Therefore, each performance of the compass may be different and should be checked by different sensors or methods.