-
Notifications
You must be signed in to change notification settings - Fork 1
Calibration (Patch Test)
This section is under development. If you have expertise in calibration (patch testing), please reach out to the admins at omcadmin@ccom.unh.edu to become a contributor today!
[IN DEVELOPMENT]
Geometric calibration (patch testing) is a process of identifying and correcting residual (small) biases in the swath data. This is performed only after all sensor offsets have been installed, surveyed, and configured to manufacturers' standards, including antenna calibration for the navigation system.
Several steps must be completed before patch testing:
-
Calibration of the navigation system's GNSS antenna baseline
- Applanix POS MV GAMS procedure (at sea; 10-60 minutes)
- Seapath GNSS Calibration Wizard (dockside; 2 hrs; see manual)
Applanix has put together an installation and survey story that illustrates several important points to consider for any navigation system.
It is important to consider what it is that we are really aligning through this process. Assuming all linear offsets are correct, the process help to identify position latency and the net effects of angular biases that exist between the navigation/attitude system and the echosounder (as observed in the swath data). It does not account for attitude latency. There are different approaches for how to attribute these results, which may depend on the sensor manufacturer's expectations for how these are applied.
Even with highly accurate offset surveys, patch tests typically reveal small (<0.5 deg) angular biases. However, the process may reveal larger issues (e.g., latency or linear offset errors) which must be addressed before a meaningful calibration can be completed. Patch test results larger than 1 degree strongly indicate the need for a review of system geometry and configurations, and possibly the need for a more accurate sensor offset survey.
A comprehensive foundation for multibeam calibration is presented in Godin (1998).
There are several concise resources for modern patch test methods:
- QPS: How to calibrate a multibeam echosounder (patch test)
- [References to be added]
The approach presented here is typically used by the MAC for deepwater systems with large TX and RX arrays.
This implies a few assumptions about the mapping system reference frame and sensor configuration:
- The mapping system reference frame is aligned with the vessel axes, with a user-selected origin (e.g., granite block)
- All sensor survey results meet manufacturer specifications for uncertainty
- All sensor offsets are reported in the mapping system reference frame
- All sensor offsets are configured in their respective software using a consistent reference frame:
- TX and RX array X, Y, Z, roll, pitch, heading
- Motion sensor X, Y, Z, roll, pitch, heading
- GNSS antenna X, Y, Z
- Translation of sensor output is acceptable with careful review
- For instance, a Seapath Monitoring Point must match the navigation input location expected in SIS
Under these assumptions, the patch test results are interpreted as follows:
-
Results represent the residual misalignment between the echosounder arrays and attitude source
-
Results are attributed to the motion sensor because:
- TX and RX array installation angle biases cannot be resolved individually from the patch test data
- Arrays typically provide longer survey baselines (and smaller angular uncertainties) than motion sensor housings
-
Results are applied in the multibeam software rather than the navigation software because:
- Results are determined from multibeam data
- Results do not necessarily apply to other systems receiving attitude data
- This distinguishes patch test results separately from array angles in the multibeam configuration
- This maintains consistent motion sensor configuration based on the survey report
Note there are cases in which the motion sensor offsets are known more confidently than the array offsets (e.g., retractable keel systems where raising/lowering is not repeatable to <0.01 deg). In these cases, patch test results might be attributed to array misalignment in the mapping frame, summed with the nominal offset survey results, and applied in the multibeam software. This may be a case-by-case consideration of the confidence in each sensor's offsets.
Add topics on pre-patch test review, site selection, and settings.
Colleagues from the MAC and NOAA have started to organize proven multibeam assessment sites into the MBES Test Site Database. Each site includes links to waypoints and any available settings or reports. This is a work in progress with some ongoing discussion about how to organize sites by depth and frequency (i.e., in a model-agnostic way). New sites will be added as they are developed or re-organized from previous SAT/QAT efforts.
The database is intended to simplify the selection of assessment sites near a vessel's operating areas and transit routes, saving significant planning effort and leading to repeat use of these sites to make more meaningful comparisons across ships and installations.
Toward this end, operators are encouraged to use the existing sites, provide feedback, and add information for new sites serving any multibeam assessment purpose:
- Calibration (patch test)
- Accuracy testing
- Swath coverage assessment
- Backscatter normalization
- (Potential: EK60/80 calibration sites)
More details on site submission are in development (e.g., a Google Form for submitting waypoints and any related reports or example from vessels using a site). In the meantime, please reach out to the Multibeam Advisory Committee or the OMC Admins with any site information you'd like to share.
Acquiring patch test lines
The 'active' navigation source provides the following components for real-time application:
- Position (latitude and longitude)1
- Attitude (roll, pitch, heading)2
- Attitude velocity (rates of rotation)2
- Heave (vertical displacement from a reference level)1
- Timing (PPS signal and ZDA messages)
1Position and heave are valid at a reference point that must be consistent with MBES configuration
2Attitude and attitude velocity are assumed to apply uniformly for the whole vessel and mapping system
All navigation data are used in real time by the multibeam echosounder, including some data streams that cannot be adjusted or reapplied in post-processing (even if logged). For example, attitude velocity is used by some systems during the RX cycle to correct for Doppler shifts perceived by individual transducer elements due to rotation of the vessel and RX array. This data stream has a fundamental impact on bottom detection and generally cannot be reprocessed after acquisition.
Accordingly, it is important to conduct independent patch tests for all navigation systems (especially if new) for the following reasons:
- Only the active attitude source is applied in real time
- Post-processing patch test data for non-active navigation source logged in the raw files may solve for bulk offsets for that sensor, but does not ensure that the multibeam functions properly in real time
- Opportunities to patch test a 'backup' navigation system may be limited or non-existent when it (eventually) needs to be put into service
- It is more efficient and far less risky to make sufficient time for testing all systems at proven patch test sites
- Logging both navigation sources with patch test values applied ensures these offsets are available in the raw data
- This can significantly simplify downstream troubleshooting of any complications with the primary navigation source and re-application of the secondary source, if needed
- This is especially important if multibeam data are not being processed quickly after acquisition; many mapping datasets have been 'saved' by the secondary, fully patch-tested navigation source when a) an issue arose with the primary source that was not detected during acquisition, and/or b) there was no opportunity to patch test the 'backup' system after detection of an issue
After verification of all offsets from the vessel survey (see Dimensional Control), vessels with multiple navigation sources should prioritize their patch test planning as follows:
- Calibration with the primary navigation source applied in real time
- Calibration with the secondary navigation source applied in real time
- Verifications (repeat calibrations with initial results applied in multibeam configuration)
- This can be done for each system in order of priority, as time allows
In both cases, only the 'applied' or 'active' navigation source changes. All navigation sources are still logged to the raw files, if possible (this may depend on vessel network constraints).
Switching the active navigation source is highly vessel-specific, depending on network setup and software versions. For example, see Selecting active navigation sources in SIS 4 for an overview of an EM122 installation with POS MV and Seapath navigation sources.
It is imperative that the PPS timing source and format be updated to match the active navigation system. See the table of PPS conventions for common installations.
Regardless of differences in the interfaces across software versions, the operator must ensure that the position, attitude, attitude velocity, and timing sources are consistent for the desired active navigation system. Patch testing (or surveying) with inconsistencies across these sources can cause data quality issues that are not addressable in post-processing. It is highly recommended to practice switching the active navigation source routinely to ensure smooth transition when necessary during survey, with review during post-processing to identify any obvious issues or mismatches between the data sources.
Notes on processing in Qimera, SIS, Caris, etc.
Applying patch test values in SIS
How to document the process and results
Supported under NSF grants 1933720 and 1933776