# S-44 and USACE Standards

Πολεοδομικά Έργα

15 Νοε 2013 (πριν από 4 χρόνια και 6 μήνες)

104 εμφανίσεις

S
-
44 and USACE Standards

Depth Stuff Only

USACE

Single Beam

QA CROSS LINE PERFORMANCE TEST:

Hard Bottom

Soft
Bottom

Max. Allowable Mean Bias

<0.1’

<0.2’

Resultant Depth Accuracy (95%):

Depth < 15’

0.5’

0.5’

15’ < Depth < 40’

1.0’

1.0’

Depth > 40’

1.0’

2.0’

Minimum # of Samples

100

100

Note that in the 1 Jan 02 version of EM 1110
-
2
-
1003, it calls for the Resultant
Elevation/Depth Accuracy. There is no way to compute this unless you have a
lock chamber. We can computer the Resultant Repeatability, which is what
most surveyors wind up using.

S
-
44

Single Beam & Multibeam

Formulae in meters:

Special Order

Order 1a

Order

1b

Areas where under keel
clearance is critical.

Areas

shallower than
100m where under keel
clearance is less critical
but features of concern
to surface shipping may
exist.

Areas

shallower than
100 where under keel
clearance is not
considered to be an
issue for the type of
surface shipping
expected to transit the
area.

Maximum
Allowable Total
Vertical
Uncertainty

A = 0.25m

B = 0.0075

A = 0.5m

B = 0.013

A =

0.5m

B = 0.013

TVU <

[ A
2

+ (B x Depth)
2

]
.5

Some terms:

Accuracy: How close is your measurement to the true value?

Repeatability: How good are your chances of going out and
measuring the same value?

Uncertainty: The range about a depth measurement that
should contain the true value at a certain confidence level
(usually 95%).

Confidence Level: The probability that the true value lies
within a specified uncertainty from the measured value.

Measured Depth = 25.0’

Uncertainty at a 95% Confidence
Level = 1.0’

We have a 95% chance that

the
true sounding is between 24.0’
and 26.0’.

The smaller the uncertainty, the
more ‘confident’ you are that your
measured

value reflects the true
value.

Uncertainty:

How can you compute uncertainty:

Uncertainty can be based on either limitations of the measuring
instruments (think TPU) or from statistical fluctuations in the quantity
being measured.

Limitations of the Measuring

Instruments:

By determining the measurement errors
in each sensor and factoring how those
errors propagate throughout the
system, you may compute uncertainty
values for your position and depth
measurements.

TVU

= Total Vertical Uncertainty

THU

= Total Horizontal Uncertainty

TPU

= Total Propagated
Uncertainty

Statistical Fluctuations in the
Quantity

Being Measured.

Determine the differences between
-

values compared to:

A known depth (Lock Chamber
or Bar Check)

A surveyed test area
(Performance Test)

A previously surveyed data set
(Cross Check Statistics)

S
-
44
vs

USACE: Some Numbers

Depth (Ft)

Depth (m)

Special Order (m)

Order 1A (m)

Special Order (ft)

Order 1A (ft)

USACE
Hard
(ft)

USACE
Soft
(ft)

10

3.05

0.25

0.50

0.75

1.53

0.50

0.50

15

4.57

0.25

0.50

0.75

1.53

1.00

1.00

20

6.10

0.25

0.51

0.76

1.54

1.00

1.00

25

7.62

0.25

0.51

0.77

1.55

1.00

1.00

30

9.14

0.25

0.51

0.78

1.57

1.00

1.00

35

10.67

0.26

0.52

0.79

1.58

1.00

1.00

40

12.19

0.26

0.52

0.80

1.60

1.00

1.00

45

13.72

0.27

0.53

0.81

1.62

1.00

2.00

50

15.24

0.27

0.54

0.82

1.64

1.00

2.00

55

16.76

0.28

0.55

0.84

1.66

1.00

2.00

60

18.29

0.28

0.55

0.86

1.69

1.00

2.00

Tests to Check Compliance

Test:

System Type:

Testing for:

What you need:

Drawbacks

Lock Chamber
or Bar

Single Beam

Accuracy

A lock chamber at a
known depth.

Hard reflective bottom
does not reflect actual
conditions.

Lock

Chamber
or Bar

Multibeam

Accuracy

A lock chamber at a
known depth.

Hard reflective bottom
does not reflect actual
conditions.

Cross Check
Statistics

Single Beam

Repeatability

Survey lines that
intersect.

Measures repeatability,
not accuracy.

Total
Propagated
Uncertainty

Multibeam

Theoretical
Accuracy

System parameters,
sensor locations.

In
thoery
, I can break

80
on a golf course. In
practice, I cannot.

Performance

Test

Single Beam

Repeatability

Performance test
area.

Measures repeatability,
between multibeam
and single beam
systems.

Performance

Test

Multibeam

Repeatability

Performance test
area.

Measures repeatability,
not accuracy.

Lock Chamber

Single Beam

Make a note of the
depth in the lock
chamber (or the depth
of the bar).

SINGLE BEAM EDITOR:

Edit spikes as you
normally would.

Click ‘File

Offline
Statistics’.

HYPACK 2012 will show
the Average Z
-
value and
the Std. Dev. for the Z
-
value.

Lock Chamber

Single Beam

You have tested your accuracy at a single depth.

The acoustic properties over the concrete bottom of the lock chamber
will be different than the same properties over a hard or soft bottom.

You have demonstrated how accurately you can determine the bottom
of the lock chamber with your acoustic system.

I could argue you have not demonstrated the accuracy of your system
over different types of bottom conditions.

Setup

+/
-

Depth Gate: Soundings outside bar
depth +/
-

the gate are ignored.

+/
-

Angle Limit: Soundings with beam angle
outside limits are ignored.

Running The Test

Run Bar Check from the Tools menu.

Click “Reset Barcheck.txt” to clear the report.

Lower the bar and enter Bar Depth.

When Measured Depth stabilizes, click
“Save Depth”.

Repeat for each bar depth.

When done, click “Barcheck.txt” to view or
print the report.

Bar Check averages depths for three
seconds then saves and graphs the result.

Lock Chamber

Multibeam: Bar Check Routine

Lower a bar to a fixed depth, or use a
lock chamber. Then run MBMAX

BarCheck.TXT Sample

Lock Chamber

Multibeam: Bar Check

If you use a bar check, it’s damn hard to hit the bar with
enough beams at a typical survey depth.

Lower the bar to 10’, we get hits on beams +/
-

60˚, but we’re going
to be surveying at 30’.

Lower the bar to 30’, we get occasional hits on beams +/
-

20˚.

In a lock chamber:

You may have to turn 90 degrees to keep the outside beams from
hitting the walls of the chamber. Now you are assuming the lock
chamber is level in that direction…..

For both:

The acoustic conditions aren’t the same as the ‘real
world’.

CROSS CHECK STATISTICS

SINGLE BEAM

SBMAX

Run the CROSS CHECK
STATISTICS program

Program

Computes:

Arithmetric

Mean: (Average
Difference)

Difference

Mean
(Average of Abs.
Value of
Differences)

Std.

Deviation

Minimum Diff.

Maximum Diff.

CROSS CHECK STATISTICS

Single Beam

Depth
Difference

Abs. Val.
Of Depth

-
1.15

1.15

0.32

0.32

0.05

0.05

0.52

0.52

-
0.26

0.26

-
0.03

0.03

0.48

0.48

0.96

0.96

-
0.32

0.32

-
0.28

0.28

0.14

0.14

-
0.37

0.37

Average:

0.01

0.41

St. Dev.

0.54

95% Conf.

1.07

Arithmetric

Mean = Average of Difference Values.

Difference Mean = Average of the Absolute Values of Difference Values

95% Confidence = 1.96 * St. Dev. (aw heck, just multiply by 2)

Arithmetric

Mean
is close to 0’.

Difference Mean is
close to 0.1’.

You are one tough
surveyor!

Arithmetric

Mean is
close to 0.

Difference Mean is close
to 2.0’.

You probably need a
heave
-
pitch
-
roll sensor.

Arithmetric

Mean
is close to 0.5’

Difference Mean is
close to 2.0’.

Something is
wrong.

CROSS CHECK STATISTICS

You should always run a couple cross check lines across your
survey lines to allow you run CROSS CHECK STATISTICS.

This is not a measure of accuracy. It is a measure of
repeatability.

How confident are you that you can go out and get the same values time and time
again.

You may find different results due to different weather conditions.

Flat a pancake…..

Light chop…..

We probably shouldn’t be surveying today…..

Can it show what % of my intersections met each standard?

Not yet, but we’re going to implement it in the next month or two.

Stay away from side slopes.

Total Propagated Uncertainty

Computes the
‘uncertainty’ for each
measurement of the
multibeam and single
beam system.

Use as a planning tool
or in real time.

Display the
uncertainty versus
survey requirements:

IHO Orders 1A, 1B &
Special

USACE Hard & Soft

During data collection, certain
estimates are replaced by actual values

Sample Graph

200 kHz
Fansweep

Overall system (as
defined only in this
example) meets IHO
Order 1 Positioning
requirements out to 55
°

Depth Uncertainty = TVU

Position Uncertainty = THU

General Parameters

Basic information concerning
the multibeam system.

Frequency

Number of Beams

Pulse Length

Along/Across Track
Beamwidth

Maximum Ping Rate

Estimated Depth of Bottom

Important when planning a survey.

Replaced by actual depth in real
time.

Survey Requirement Selection:

IHO Order 1 and Special Order

USACE Hard and Soft Bottom

Environment Tab

Contains items that
describe the
environment where the
data will be collected.

Estimated Speed of Sound

Peak
-
to
-
Peak Swell

Maximum Depth that SV

Other Stuff

Sensor Info
Tab

Contains:

Offset information for
HYSWEEP HARDWARE, if
possible)

Info on how accurately
you measured each
sensor position.

Estimated accuracy of
each sensor.

There are a lot of parameters in the TPU EDITOR. Most of you have
no clue on what to enter for 50% of them.

Surveyors just monkey with the parameters until they can justify
using all their outside beams.

Some of the parameters cause BIG changes to the computed TVU
and THU:

Pulse Length

Water Level Uncertainty

Spatial Tide Prediction

Fixed Heave Uncertainty

GPS RMS Error Estimate

IMHO: TPU is just one tool. I prefer the Performance Test which
uses real world measurements to determine the repeatability of

Performance Testing

Creating a Test Area

To Make a Reference Surface:

Pick a flat area. Small changes in position
will not affect your overall repeatability.

Set up a grid pattern to survey with your
multibeam vessel.

5 * Depth x 5 * Depth for up to 65
degre

systems.

7 * Depth x 7 * Depth for 65+ degree systems.

Create a MTX file with 1’ (30cm) cells around
the test area.

Do a velocity cast immediately before survey,

In MBMAX:

Process it, using only beams out to 45 deg.

Going into Stage 3, use the MTX file you created.

Do a great job of editing.

Save out an XYZ file of average values for each cell
where you have 3 or more samples.

Performance Testing

Multibeam

Run two lines through your test
area.

Edit the data in MBMAX, using all
of the beam data.

The better your editing job, the better

However, you just can’t delete all the
outside beam data, as you won’t have
enough samples to compute statistics
on them.

When finished editing in Phase 3,
run the Beam Angle Test

Beam Angle Test Graph

The ‘X’ markers show
the bias (absolute
value) between the
Reference Surface and

The ‘O’ markers show
the 95% confidence
for each beam.

Typically pretty flat out
to 45˚…..

….starts sloping up
between 45˚ and 60˚ …

….starts zooming up
beyond 60˚.

Beam Angle Test: Details Tab

Details Tab is Result at Specific Angle

Histogram is the distribution.

Max Outlier = Max difference.

Mean Difference shows bias between
reference and the check lines.

95% confidence = 2 * Standard Deviation.

Hard Bottom

Soft Bottom

Maximum
Outlier

1.0’

1.0’

Maximum Bin
Size

1.0’

1.0’

Maximum
Allowable
Mean Bias

<0.1’

<0.2’

Depth

Accuracy
(95% conf.)

D<15’ = 0.5’

15<D<40=1.0’

D>40

= 1.0’

D<15’ = 0.5’

15<D<40=1.0’

D>40

= 2.0’

BeamAngle.TXT Report

Sample Beam Angle Test Results

Which system do you have more confidence in?

Performance Test

Single Beam

Survey two single beam lines over
the Performance Test area you

Edit the data in SBMAX and save the
data to Edited ALL Format.

Take the Edited ALL files into EXPORT
and save them as a single XYZ file.

Load the XYZ file into MBMAX. It will
take you directly to Stage 3.

Use the same MTX file you used
when you created the Reference
Surface.

Select the Single Beam Test from the

Performance Test

Single Beam

Running the Test

Select Single Beam Test from the Tools menu in MBMAX. Open
the Reference XYZ file.

Results

Once again using difference
statistics = reference minus check.

Histogram shows the distribution
and bias.

Max Outlier = Max Difference.

Mean Difference = Bias. This is
the
main point of the test
!

Standard Deviation (Sigma):
distribution.

95% Confidence = 2*Sigma. 95%
of the differences fall within
2*Sigma.

Performance Test

Take a velocity cast immediately before you survey.

Run

the test at low tide.

Chances

are the velocity profile will remain constant
during the test.

Your

inability to accurately measure the tide will be
factored out.

Run

the test in the deepest part of the channel.

Stay

away from vertical changes in your test area.

Sometimes, you get what you pay for…

The End