Abstract -------- University of Hawaii ADCP data acquisition and processing software "UHDAS" has been installed on two Antarctic research vessels, two NOAA ships, 8 UNOLS ships, and one other vessel. A system status report status and subsampled data are emailed from these ships daily, allowing the health of the systems to be monitored. In general, the ADCPs using UHDAS are doing well. Two installations are anomalous in that they require a scale factor to be applied to the measured velocities before use. This is unusual and warrents concern. The two instruments are the OS75s on the Thompson and the Wecoma. Recent refurbishing of the HDSS sonars on the Revelle has improved their characteristics, but there is evidence of small biases at low speeds. These are not understood. High quality ADCP data relies on a good installation of the transducer, and consistent feeds of gps position and high quality headings. Many different heading devices are used, including mechanical gyro, optical gyro, gps arrays (eg. Ashtech, gps compass), and more expensive systems that combine inertial sensors and gps arrays with Kalman filters (eg. POSMV, Seapath). Heading errors result in cross-track errors in the ADCP data which are proportional to the ship's speed. Heading errors should be kept to under a few tenths of a degree to keep the errors in the ADCP data under a few cm/s (Most open ocean velocities are under 20cm/s.) We use a gyro for smoothness and reliability and correct those headings with another instrument. Ashtech, Seapath, and POSMV are capable of high accuracy and are the preferred final heading reference. Some other devices (Phins, Marinus, Mahrs) may be adequate, with errors of a few tenths of a degree. Some gyros are quite bad, especially in high latitudes, having errors of several degrees. Newer gyros with a gps feed seem to be much better than older gyros. Preliminary tests of two "gps compass" devices indicate that they are not a replacement for the high-accuracy devices. Limited data from a Furuno "satellite compass" suggests that it is comparable to the better mechanical gyros. The CSI "gps compass" is far worse than any other device tested, with large excursions unflagged by its own QC. /////////////// // Section // /////////////// Outline: --------- - Ships with UHDAS installed - ADCP Lessons Learned - ADCP Status . Introduction to ADCP Velocity Determination . ADCP Errors and solutions . Notes on Specific Installations . Heading Devices: Overview and Suitability for ADCPs - Summary /////////////// // Section // /////////////// ADCPs with UHDAS ---------------- .UHDAS installations `----------------`-------------------`-------------`--------------- Ship Instruments Institution Heading Devices ------------------------------------------------------------------- N.B.Palmer NB150, OS38 NSF gyro, Seapath L.M.Gould NB150, OS38 NSF gyro, Ashtech Atlantis OS75 WHOI gyro Knorr NB150, OS75 WHOI gyro, POSMV Oceanus NB150, OS75 WHOI gyro, Ashtech Revelle NB150, HDSS(140,50) Scripps gyro, PHINS, Marinus, Ashtech Melville NB150, OS75 Scripps gyro, Ashtech Thompson OS75 UW gyro, POSMV Wecoma WH300, OS75 OSU gyro, Ashtech Kilo Moana WH300, OS38 UH gyro, Ashtech, POSMV KOK NB150 UH gyro, Ashtech Hi`iakalai OS75 NOAA gyro, POSMV Ka`imimoana OS75 NOAA gyro, POSMV, CSI ---------------------------------------------------------------------- /////////////// // Section // /////////////// ADCP: Lessons Learned ---------------------- . Installation (ADCP) - if possible, accessible from the dry side - don't block transducers (edges) - window parallel to transducers, not tilted - bubbles can kill the data + . Acquisition and Processing - BB mode has higher resolution but shorter range than NB mode - bubbles: for some ships, single-ping editing is necessary - accurate and reliable heading necessary - issues with clocks: + -- do not change the computer clock time during data acquisition + -- perhaps turn off "Windows Time Service"? + . ADCP Errors and solutions - gps: (pretty good now) - ringing (increase blank or change installation) - bubbles (change installation, may be able to edit velocities) - instrument errors - heading (correct gyro to other device) - scale factor (eg. temperature correction expected for wh300, nb150, not OS) /////////////// // Section // /////////////// Determining ADCP velocities ---------------------------- - 4 beams of sound - measure Doppler frequency shift - obtain horizontal velocities from opposing beams - use transducer angle (from bow) to get velocity in SHIP coordinates - use heading of bow to orient with Earth - use positions to take out ship's speed - show two kinds of errors: * angle: 1 degree = 10cm/s underway * scale: 2% scale factor = 10cm/s underway The ADCP is mounted to the hull and measures velocities from the moving ship. In an ocean with no currents, the measured velocity should be the same as the ship's velocity but in the opposite direction [[FIG_VelCoords]] image::figs/shipsteaming.png[figure depicting measured velocity] Two common sources of error in ADCP ocean velocity calculations are the total angle used to transform beam coordinates to earth coodinates, and the extent to which the measured velocities are consistently too small or too large. These errors are nearly independent, with the angle error primarily affecting cross-track velocity component and the scale factor affecting along-track component. The following figure illustrates the way in which angle and scale factor errors result in cross-track and along-track errors, respectively. [[FIG_MeasVel]] image::figs/head_amp_err.png[effect of errors in heading and magnitude] Accurate ADCP velocities require - angle errors below .2deg (2cm/s error underway) + - scale factor under 1/2 percent (2-3cm/s error underway) /////////////// // Section // /////////////// Specific Issues/Instruments ---------------------------- . WH300 (Kilo Moana and Wecoma) - depth range disappointing (but in spec) - synchro gyro input (converter, not input to data) . Scale Factor Errors: - Thompson scale factor (BB=NB=BT; temperature dependent, up to 3-4 percent) - Wecoma (BT=NB, BB up to 3-4 percent) . NB150 vs/ HDSS evaluation - HDSS 50kHz: new transducers in January 2006 - HDSS new electronics of some kind - biases in HDSS 50Khz due to strong scattering layers - biases at low speeds in both instruments (see link:http://currents.soest.hawaii.edu/reports/hdss_nb150_2006/index.html[report]) /////////////// // Section // /////////////// Figure: Thompson Scale Factor ------------------------------ [[FIG_Thompson]] image::figs/P16N_orig_ampcalT.png[Thompson scale factor] Watertrack calibrations using CODAS ADCP processing software showed a scale factor was necessary for measured velocities on the Thompson. A scale factor should not be necessary for an Ocean Surveyor ADCP, and it is particularly odd that this seems to vary with temperature. /////////////// // Section // /////////////// Figure: HDSS 50kHz and scattering layer --------------------------------------- [[FIG_ScatHDSS]] image::figs/BOSL.png[HDSS 50 scattering bias from scattering layer] A weak scattering layer at 700m exists below a region of lower scattering (400-650m). The 50kHz HDSS beam pattern responds badly to this combination, resulting in an along-track bias of 50cm/s in this example. /////////////// // Section // /////////////// Heading Devices: Overview and Suitability for ADCPs ---------------------------------------------------- - POSMV (description, installations; failures) - Ashtech (description, failure modes) - Seapath (description, failure modes) - gyro; gyro with gps - gps compass /////////////// // Section // /////////////// POSMV installations -------------------- link to link:http://currents.soest.hawaii.edu/reports/posmv_evaluations/index.html[online report] .Summary of POSMV `--------------`-------------------`------------`-------------`------------ ship accuracy overall percent comment cutoff percent not perfect (from data) GAMS accuracy ---------------------------------------------------------------------------- Kilo Moana(1) .013 >95 10 very good Kilo Moana(2) .013 >99 1 excellent Knorr .018 75 17 good Ka`imimoana .015 60 15 good Thompson .055 80 1 very good Hi`ialakai(1) .018 50 50 bad Hi`ialakai(2) .018 50 35 less flakey ----------------------------------------------------------------------------- /////////////// // Section // /////////////// Ashtech and Gyro Behavior ------------------------- [[FIG_Headings]] image::figs/dh_pos.png[Kilo Moana headings] Ashtech has a reacquisition flag that does a very good job of eliminating most bad heading measurements. The red dots in the top panel are unflagged Ashtech data. The second panel has QC applied and the Ashtech remaining values are much cleaner. The bottom panel shows the difference between gyro and POSMV, and between Ashtech and POSMV. The gyro shows a heading-dependent error of up to 2 degrees. The POSMV and Ashtech difference is nearly constant. /////////////// // Section // /////////////// CSI Behavior ------------ image::figs/csi_errs.png[CSI heading accuracy] The CSI has large excursions that still exist even after QC has been applied. This is not a device suitable for use with an ADCP. /////////////// // Section // /////////////// Heading Device Quality: What works?? ------------------------------------- .Comparison to POSMV and/or Ashtech `----------`-----------------`-----------`--------`-------------- ship instrument max 30min stddev location comparison deviation ----------------------------------------------------------------- Ka`mimmoana Ashtech/Posmv 0.11 0.04 Kilo Moana Ashtech/Posmv 0.11 0.04 Kilo Moana Ashtech/Posmv 0.11 0.04 N.B.Palmer Seapath/Ashtech 0.18 0.06 Ka`imimoana Marinus/Posmv 0.40 0.10 10S-10N Kilo Moana AnschutzPosmv 0.69 0.22 Hawaii (22N) R.Revelle Marinus/Ashtech 0.92 0.18 20-35N R.Revelle Phins/Ashtech 0.91 0.18 20-35N Tioga Furuno/Ashtech 1.13 0.26 40N Kilo Moana Anschutz/Posmv 1.32 0.69 Hawaii (21N) R.Revelle Sperry/Ashtech 1.49 0.29 20-35N Atlantis gyro/Ashtech 2.41 0.73 north atlantic N.B.Palmer gyro/Ashtech 6.76 2.03 southern ocean Atlantis Octans/Ashtech 2.69 0.67 north atlantic Ka`imimoana CSI/Posmv 269.09 16.61 ----------------------------------------------------------------- [[FIG_Accuracy]] image::figs/dhsummary.png[summary:heading accuracy] Summary of Heading Devices 1. When they work, Ashtech POSMV, and Seapath are comparable. 2. The Ka`imimoana Marinus was close behind, but the POSMV does improve the data quality. 3. Newer electro-mecahanical gyros appear to be reliable, but with errors of almost a degree, require correction. 4. Older mechanical gyros can have obvious heading-dependent errors and definitely require correction by a #1 device. 5. The Furuno gps compass was compared to an Ashtech. No long time series was available, but it seems to be reliable and should be grouped in quality with the the mecahical gyros, coming in between the older mecahincal gyros and the newer gyros (with a gps interface). It is not a suitable device to improve heading over that of a gyro. 6. The Octans had a gps interface that was not connected. Presumably its performance was poor because of this. 7. The CSI is not a substitute for any of these instruments. It short wild swings in heading that can be nearly any size, after applying its own QC flag. This makes further quality control impossible. Thus the instrument is rated as poor.