• Aucun résultat trouvé

Bridging the gap : creating near-shore bathymetric maps from multibeam swath of systems and conventional data

N/A
N/A
Protected

Academic year: 2022

Partager "Bridging the gap : creating near-shore bathymetric maps from multibeam swath of systems and conventional data"

Copied!
18
0
0

Texte intégral

(1)

Dist.:

LIMITED

E/ECA/NRD/CART/210 09 February 1993 UNITED NATIONS

ECONOMIC AND SOCIAL COUNCIL Original: ENGLISH

Economic Commission for Africa Eighth United Nations Regional Cartographic Conference for Africa Addis Ababa, Ethiopia

22-27 February 1993

BRIDGING THE GAP: CREATING NEAR-SHORE BATHYMETRIC MAPS FROM MOLTIBEAM SWATH SONAR SYSTEMS AND CONVENTIONAL DATA

(2)

Economic and Social Council ?i5™«

LIMITED

WP.7

ORIGINAL IN ENGLISH

EIGHTH UNITED NATIONS REGIONAL CARTOGRAPHIC CONFERENCE FOR AFRICA

Addis Ababa, 22-27 February 1993

REVIEW OF THE LATEST TECHNOLOGY IN CARTOGRAPHIC DATA ACQUISITION, MANIPULATION, STORAGE AND PRESENTATION, WITH SPECIAL

EMPHASIS ON POTENTIAL APPLICATIONS IN DEVELOPING COUNTRIES:

AUTOMATIC CARTOGRAPHY: DEVELOPMENT AND APPLICATION OF DIGITAL CARTOGKAPHIC DATABASES, INCLUDING DIGITAL TERRAIN MODELLING

Bridging the Gap: Creating Near-shore Bathvmetric Maps wultibeam Swath Sonar Systems and Conventional Data

f;Submitted by the United States of America)*

E/CONF.86/1

* Prepared by Stephen Paul Matula, National Oceanic and

Atmospheric Administration, National Ocean Service, Rockville,

Maryland

(3)

E/CONF English Page 2

SUMMARY

outputs in both digital and analog forms.

(4)

INTRODUCTION

In the past seven years, NOAA vessels have collected 106,000 square nautical miles of digital multibeam swath bathymetric data, in

waters primarUy deeper than 150 meters. These data have resulted I the production of twenty-four high quality deep water bathymetric maps, which have been compiled exclusively of multibeam

data.

Although modern multibeam surveys continue to collect data and expand knowledge along the relatively unexplored contj;n^^4^oP®

areas, other parts of the continental margins are also becoming of increasing interest. In fact, expression for information in the

near-shorl regions by scientists in state and local government

academic institutions, and commerce enterprises has grown for the

oast several years. Again, one of the conclusions from the

National Research Council's Commission on Engineering and Technical

Systems Marine Board report, dated October, 1991 *"" ;•?{";

primary interest expressed by the states and industries areJLn_ the shallow-nearshore regions." As a result, in order to attempt to address these needs, continued efforts will be made to ass^leJ;":

shore bathymetric maps which must be constructed of both off-shore multibeam data and in-shore hydrographic information.

MULTIBEAM DATA PROCESSING

Over the course of the past several years, methods have been developed to efficiently process multibeam data into a usaaie product. The schematic in figure 1 gives a b5^ef end-to-end overview on the current structure employed in creating bathymetric maps. Two procedures are used in the selection/culling process (Fia IB) . The first is a generalized filtering program to remove anomalous data. During this prefiltering process, rare erroneous data values are placed in a separate file which Prevents their influence during additional future examinations. The remaining data are passed to a culling program which has been designed to:

(1) statistically validate all data within selection areas, usually 250 square meters, (2) choose representative sounding values data

within the selection areas, (3) and produce a 5Gduc;l j?a** «t?

consisting of a maximum (deepest) depth, a minimum (shallowest) depth, and a depth value that is closest to the average for that particular selection area. Current processing retains between two

and twenty-five percent of the original dat* dur*"* *£*?The

selection and culling process, depending on the water depth. The

selected sounding values are transferred to a conuurcial ™*™F*

package, Radian Corporation's Contour Plotting System 1 (CPSl).

bPSl is used to derive the 250-meter gridded coordinate values on

a Universal Transverse Mercator (UTM) projection with the

horizontal North American Datum of 1983 (NAD83) (Fig. ID), and also

employed to develop and plot the resulting contour data (Fig. 1G).

(5)

!I'! 1!! I'!

'!! 'ii !i!I

! i I 1

\\\!

" ! I 1 '! I ' 1 i I !• ! ! .1 ' : 1 i ! ! I ! '! '! i 1

!'!I

! I ! 1

!'!•!t

'! 'I I 1

!

!! !I!

Ml\\ »»1>1U>S 111»»1 >< 11It1>

UN|iBiim»fnr•'Hfl«*»|IW»iiIOhhi»'hi I*iihmkaiVtt"i»«*iWMtiWi"«'»<••n ^HrhU,.W.,W.).l,'1*^MNIII.*I- IHHb11iltill_Za<>-ki^'i

icniiiii"ifiiilviino innmniiimtinii*i111™ ^Mimiwitiiiniirtiii.imwmiMmMmiwBMr.nti ■>Mitiiraiintmntiiatniiiimatti«iiiiiiiiNRiH»H»iimni •■MramntmimnBiintinnniiiiatiiinHnni«iMiaMiu«M HmmmainiHfmmmiinnnnhhiuihmhmmminwihhii «hiHtinntnt«itimi(tnwmfnnminniHimmnrtHmtmi HHHWUtKttUIHIINaHlUHMHUibllttl iiHtnaimt»wtumiwitaimuauiwuuuuiiu«utu;iitiuii»i HH

o o

(6)

Bathymetric map composition would be greatly simplified if all the digital data along the coast were homogenous. Under ideal conditions data would be of the same quality, quantity, and would use highly-accurate navigation systems such as those employed to collect the multibeam data. However, in reality many types of collection and navigation systems have been used over time to conduct in-shore surveys. As a result, these variations may cause added considerations during the construction of these in-shore

maps.

HYDROGRAPHIC COLLECTION AND SYSTEM HISTORY

Through an Act of Congress, hydrographic surveys have been used to chart the near-shore U.S. coastal waters, since 1807. These surveys have normally investigated particular features or navigational channels using site-specific collection methods.

Reconnaissance surveys have been conducted to evaluate little-known regions and are generally used in advance of detailed surveying.

Technologies and methods of data collection have evolved over the years. (Fig. 2) Initial surveys were collected near shore with leadline systems deployed vertically overboard to derive depths, which provided relatively good quality sounding values. As leadline technology matured, confidence increased in conducting off-shore investigations. Yet problems persisted in these deep water areas, such as: difficulty in obtaining non-vertical depth samples due to vessel movement, leadline migration because of unseen sub-currents, misreading depth values on the leadline, and calling out and recording incorrect depth values aboard the vessel (in early surveys values were actually called out by a leadsman and transcribed by a recorder). Because of these and other difficulties, there was a need for improved, less time-consuming systems to gather off-shore information.

Starting in the early 1920's the first echo-sounding device, the Sonic Depth Finder, was introduced to sample the ocean floor. In this system, depth values were determined by having an operator send a signal at a predetermined time and then listening on headphones for the returning echo. The operator would then examine a variable-control mechanism to judge the actual depth. Much skill was required to determine results, but it was a marked advance in

depth determination.

In 1925 the first fathometer, known as the 312 fathometer, was developed by the Submarine Signal Company of Boston, Massachusetts.

This system allowed an operator the ability to determine depth values by correlating the echo heard in his headset while noting the reading on a continuous-rotating flashing white-light circular depth scale. This system was more accurate and unlike the previous system, could be operated while the vessel was underway.

(7)

OFF SHORE COLLECTION SYSTEMS RANOE:

150 TO 10,000 METERS

MULTIBEAM

MULTIPLE AREAS SAMPLED

IN SHORE DATA COLLECTION SYSTEMS

RANGE:

0.8 TO 500 (NARROW) 1.5 TO 800 (WIDE) METERS

(LARGE &. SMALL FOOTPRINT) DUAL FREQUENCY SINGLE BEAM

BEAM ANCLES:

8 (NARROW) 10 (WIDE) DECREES

FREQUENCY:

200 (NARROW) 40 (WIDE)KHZ

RANGE:

I FOOT 150 FATHOMS

(SMALL FOOT PRINT)

NARROW SINGLE BEAM

BEAM ANGLE:

6 TO* DEGREES

FREQUENCY:

WTO200 KHZ

RANGE:

1 FOOT 4000 FATHOMS

WIDE SINGLE BEAM BEAM ANGLE:

20 TO 60 DEGREES FREQUENCY:

1 KHZTO21KHZ

(LARGE FOOT PRINT)

RANGE:

NEAR SHORE LEAD WEIGHT

LEADLINE

CALIBRATED ROPE/ CABLE

E/CONF English Page 6

— PRESENT

1984 MULTIBEAM DATA COLLECTION BEGINS

PRESENT

1980 DUAL BEAM SYSTEMS DEVELOPED

— 1970 NARROW BEAM SYSTEMS

- 1965 SURVEYS BEGIN DIGITAL DATA COLLECTION

— 1940 MID-RANGE 808 FATHOMETER

■ 1939 AUTOMATIC RECORDING FATHOMETER HUGHES VESLEKARI

■ 1933 HIGH-PRECISION DORSEY #1 FATHOMETER

■ 1925 312 FATHOMETER DEVELOPED

- 1920 ECHO SOUNDING DEVICE INTRODUCED

1837 FIRST HYDROGRAPHIC SURVEY

FIGURE 2: DEPTH SOUNDING TECHNOLOGY AND EVOLUTION

(8)

The first precision instrument deemed accurate enough for hydrographic work was the Dorsey fathometer #1, developed in 1933 and deployed on the Lydonia in 1934. This was truly a revolutionary system for its time as it incorporated the transmitter and receiver into one seawater-tolerant hull mounted unit called a transceiver, used an supersonic operating frequency of 17 kHz, incorporated a mechanism for vibration dampening, and had an operating range from 3 feet to 150 fathoms.

The ability to record depths automatically was a breakthrough introduced in 1939 with Hughes Veslekari graph-recording instrument. This was a 16 kH2 system that employed a 30 degree wide beam angle with an operating range from 7 to 1000 fathoms.

A workhorse for shallow to intermediate surveying was the 808 fathometer. Introduced in 1940, both permanently attached and portable variations of the device were used in off-shore and near- shore work* The 808 would be the primary echo sounding mechanism in these water depths for the next twenty-five years.

The mid-1960's saw the first revolution in digital collection of hydrographic data. This was a high water mark which would result in computerized data acquisition for subsequent systems. During this period several systems were deployed, with most of the instruments in the low frequency and relatively wide beamwidth (45 to 60 degrees) category. These systems were able to sample larger ocean floor areas, but did not always provide for the best reproduction of some features.

Narrow beam reflector systems, with beamwidths on the order of 6 to 8 degrees coupled with higher frequencies, were designed to more sharply define the bottom's character in the mid-1970's. These systems were able to achieve better feature fidelity in areas of sloping and rough terrain. However, because of the narrow sampling nature of the instrument, this compromised the amount of area

covered.

To address the limitations and innovations gained from previous single beam experience, dual frequency systems were employed in the 1980's. These systems, composed of a narrow, high-frequency beam for precise depth measurements, and a wider low-frequency beam to sample the surrounding area, were employed to improve the confidence in depths resolved in shallow waters. These modern systems are now collecting the highest quality data for in-shore

surveys.

(9)

E/CONF English Page 8

CONSIDERATIONS

In addition to the properties of the echo sounding devices, other elements are also critical in obtaining correct depth

determinations.

There are many factors in the water column that contribute to depth resolution such as: salinity, temperature, and pressure of the water; bubbles entrained through natural events or vessel movement;

organic materials and suspended sediments; vessel attitude and

associated turbulence.

The ocean floor's characteristics such as: substrate composition, orientation, and sea-floor roughness can all contribute in resolving vertical depth values.

Perhaps the greatest factor in depth determination is the association of geographical positions to depth values. Latitude and longitude coordinates, or geographical positions, are attained through the knowledge of fixed standard locations from which navigational processes can derive new positions. Navigation which comes from the Latin derivative "navigare", means "to direct ship."

This ship direction takes the form of making a determination of the present location, recording its location, planning the course to the next position, attempting to regulate or control the vessel to the ensuing position, and again making a determination of the new location.

Navigational techniques, like the methods used to calculate depth values have also progressed over time. The initial historic near- shore surveys used visual means along with sextants to resolve hydrographic positions. These early surveys' positions were derived from fixed known land features which could be used as standards. Off-shore positions were collected using dead-reckoning skills which required sailing from original known positions and

making estimates along the course.

In the late 1920's, a method of navigation control beyond visibility of shore signals was developed. It was called Radio Acoustic Ranging (RAR). In RAR, a charge of TNT was detonated close to the vessel conducting the survey. The explosion caused acoustic waves to be radiated through the water. Hydrophone listening devices were placed at known locations near shore radio stations. Once the detonation was received at the shore stations, the time was recorded and then transmitted by the radio shore station to the collection vessel at sea. Thus, through a sequence of timing events of sound propagation through water, it was then possible to provide more accurate positions for vessels of the era which were beyond the sight of land.

(10)

At the end of World War II, a special type of radar system call Shoran was developed to improve navigational accuracies. In this system, vessel position was obtained by precisely measuring the elapsed time between a transmitted pulse and a return signal from two fixed shore stations.

Improved long- and medium-range hyperbolic navigation systems were developed in the 1950's and 60's. The off-shore systems included Hi-Fix, Sea-Fix, Raydist, Loran and others. These systems were characterized by measuring phase shifts from two continuous-wave signals to determine position. This was accomplished by emanating signals from both a vessel conducting the survey and two or more shore stations radiating the same signal. These systems also had the capability of operating in a range-range mode, where the distance could be determined from radio signals emanating from two or more precisely-positioned shore stations.

The newest satellite navigation systems, such as the differential mode of global positioning satellites (GPS) have a wich higher probability of accuracy and repeatability for position locations.

".. .navigation is not an exact science. A number of approximations which would be unacceptable in careful scientific work are used by the navigator, because greater accuracy may not be consistent with the requirements or time available, or because there is no alternative," (Bowditch) Position approximations are generally made because of the relative uncertainty of the original position, equipment incapable of attaining sufficient precision, incorrect instrument readings, random error from noise sources, and errors from systematic system biases. Most mislocated positions are usually a result of one or more of these errors. It is not because the navigator intentionally wishes to misrepresent the position, rather these and other factors are beyond his control.

COMBINING MULTIBEAM AND HYDROGRAPHIC DATA

When completing in-shore bathymetric map areas, the current goal is to produce sufficient coverage to create a uniform modeled (gridded) surface while preserving as much integrity as possible from individual surveys. Although multibeam systems are used to attain patterns of consistent coverage below the 150-neter depth curve, shoaler areas must be augmented with hydrographic data.

Thus far, the primary purposes of hydrographic survey data have been to support nautical charting activities, such as deriving contour lines or providing spot soundings as needed on the given manuscript. Rather than a systematic sweeping of the ocean floor, which has been technologically impossible until recently, data have been primarily collected to check for the depths that would be

(11)

E/CONF English Page 10

considered hazards to navigation (Fig. 3). Thus, shallow areas have received both more (closer line spacing) and frequent (on periodic basis) coverage- Substantial efforts to review, select, and apply these values on the nautical charts have been, for the most part, through manual rather than automated methods.

SUGGESTED LINE SPACING (duked liooi) FOR AREA TO BE SURVEYED

FIGURE 3

(12)

To date, approximately 10,000 surveys have been conducted by the Coast and Geodetic Survey in the coastal waters (Fig. 4). As fiaure 4 illustrates, approximately 10 percent of the surveys have been collected and processed digitally, from roughly 1965 to the present; another 30 percent of the surveys have been converted to a digitized form, covering the period from 1930 to 1965; and the vast majority of surveys prior to 1930, some 60 percent, are in a

non-digital manuscript form. Some of the major coastal areas of

the continental U.S., such as the areas illustrated in figure 5, have not been surveyed since the late 1920's.

HYDROGRAPHICNOS DATA

10%

30%

60%

DIGITAL DIGITIZED NON-DIGITAL

1992

1965 1930

1837

10.000 SURVEYS

FIGURE 4

The surveys, once collected, act as historical records reflecting

the current technology and processing for that particular era.

When including hydrographic data, the preparations are somewhat like the multibeam methodologies with the following exceptions:

(1) If the data required in a map area are not in a digital form, decisions must be made in order to incorporate the information.

Presently a couple of alternatives can be exercised. First, the original manuscripts can be retrieved from the archives and a digitization process can be undertaken. This can often be a long and laborious process that requires a great deal of quality control to ensure the digital representation creates an authentic reproduction. A second approach is to have an experienced cartographer review the information, manually draw sufficient contour levels which satisfy the orientation of the survey, and transpose the contours though a digitization or scanning process.

in addition, software exists which can convert the contoured lines into a gridded model. However, depending on the data input and

personal interpretation, results can vary. As a result, clearly the

first method is preferable not only to preserve the data for

(13)

E/CONF English Page 12

SURVEY DISTRIBUTION

SAN FRANCISCO TO CAPE FLATTERY

(14)

perpetuity, but also to allow future interpretations and manipulations.

(2) The digital data must be converted to a common form in terms of depth units and reference locations for agreement with the multibeam data. Original depth values may have been collected in fathoms, feet, quarter-feet, or several other units of measurement and must be converted to meters. One must be aware that due to changing technologies the precision of measurement has generally increased over time. As an example, (Fig. 6) a newer system may have the ability to discern finer units of resolution, which may in turn yield slightly different results. In addition if not already adjusted, each position must be converted to a common set of coordinates on the North American Datum of 1983 (NAD83) for proper reference.

* 1 METEk—

UNITS COLLBCTBD

IN .

WHOLE * METERS

- t METER.

NEW

* l.tZUUC METERS"

_

- 1.82U036 METEK m

SYSTEM PRECISION Figure 6

"♦ I FATHOM

UNITS COLLECTED INWHOLE FATHOMS

B- 1 FATBOM OLD

(3) The data must be examined for density, orientation, and valid data values. Data density can vary quite widely between historic hydrographic and multibeam data. In figure 7, the upper left portion of the map contains the outer edges of a hydrographic reconnaissance survey, while the center of the map represents a portion of a multibeam survey with only selected soundings. Thus, the data examination is accomplished by creating depth position plots for each separate survey? selecting parameters for grid and contour generation based on coverage and orientation; removing definite erroneous data points; and re-visiting the entire process once acceptable data have been obtained.

(4) Survey overlap and influence must be determined to establish data coverage and resolve conflicts in all surveys in the region.

(15)

E/CONF English Page 14

r~**±sr^-**Z.

DATA DENSITY TOP: CONVENTIONAL DATA MIDDLE: CONTOUR DATA BOTTOM: MULTIBEAM DATA

(16)

The decision process can be further complicated with multiple overlapping irregular survey junctions (Fig. 8). Additional data selection is generally predicated on a progression of the most current and densest hydrographic data ranging down to the most historic and sparsest.

NOS SURVEY INDEX

36*30'

FIGURE 8

In a theoretical representation in figure 9, a section of multibeam data (lower left) is directly merged with a historical hydrographic data (upper right). The result is a compromised area that will most likely yield undesirable results when contoured. The solution for this example is to remove the intruding interior lines of data along with most of the historic data at the boundary to produce a reasonable solution. Although data elimination is an acceptable conclusion for most multibeam-to-hydrographic data interfaces, other hydrographic-to-hydrographic data areas may require further attention. In figure 10 two fairly sparse surveys are required for inclusion. Investigations and examinations have shown each data set to be consistent, significant, and required to derive an optimal depiction. As a result, the intruding data cannot be automatically eliminated, nor can the boundary data be reduced.

{5) Surveys are gridded on an individual basis to retain the feature integrity and reduce potential bias from adjacent data.

Surveys can be grouped together and processed if they have used the same type of instrumentation, navigation, line spacing, and have good agreement among them as the net result will be the same.

(6) Once all grids have been resolved for the individual surveys and the junction areas, they are then mosaiced together to form a

(17)

LINE INTRUSION

OVEKLAP AREA LINE OF DATA HISTORIC HYDROGRAPHIC DATA

SURVEY BOUNDAST

POSSIBU MULTIBEAM TO HISTORIC HYDROORAPHIC DATA INTERFACE Figure 9

IIHIVET

■oaaduf val***

POSSIBLE HYDROGRAPHIC TO HYDROGRAPHIC INTERFACE Figure 10

complete surface model. Under normal conditions, multibeam data are always given the highest precedence due to its increased density, higher quality, and its contemporary nature. This modeled surface is used to create the final map representation.

CONCLUSION

Throughout the history of the Coast and Geodetic Survey, hydrographic surveys have evolved and improved over time due to advances in echo sounding technologies, better navigation systems,

incorporation of progressive technologies and applications of advanced methods to process the information. This evolution has lead to better support of the traditional mission of discovering and reporting hazards to navigation for nautical charting activities- Another direct result of the progression in surveys is the development of higher-quality data products.

Most hydrographic surveys were not specifically designed for computer operations, particularly from the time period prior to 1965, yet future trends indicate that computer-assisted methods will predominate in the future. As a result, one of the biggest obstacles in creating near-shore computer-generated bathymetnc maps and associated digital products may be the conversion of

hydrographic data to a digital form that can be readily

incorporated with the off-shore multibeam digital data.

(18)

For in-shore areas the hydrographic challenges are: to provide for adequate digital data coverage; attempt to retain individual survey characteristics and integrity; resolve survey differences and eliminate spurious data; and provide for gradual progression between differing survey collection techniques and technologies.

The desired result is to produce an aesthetically pleasing bathymetric product that can be used by a wide range of scientists,

oceanographers, engineers, environmental planners, educators and

the general public in both a printed cost-effective paper map form and computer ready digital data set.

Currently, in-shore hydrographic data are fit to the same 250-meter grid model that has been used in the off-shore regions. Although this methodology may be satisfactory for some applications by providing a continuous homogenous surface from the shallow to deep waters it should be noted that small micro-topographic features may be attenuated during the selection, gridding, and contouring

processes.

Through planning, examination, review, and consolidation, it is

indeed possible to produce near-shore bathymetric maps by blending both off-shore multibeam data along with in-shore hydrographic data. However, the precision and accuracy of the combined bathymetric map is always a direct result of the information and adequacy of the individual surveys from which it is compiled and therefore cannot be more accurate.

REFERENCES

Adams, K. T., Hydroaraphic Manual. Special Publication Number 143, Coast and Geodetic Survey, U.S. Government Printing Office,

Washington, D. C, 1942.

Bowditch, N., American practical Navigator, Defense Mapping Agency Hydrographic Center, Publication Number 9, Vol. II, p.429, 1975.

Jeffers, K. B., HYHr»CTT*aphic Manual. Publication Number 20-2, Coast and Geodetic Survey, U.S. Government Printing Office, Washington,

D. C, 1960.

Matula, S.P., using Exclusive Economic Zone Digital Swath Data to ct Effective Bathymetry. Proceedings: OCEANS 86, November 1986.

NOTE

No product endorsement is intended or implied by the National

Oceanic and Atmospheric Administration.

Références

Documents relatifs

These two examples raise indeed several questions about the relationship between tourist consumption and heritage preservation, about the right to visit heritage sites and about

Winter mixed layers observations are rather scarce in the Western Nansen Basin. In particular, favorable comparisons with A ‐ TWAIN mooring time series of deep winter mixed layers

d) Explain one important difference between a datum transformation and a coordinate transformation. a) Sketch the graticule appearance of the Universal Transverse Mercator (UTM)

Although programmable calculators may be used, candidates must show all formulae used, the substitution of values into them, and any intermediate values to 2 more significant

Received: 29 April 2020; Accepted: 28 June 2020; Published: 3 July 2020 Abstract: Automatic cleaning of MultiBeam EchoSounder (MBES) bathymetric datasets is a critical issue in

Principle 8: Ensure that all teams within a company, especially the BPM team and the SAP team, develop the same processes and process maps and that only one process

This process model should list necessary steps for data profiling, describe the involved components (i.e., algorithms, tools, people) and their interactions, and demonstrate the

6.2.2 A Weaving Model between Enterprise Goals and the Data Warehouse Model The first approach to create business metadata for DWHs exploits the relationship between decision