Uncertainty Uncertainty Workshop: Workshop: Sounding Sounding attributes attributes Rob Hare Rob Hare Manager, Hydrographic Surveys Manager, Hydrographic Surveys Canadian Hydrographic Service, Pacific Canadian Hydrographic Service, Pacific Region Region CHC2004 May 24, 2004 CHC2004 May 24, 2004
43
Embed
Uncertainty Workshop: Sounding attributes Rob Hare Manager, Hydrographic Surveys Canadian Hydrographic Service, Pacific Region CHC2004 May 24, 2004.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Rob HareRob HareManager, Hydrographic SurveysManager, Hydrographic Surveys
Canadian Hydrographic Service, Pacific RegionCanadian Hydrographic Service, Pacific Region
CHC2004 May 24, 2004CHC2004 May 24, 2004
AbstractAbstract
A discussion of attributes on A discussion of attributes on soundingssoundings, sources of error and , sources of error and computation of sounding error. computation of sounding error. Plus a whole lot morePlus a whole lot more
An error prediction tool along with An error prediction tool along with the characteristics and capabilities the characteristics and capabilities of (a few) sounding systems will be of (a few) sounding systems will be reviewedreviewed..
Data Quality ElementsData Quality Elements CompletenessCompleteness Logical consistencyLogical consistency Positional accuracyPositional accuracy
Has Has horizontal and verticalhorizontal and vertical components components Sub-elements: Sub-elements: absolute accuracyabsolute accuracy, attribute , attribute
completeness, shape fidelity, time accuracy, completeness, shape fidelity, time accuracy, topologic consistencytopologic consistency
Estimation of Quality Estimation of Quality metricsmetrics
Direct and indirect methodsDirect and indirect methods Require absolute coordinate Require absolute coordinate
reference frame (e.g. WGS-84)reference frame (e.g. WGS-84) Most direct methods impracticalMost direct methods impractical
source inter-comparison is an exceptionsource inter-comparison is an exception Indirect methods require validationIndirect methods require validation
e.g. deductive error estimation (forward e.g. deductive error estimation (forward error prediction)error prediction)
Statistical estimatorsStatistical estimators
Measures of central tendencyMeasures of central tendency Sample meanSample mean Median, ModeMedian, Mode
Measures of dispersion (1-D and 2-D)Measures of dispersion (1-D and 2-D) Standard deviation (or variance), Standard deviation (or variance), RMSRMS CEP, MSEP,CEP, MSEP, drmsdrms
Total Propagated Error (TPE)Total Propagated Error (TPE) CAUTIONCAUTION - many hydrographic - many hydrographic
measurements are correlated (e.g. H&V measurements are correlated (e.g. H&V components of soundings)components of soundings)
PreanalysisPreanalysis Will my system meet specifications?Will my system meet specifications? Do I purchase a new … C/B analysis?Do I purchase a new … C/B analysis?
Real-time QA Real-time QA Am I collecting enough data to meet specifications?Am I collecting enough data to meet specifications? Do I modify my sampling/processing strategy, discard outer Do I modify my sampling/processing strategy, discard outer
beams, increase overlap, take more sound speed profiles, beams, increase overlap, take more sound speed profiles, etc.?etc.?
Post-mission assessmentPost-mission assessment Did I meet specifications? Classification - what Order did I Did I meet specifications? Classification - what Order did I
achieve?achieve? Provide metadata for informed decision making/risk Provide metadata for informed decision making/risk
assessmentassessment Data attribution for integration/validation/comparison of Data attribution for integration/validation/comparison of
different data setsdifferent data sets Assessment of historic sourcesAssessment of historic sources
Initialization of CUBE estimator?Initialization of CUBE estimator? Create Source Classification or Reliability Diagrams or ZOCsCreate Source Classification or Reliability Diagrams or ZOCs
analogue soundersanalogue sounders digital soundersdigital sounders
Sweep (multi-transducer)Sweep (multi-transducer) Lidar Lidar Swath (multibeam echosounder - MBES)Swath (multibeam echosounder - MBES) Other – e.g. TIBS, wire or bar sweep, Other – e.g. TIBS, wire or bar sweep,
diverdiver
Sources of error - water Sources of error - water levelslevels
gauge measurement precisiongauge measurement precision method of filtering sea surface wavesmethod of filtering sea surface waves timing synchronisation of gauge and timing synchronisation of gauge and
measurement measurement vertical datum precision vertical datum precision spatial extrapolation to the location of the spatial extrapolation to the location of the
vertical measurement, or in the case of vertical measurement, or in the case of predictions,predictions,
quality of constituent set (length/quality of quality of constituent set (length/quality of observations)observations)
correction for environmental effectscorrection for environmental effects
Other sources of vertical Other sources of vertical errorerror
Draft and squat or settlementDraft and squat or settlement Antenna to transducer offset Antenna to transducer offset
(GPS)(GPS) Datum separation model (GPS)Datum separation model (GPS) Sounding measurementSounding measurement HeaveHeave Sound speedSound speed
Traditional sounding Traditional sounding reductionreduction
Dynamic draft
ChartedDepth, D
Chart datumMeasuredDepth, d
Tide, WL
θr
D = d + draft – WLd = r cos (θ+R) cos P
r = rangeΘ = beam angleR = Roll angleP = Pitch angle
includes squat, settlement, change of includes squat, settlement, change of trimtrim
Heave (measured and induced)Heave (measured and induced) Tides or water levelsTides or water levels Roll, pitch, headingRoll, pitch, heading Calibration (patch test) offsetsCalibration (patch test) offsets
Confidence levelsConfidence levels For a normal distribution, the probabilities of For a normal distribution, the probabilities of
univariate random errors of a single univariate random errors of a single measurement falling within a certain level of measurement falling within a certain level of error (number of standard deviations, error (number of standard deviations, ) are ) are
given in the following table.given in the following table.
EM1002 depth error estimates (95%)(view looking aft, EA mode)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
-250 -200 -150 -100 -50 0 50 100 150 200 250
Cross-track distance (m)
Dep
th e
rro
r (m
)
Sounder Heave Roll Pitch
Refraction Total depth measurement Dynamic draft Water level
Total reduced depth (EA mode) IHO Special Order IHO Order 1
transducer shape, orientation, roll-modulationtransducer shape, orientation, roll-modulation roll, pitch and roll, pitch and headingheading error error latency between systems latency between systems
GPS, sounder, VRUGPS, sounder, VRU
Sample position error mappingSample position error mappingequationsequations
1sp P r2 1 2 2 ( (cos cos ) )
2sp r P2 2 1 2 2 ( (cos cos ) )
Sounding position error mapSounding position error mapp
Pv
R R v
r
v v
m
pitch stabilizationR
non-orthogonally-steered beams
PPrm
vm
sm
m
s
p s
ps ps ps ps4321
vm
vm
m
Mapping total position errorMapping total position error
p
drms pp p
speed latency P offsets P R
s at
Mapping position errorsMapping position errors
Confidence levelsConfidence levels
EM1002 position error estimates (2MSEP)(view looking aft, EA mode)
0
2
4
6
8
10
12
-250 -200 -150 -100 -50 0 50 100 150 200 250
Cross-track distance (m)
Po
siti
on
err
or
(m)
Positioning System Sounding Refraction Roll Pitch Heading Sounding position IHO Special Order IHO Order 1
Position error estimatesPosition error estimates
Field Sheet processesField Sheet processes Horizontal datumHorizontal datum Manual processesManual processes
Materials and constructionMaterials and construction Horizontal controlHorizontal control Station plottingStation plotting
Manual/OpticalManual/Optical Sextant (eccentric circle LOP)Sextant (eccentric circle LOP) Subtense (eccentric circle LOP)Subtense (eccentric circle LOP) Range poles (straight-line LOP)Range poles (straight-line LOP) Azimuth (straight-line LOP)Azimuth (straight-line LOP)
Electronic positioning systems (EPS)Electronic positioning systems (EPS) Range-bearing (hybrid - circle and line) Range-bearing (hybrid - circle and line) Two-range or range-range (later multi-range) - concentric Two-range or range-range (later multi-range) - concentric
circle LOPcircle LOP Hyperbolic (2 and later multi-hyperbola) LOPHyperbolic (2 and later multi-hyperbola) LOP Transit satellite (Doppler) - spherical LOPTransit satellite (Doppler) - spherical LOP GPS (spherical LOP)GPS (spherical LOP)
PlottingPlotting 3-arm protractor3-arm protractor latticeslattices collector registrationcollector registration fixes and ‘tweeners’fixes and ‘tweeners’
InkingInking on collectoron collector on field sheeton field sheet registrationregistration
DigitizingDigitizing digitizer resolutiondigitizer resolution registration (rms of fit on HCP) - transformation usedregistration (rms of fit on HCP) - transformation used datum shiftdatum shift projectionprojection
Data inter-comparisons:Data inter-comparisons:validating error prediction validating error prediction
assumptionsassumptions MethodsMethods
compare same data set at different process stagescompare same data set at different process stages difference DTM (soundings)difference DTM (soundings) inter-comparison of point features (shoal examinations, inter-comparison of point features (shoal examinations,
LimitationsLimitations small statistical samplesmall statistical sample is it the same feature?is it the same feature? separation of depth from position errorseparation of depth from position error
22
21 mm
Quality ImplementationsQuality Implementations Data collection - ASCII, Simrad, Hypack, NMEAData collection - ASCII, Simrad, Hypack, NMEA Data processing - HIPS (Quality flag, coverage, Data processing - HIPS (Quality flag, coverage,
standard deviation), CUBE (stochastic surface, standard deviation), CUBE (stochastic surface, etc.)etc.)
Data storage - Caris ASCII, GSF, SDS, …Data storage - Caris ASCII, GSF, SDS, … Paper Charts - Source Classification/Reliability Paper Charts - Source Classification/Reliability
Diagrams, Explanations in Sailing Diagrams, Explanations in Sailing Directions/PilotsDirections/Pilots
SummarySummary Depth and position errors can be estimated Depth and position errors can be estimated
using forward error predictionusing forward error prediction These estimates can be validated by inter-These estimates can be validated by inter-
comparison of data setscomparison of data sets Estimates can be used:Estimates can be used:
To make decisions regarding equipment To make decisions regarding equipment selection and purchaseselection and purchase
To determine if specifications can be/have been To determine if specifications can be/have been metmet
To adapt your sampling strategyTo adapt your sampling strategy As an input to statistical processing algorithmsAs an input to statistical processing algorithms
These are but two of many These are but two of many Quality/Uncertainty MeasuresQuality/Uncertainty Measures