The Mary Rose Site–Geophysical Evidence for Palaeo-Scour Marks

The hilernationd J014r?ifllqf Na~iticalArclineologj~ ( 1997) 26.1: 3-16

The Mary Rose site-geophysical evidence for palaeo-scour marks

R. Quinn, J. M. Bull, J. K. Dix* Deprirtnieiit of Geology, *Departnient of Oceanography, Southainpton Oceaiiographv Centre, University of Sozrtlianpon, Etnpress Dock. Soutliunipton SO14 3ZH, UK J. R. Adams Department of Archneology, University of Soirtkcimpton, Higllfield SO1 7 1BJ, UK

Introduction co-ordinated surveys. This trend is also Since the development-led boom in rescue reflected in the UK Government’s inter- archaeology in the 1960s and 1970s, the pretation of current legislation, in particu- vicissitudes of funding and the growing lar with respect to historic wrecks, as well recognition of the non-renewable nature as dove-tailing with the recently estab- of the archaeological resource, have lished database for maritime sites at the prompted much more circumspect atti- National Monuments Record Centre at tudes to excavation (Cleere, 1989; Hunter Swindon, UK. & Ralston, 1993). Hence the justification Over the past 30 years, various marine of decisions to excavate are generally based seismic reflection techniques have been either on the threat from development or used in the investigation of sites of other agency, or the need for research, archaeological interest, (McGhee et al., where excavation is viewed as the best or 1968; Frey, 1971; Chauhan & Almeida, only way to answer certain questions, or 1988; Rao, 1988; Redknap, 1990). To date, occasionally both. As a result, greater the application of sub-bottom profiling emphasis is now based on desk-top assess- systems to maritime archaeology has ments, non-intrusive evaluations such as been restricted by poor resolution and an fieldwalking and geophysical surveys, and inability to image the seabed in very on the associated task of building Sites’ shallow water depths. Recent advances in and Monuments’ Records (SMRs) to both acquisition and processing techniques better quantify and manage the resource. have culminated in the development of Over the same period archaeology a system known as ‘Chirp’, a digital, underwater focused on the development frequency-modulated (FM) sub-bottom of appropriate excavation and recording profiling system. methods and the collection of data. Since The principal objectives of this paper are the mid 1980s, this area of research has to present the results and associated inter- moved in a broadly similar direction, pretation of a three-dimensional Chirp concentrating less on the excavation of sub-bottom survey of the excavated Mary single sites (often as a reaction to chance Rose site carried out in conjunction with discoveries) and more on assessment and the Mary Rose Trust, and discuss the recording. An increasing number of application of Chirp technology to mari- projects in coastal regions are more pro- time archaeology. In order to familiarize active in character, involving regionally the reader with the principles of operation 1057-2414/97/010003 + 14 $25.00/0 na970060 1997 The Nautical Archaeology Society NAUTICAL ARCHAEOLOGY, 26.1

I I Transmitted ray

Figure 1. Reflection and transmission from a boundary in the sub-surface. A boundary giving rise to a reflection will be marked by an acoustic impedance contrast across it. The incident ray is partitioned into a reflected ray and a transmitted ray. The proportion of incident amplitude that is reflected is controlled by the reflection coefficient. See text and Equation 1 for more details. of the sub-bottom profiling system, the of the reflected signals. Assuming a vel- following two sections concentrate on ocity at which the acoustic signal propa- seismic reflection theory and introduce the gates through the sub-bottom sediments, it aspects of Chirp sonar which set it apart is possible to estimate the depth of reflec- from conventional sub-bottom profiling tors. In seismic reflection theory, a material technology. For those interested, more is characterized by its compressional wave detailed discussions on the principles of velocity (V,) and density (p). The product acoustics can be found in many general pV, is known as the acoustic impedance. geophysics texts (including Anstey, 1981; If a contrast exists between the acoustic and Sheriff & Geldart, 1995), and the impedance of two materials (plVpl and systematics of Chirp design and operation p2V,,), then a reflection occurs at the is dealt with more comprehensively in boundary of the media (Fig. 1). The Schock & LeBlanc (1990) and Parent & strength of this reflection is governed by O'Brien (1993). Furthermore, a Notes the reflection coefficient, K,, where: section is provided at the end of the paper providing definitions for some of the KR = p2vp2 - plvpl (Equation 1) terminology and concepts introduced. PJp2+ PJ,, This equation means that the higher the The seismic reflection method contrast between the acoustic impedances The seismic reflection method utilizes the of the media, the stronger the reflection propagation of waves through the earth. from the boundary. Values of K, range Seismic reflection profiling is accomplished between - 1 and + 1. If KR= 0, then the by transmitting an acoustic signal through incident compressional wave is entirely the water column into the underlying sedi- transmitted, and no reflection occurs. Con- ments and measuring the time interval versely, if K,= - 1 or + 1 then complete between pulse transmission and the arrival reflection of the incident wave occurs,

4 R. QUINN ET AL.: THE MARY ROSE SITE

- 0 32 Time (ms) Figure 2. An example of a 32 ms frequency-modulated Chirp pulse, sweeping between 2 and 8 kHz (reproduced courtesy of GeoAcoustics Limited, Great Yarmouth, UK). Note the frequency content increases to the right. and no energy is transmitted across the ics is the advent of powerful and affordable boundary. digital electronics and desktop computing In geological terms, values of K, tend to facilities. Chirp technology utilizes these fall in the range of h0.1 (Anstey, 1981) developments to aid processing and acqui- with the majority of the energy being trans- sition techniques to produce high quality, mitted. Where anthropogenic materials high-resolution sub-bottom images in real such as wood are brought into contact time. The most obvious aspects of Chirp with unconsolidated marine sediments, sonar that set it apart from conventional values of K, tend to have a broader range systems are the increased vertical resolu- (Quinn et al., 1996). Historically, wooden tion"] of Chirp systems, pulse repeatability wrecks are composed of oak, with lesser and the acquisition of data with a high components of pine. mahogany and elm. signal-to-noise ratio"] (SNR). Quinn et al. (1996) demonstrated that Unlike conventional short-pulse, single- theoretical reflection coefficients between frequency systems (e.g. pingers and oak and unconsolidated marine sediments boomers), the Chirp sonar transmits a range between - 0.03 and - 0.63 (values linearly swept, frequency modulated (FM) considerably higher than those found in pulse, i.e. the frequency of the pulse normal sub-surface geological situations) changes linearly with time (Fig. 2). In indicating a high probability of imaging conventional sub-bottom profilers, a trade- wooden artefacts using an appropriate off occurs between penetration and vertical sub-bottom profiling system. resolution. A high frequency source produces high resolution images of the Chirp technology sub-bottom, but penetration is limited. Perhaps the most important recent devel- Conversely, low frequency pulses penetrate opment in high-resolution marine geophys- deeper, but produce less-detailed sections. 5 NAUTICAL ARCHAEOLOGY, 26.1

Pinger’s typically operate in a frequency signal or noise energy does not match the range of 3 to 12 kHz, corresponding to outgoing Chirp pulse, the matched filter vertical resolutions of the order 20 cm. The attenuates the unwanted signals thus pinger’s depth of penetration is very lim- improving the SNR of the acquired data. ited however, typically tens of metres in fine muds and only several metres in coarse The Mary Rose site sediments. Boomers, with a higher energy The Mary Rose, King Henry VIII’s flag- output, offer higher depths of penetration ship, was built in 1509 and subsequently but vertical resolution is typically limited wrecked on 19 July, 1545. The wreck was to 0.5 to 1 m. The Chirp systems wide rediscovered in the Solent during the late bandwidth (typically of the order 10 kHz) 1960s (Fig. 3), and site excavation culmi- limits this trade-off between penetration nated with the raising of the hull remains and vertical resolution by transmitting a in 1982. Events leading to the location, range of frequencies, thus ensuring opti- excavation and subsequent raising of the mum penetration aid resolution. The Mary Rose are well documented (McKee, vertical resolution of the Chirp system is a 1982; Rule, 1982; Dobbs 1995). Today, function of the bandwidth of the trans- the raised hull structure is on display in mitted pulse, whereas the resolution of Portsmouth Dockyard, while the Mi4r-v conventional systems in entirely dependent Rose site itself remains one of 42 sites on the pulse length. Therefore relatively currently designated under the Protection low frequency wideband Chirp pulses can of Wrecks Act, 1973 (Archaeological be used to achieve high vertical resolution. Diving Unit, 1996). Elements of the wreck For example, a Chirp system bandwidth of remain buried on the seabed and the site is 10 kHz corresponds to a theoretical verti- constantly monitored by the Mary Rose cal resolution of 7.5 cm, assuming a com- Trust. pressional wave velocity of 1500 ms ~ ’ Prior to the positive identification of the (typical velocity value for the water wreck in the late 1960s, a number of geo- column). physical surveys were conducted over the With conventional systems, variations in supposed wreck-site in an attempt to locate the amplitude and phase of the transmitted the exact position of the Mary Rose (Rule, pulse lead to a problem in source repeat- 1982). In 1967, an anomaly was recorded ability. In the Chirp system, the problem of in the supposed position of the wreck by pulse repeatability is addressed by pre- John Mills using a combination of dual- processing the transmitted pulse, i.e. channel sidescan sonar and pinger. A year amplitude and phase weighting is applied later, John Mills was joined by Professor to the outgoing Chirp pulse. Source repeat- Harold Edgerton of the Massachusetts ability greatly aids post-processing of the Institute of Technology, and a second acquired sub-bottom data, as many of geophysical survey was conducted over the the algorithms applied to seismic data are site, using two pingers operating at fre- designed around the signature of the quencies of 5 and 12 kHz. Four anomalies transmitted signal. were recognised in close proximity to the Seismic data are usually corrupted with supposed site, one of which was inter- various forms of unwanted noise. FM preted as the buried wreck. The target pulse compression in the Chirp system is identified by the sub-bottom profilers performed using a matched filter (Schock provided sufficient impetus for the & LeBlanc, 1990), which correlates the archaeologists to excavate the site. recorded sub-bottom reflections with the During November 1994, 27 years after transmitted pulse. If any of the recorded these initial geophysical surveys, the

6 R. QUINN ET AL.: THE MARY ROSE SITE

Figure 3. Location map of the Mary Rose site located at Spithead, East Solent. The location of the differential GPS land receiver station at Warsash is indicated. authors surveyed the excavated Miivy Rose A pseudo three-dimensional survey, of site using a Chirp system. The aim of 10 m line spacing, was conducted over the survey was to investigate the site for the site (Fig. 4), covering an area of ap- remaining wreck structure, to put the proximately 350 x 250 m centred on the excavated site into a sedimentological con- excavation hole. Eleven east-west and 18 text with the advantage of detailed site north-south lines were acquired, totalling knowledge supplied by the Mary Rose over 6 km of Chirp data. The survey grid Trust and to assess the applicability of the was designed to provide close three- Chirp system to marine archaeology. dimensional cover in order to image potential remaining wreck artefacts. Tight three-dimensional coverage ensured any Survey equipment and methodology artefact material recognised in the east- The geophysical survey was conducted on west sub-bottom profiles could be readily 22 November 1994 aboard the research tied in with the north-south lines. vessel Mary Lisu, utilizing a 2 to 8 kHz swept frequency Chirp system. Through- out the survey, a 32 ms pulse length and a Data processing system transmit rate of four pulses per Sub-bottom profilers are capable of re- second was used in the acquisition of cording acquired data in digital format, digital sub-bottom data. Survey naviga- and can be subsequently processed to tion, with an accuracy of k 1 m, was pro- increase the SNR in order to aid data vided by a differential global positioning interpretation. This post-processing can be system (DGPS). purely cosmetic in terms of improving the NAUTICAL ARCHAEOLOGY, 26.1

96450

96400

96350 B c 2 z2 96300

96250

96200

I 0 50 100 150 200 m Figure 4. Shotpoint map of the three-dimensional Chirp survey of the Mury Rose site, Spithead, East Solent. The labelled lines are referenced in the text. type of display (e.g. colour or grey-scale The Chirp data was correlated with the as opposed to black and white), or known 2-8 kHz source sweep in real-time much more fundamental as is described during acquisition. Each survey line is cor- below. The most widely accepted digital rected for tidal variations over the survey data format for the acquisition of sub- period, and subsequently converted to a bottom data is SEG-Y format (the time-equivalent ordnance datum. Further Society of Exploration Geophysicists- processing of the correlated Chirp data Y format), and recording in this format was completed using ProMAX@ software ensures that the data can be read by any mounted on a Unix workstation. The pro- seismic processing software (as developed cessing sequence included true amplitude by the exploration industry). The princi- recovery[’]], bandpass filtering[41, FX de- pal aim of seismic processing is to convolution[51 and the application of a manipulate the acquired data in order dynamic signal-to-noise (S/N) filter (Fig. 5). to resemble as closely as possible the The first three algorithms in the process- sub-surface stratigraphy and to aid data ing sequence are standard seismic pro- interpretation. cessing applications (Hatton et d. 1986; 8 R. QUINN ET AL.: THE MARY ROSE SITE

7. Figure 6 is a section of the east-west Chirp sub-bottom profile MREW-10 (Fig. 4) showing a cross-section of the present corrections day morphology of the excavation hole. From divers logs of the site (Adams, 1988) True amplitude and the aqcuired Chirp sections, the local geology is interpreted as horizontal/ sub-horizontal intercalated muds, clays Bandpass filter and sands overlying bedrock of varying topography. The most striking feature recognised in I F-X deconvolution I the sub-bottom profiles over the entire site is the discrete bright-spot (high amplitude Dynamic signal-to- noise filter reflector) observed in MREW-8 (Fig. 7). A second such discrete anomalous reflector is observed approximately 30 m to the south of the two lines. These brightspot reflectors are unconformable with the local geology, where the horizontal/sub-horizontal soft- Figure 5. Flowchart of the sequence used in the sediments horizons terminate abruptly seismic processing of the correlated Chirp data. adjacent to the high-amplitude reflectors. Yilmaz, 1987). The application of the The base of the anomalies form erosive dynamic S/N filter as a final step is less surfaces, clearly post-dating the local stra- standard, and is particularly suitable to tigraphy. Examination of Fig. 7 indicates Chirp data as it eliminates the need for the majority of the incident energy is re- a conventional ‘late-stage’ time-variant flected at this horizon, indicating a very bandpass filter which is exceedingly diffi- high acoustic impedance contrast between cult to design due to the short data window the comprising material and the surround- (average 25 ms). ing sediments. The dynamic S/N filter is applied to en- In order to examine the spatial relation- hance the lateral coherency (or resolution) ship between the excavation hole and of the data by weighting each frequency by the brightspots referred to above, two- a function derived from the local signal-to- dimensional contour maps of the inter- noise ratio. Dynamic S/N filtering differs preted horizons were produced. Figure 8a from conventional coherency-enhancing is a contour plot of the sea-floor, produced methods as the standard methods include a by digitizing the seabed on each sub- portion of neighbouring traces through an bottom profile and interpolating across the addition process, which inherently leads to entire survey area. The most obvious fea- some form of lateral smearing (ProMAX ture on the contour plot is the excavation User Manual, 1993). In this case, however, hole located at the centre of the grid, although the filter is derived from sur- measuring an average of 50 m x 20 m. A rounding traces, it is applied to each trace similar method was used in the spatial in turn as an amplitude-only convolutional mapping of the brightspots. The maximum filter. interpolated amplitude value was extracted from each trace in a 15 to 19 ms time Results and interpretation window (thus encompassing the anom- Examples of two post-processed sub- alies), normalized and subsequently inter- bottom profiles are displayed in Figs 6 and polated across the survey area to produce 9 NAUTICAL ARCHAEOLOGY, 26.1

Figure 6. Processed portion of east-west orientated Chirp sub-bottom profile (acquired using a GeoAcoustics model 136A towed transducer system) MREW-10 showing a cross-section of the present day morphology of the excavation hole.

Figure 7. Processed portion of the Chirp sub-bottom profile MREW-8, displaying one of the ‘brightspot’ anomalies (highlighted by the black box). The anomaly is characterized by a discrete high amplitude reflector of length 60 m. the contour plot of Fig. 8b in which two the southern anomaly. The base of these east-west trending amplitude ‘highs’ are erosive features varies between 4.5 and 6 m seen towards the western flank of the sur- below the seabed, while relief is limited to vey grid. The sub-parallel anomalies are approximately 1.5 m. approximately 50 m in length, while the Caston (1979) identified wreck- northern feature (with a maximum width associated scour features from sonographs of 20 m) is approximately 5 m wider than acquired in the outer part of the Thanies

10 R. QUINN ET AL.: THE MARY ROSE SITE

Ffguue 8. (a) A two-dimensional time contour plot of the sea-floor over the M~yvRose site. Topographic lows are shown as black, and highs as light-grey. (b) A two-dimensional contoured amplitude map of the 15-19 nis time horizon. The normalised amplitude scale is from light-grey (minima) to black (maxima).

11 NAUTICAL ARCHAEOLOGY, 26.1

Gravel Wreck

~ 1Sand + Peak tidal flow

Figure 9. Diagrammatic representation of the end-member cases of the gradational series of longitudinal wreck marks observed from sonographs acquired in the outer Thames Estuary (modified from Caston, 1979).

Estuary. These erosional marks are associ- the current flow (Fig. 9a and b). The ated with scour around wreck obstacles shortest and narrowest scour features are lying on, or partially buried within, the those that emanate from wrecks aligned seabed. Caston showed that longitudinal precisely along the current, thus presenting wreck marks extend parallel to the peak a streamlined shape to the flow (Fig. 9c tidal flow on the downstream (stronger of and d). Wreck mark morphology is also the peak ebb or flood current) side of the dependent on seabed lithology; wreck wreck. Where such features occur on the marks on a sand floor consist of broad, upstream side of the wreck they are much shallow troughs whereas those on gravel smaller in dimension. The occurrence of floors are narrower and less extensive (Fig. single or double wreck marks was shown 9). A modern example of wreck marks is to reflect the orientation of the wreck shown in the sonograph of Fig. 10. The with respect to the peak tidal current flow. unequal development of the wreck marks The most common type of wreck mark in each direction is due to the oblique observed consists of twin scour shadows alignment of the wreck relative to the peak emanating from wrecks lying broadside to tidal flow (from left to right).

1’ R. QUINN ET AL.: THE MARY ROSE SITE

Figure 10. Sonograph showing a wreck (55 m long and a maximum of 14 m high) with scour hollows emanating from the bow and stern of the wreck (modified from Belderson et ul., 1982). The unequal development of the wreck marks is due to the oblique alignment of the wreck relative to the peak tidal flow.

96450 Peak tidal N - flow

96400

96350 i=o 30 // k

z 96300

96250

0- 50 100 96200 m 1 I I I I I I 463100 463150 463200 463250 463300 463350 463400 463450 Eastings Figirre 11. Composite map of the Mury Rose site displaying the position of the excavation hole (black) overlain on the spatial extent of the anomalies (grey) from Fig. 9.

Figure 11 is a composite map of the (275 to 279", Mary Rose Trust, pers. Mary Rose site, where the position of comm.) is also indicated. When the Mary the excavation hole (former position of the Rose sank in 1545, the hull lay heeled on its Mary Rose) is laid over the east-west starboard side in the seabed at an angle trending anomalies described above. The of 60" from the vertical (Rule, 1982). present-day peak tidal flow over the site The nature of the obstruction the hull 13 NAUTICAL ARCHAEOLOGY, 26.1 presented to easterly and westerly tidal geometry between the position of the flows was therefore very different. East- wreck and the morphology of the wreck west currents passed over and around the marks formed on a sandy floor. irregular surfaces of the castling and decks, The presence of these longitudinal scour while those flowing west to east would features was previously unrecognised on have been deflected by the smoother form the Mary Rose site. Large segments of the of the hull. The authors propose that hull superstructure and the stern and bow- the anomalies recognised on the western castles of the wreck eroded subsequent to margin of the excavation hole are the sinking. The authors propose that a pro- manifestation of palaeo-scour features portion of this material survives as wreck formed around the wreck of the Mary Rose fragments in the scour pits to the west of soon after she was sunk. the excavation. In the years after sinking, Caston (1979) further demonstrated that the hull structure standing above the longitudinal wreck marks are shallow seabed in the water column gradually (typically 1.5 m to 2 m) relative to their eroded and collapsed, so constituting pro- widths and lengths and may be curved near gressively less of an obstruction to tidal to the wreck but the distal portion is flow (Rule, 1982). Hence, any depressions always straight and parallel. This descrip- caused by the scouring action gradually tion is consistent with observations from filled in, sealing and preserving any timber the sub-bottom profiles (where maximum elements and other objects deposited relief on the anomalies is 1.5 m) and the within. The resulting fill in the scour pits shape of the anomalies in plan view (Fig. is therefore a stratigraphic record of this 11). When Fig. 11 is viewed alongside Figs interrelated process and indicates a direct 9 and 10, good correlation is noted in the relationship between the morphology of spatial relationship between the anomalies the seabed obstruction and its associated and excavation hole of the Mary Rose site, scour features. Furthermore, the authors and the scour marks and the position of propose that the strong reflection from the the wrecks from Figs 9 and 10. The dimen- interpreted wreck marks is due to the con- sions and erosive base of the anomalies are trast between the fill material (a combina- also consistent with wreck-associated scour tion of oak wreck fragments and coarse features. Additional supporting evidence grained sedimentary material deposited for this interpretation comes from the mor- within these pits due to preferential depo- phology of the wreck marks in relation to sition in topographic lows on the sea- the seabed lithology over the site. Caston bed) and the surrounding unconsolidated (1979) states that wreck mark morphology sediments. is dependent upon seabed lithology, where These proposals indicate the authors’ scour features formed on sandy floors belief that the Mary Rose site has had a around wrecks consist of longitudinally highly variable depositional history. This extensive, broad, shallow troughs as op- ranged from a highly dynamic erosive en- posed to narrower, deeper and less exten- vironment subsequent to the initial wreck- sive features on gravel floors (Fig. 9). The ing. followed by a period of relatively high outline of the interpreted wreck marks on deposition rates in the scour pits once 1 he Fig. 11 indicate these scour features are wreck structure had degraded to a level broad in comparison to the dimensions of concordant with the seabed, to a more the excavation hole. The Tudor seabed on acquiescent period subsequent to the depo- the Mary Rose site was a mix of clays and sition of a ‘hard shelly layer ’(Rule, 1982) sands (Adams, 1988), and a comparison of which sealed the Tudor levels over the site Fig. 11 with Fig. 9b indicates a similar in the late 17th or early 18th century.

14 R. QUINN ET .4L.: THE MARY ROSE SITE

Recognition and preservation of longi- for both the maritime archaeologist and tudinal scour features in the sedimentary the geophysical surveyor. Information re- record conveys important implications for garding the degradation and preservation the maritime archaeologist and the survey- of a wreck may be gleaned from the mor- ing geophysicist. For the archaeologist, phology and systematic excavation of these the buried longitudinal scour pits may be material traps. The dimensions of the filled high profile target sites where fragmented scour marks indicate these features may material from a degraded wreck is de- be more easily recognised on sub-bottom posited and preserved. The scours may profiles than the partially degraded wrecks also contain important evidence for the that were originally responsible for their archaeologist regarding the nature and formation. chronology of wreck degradation within a The Chirp profiling technique has dynamic environment. For the surveying proven itself as a rapid, non-invasive inves- geophysicist, the sheer dimensions of these tigative technique for sites of maritime material traps imply they may be more archaeological interest, and is therefore of easily recognised on the sub-bottom profile particular utility in the protection and than the actual target wreck. In extreme management of submerged archaeological cases, where a wooden wreck is so com- material. Rather than merely providing pletely degraded that the majority of the a technique for locating sites, high- superstructure is fragmented, the wreck resolution sub-bottom data can provide may not provide the concentration of information regarding the nature and coherent material required to produce a chronology of wreck degradation and a strong reflection, but material in associated more accurate definition of the extent of a scour pits may. site.

Conclusions Acknowledgments The Chirp profiles of the excavated Mqi The authors wish to thank the following Rose site have imaged structures pre- people for useful discussion and continued viously unknown on site, interpreted as support: David Peacock (Department of infilled longitudinal scour features associ- Archaeology, University of Southampton), ated with the sinking and subsequent deg- Chris Dobbs and Alex Hildred (The Mary radation of the Mary Rose in a dynamic Rose Trust), Neil Kenyon (Southampton tidal environment. The authors propose Oceanography Centre), and Sarah Draper that the fill material includes wreck frag- (The Hampshire and Wight Trust for ments from the superstructure of the Mcrrj' Maritime Archaeology). Finally, the com- Rose and coarse grained sediment prefer- ments of an unknown reviewer substan- entially deposited in scour hollows on the tially improved the style and content of the portside of the ship. final manuscript. This research is funded The preservation of longitudinal palaeo- by NERC Grant GR3/9533. scour features has important implications

Notes [ 11 Vertical resolutiorz: The minimum distance between two distinguishable adjacent reflectors. [2] Signal-to-noise ratio: The term signal is used to describe any event from which we require information, everything else in the seismic record may be attributed to noise. Seismic data quality is defined by its signal-to-noise ratio (SNR), i.e. the ratio of signal energy to noise energy in the acquired data. Good quality data will therefore have a high SNR. 15 NAUTICAL ARCHAEOLOGY, 26.1

[3] True ainplitude recovery: Applies a single time-variant function to the sub-bottom data to compensate for loss of amplitude due to wavefront spreading. [4] Bandpassjilter: This is applied to seismic data in order to suppress low- and high-frequency noise. [5] Deconvolution: This is the process of undoing one or more filters which are considered to have had an undesirable effect on the data. The successful application of deconvolution frequently increases the resolution of seismic data.

References Adams, J., 1988, Unpublished Mary Rose Trust Internal Report. Anstey, N. A,, 198 1, Seisniic Prospecting Instruments: Signal Cl~uracterrstrcsand Iiistrument Specgca- tioizs, Vol. 1, 2nd. edn. Berlin, Stuttgart. Archaeological Diving Unit, 1996, Guide to Historic Wreck Sites http://www.st-andrews.ac.uk/ - www-shir/deswrecks.html Belderson, R. H., Johnson, M. A. & Kenyon. N. H., 1982, Bedforms. In Stride, A. H. (Ed.), Offshore tidal sands: processes and deposits, 27-57. London. Caston, G. F., 1979, Wreck marks: indicators of net sand transport. Marine Geology, 33: 193-204. Chauhan, 0. S. & Almeida, F., 1988, Geophysical methods as a tool to explore submerged marine archaeological sites. In Rao, S. R. (Ed.), Marine Archaeologuvoj Indian Ocean Coiintries, Nat. Inst. Oceanog., 3-5. Goa, India. Cleere, H., (Ed.), 1989, Archaeological heritage management in the modern world. One World Archaeology, 9. London. Dobbs, C. T. C., 1995, The raising of the Mary Rose: archaeology and salvage combined, Underwater Technology, 21: 29-35. Frey. D., 1971, Sub-bottom survey of Porto Longo Harbour, Peleponnesus, Greece. IJNA, 1: 170-175. Hatton, L., Worthington, M. H. & Makin, J., 1986, Seisniic Data Processing: Theory und Pructice. Oxford. Hunter, J. & Ralston, I. (Eds), 1993, Archaeological Resource Munrzgetnent in the UK: An Introduction. Stroud. McGhee, M. S., Luyendyk, B. P. & Boegman, D. E., 1968, Location of an ancient Roman shipwreck by modern acoustic techniques-A critical look at marine technology, Marine Tec/z~rologySociety, 4th Annual Conference, Washington, D.C. McKee, A,, 1982. How We FOUIZ~the Mary Rose. London. Orsi, T. H. & Dunn, D. A,, 1991, Correlations between sound velocity and related properties of glacio marine sediments: Barents Sea. Geo-Marine Letters, 11: 79-83. Parent, M. D. & 0 Brien, T. F., 1993, Linear swept FM (Chirp) sonar imaging system. Sea Technolo,gy, 34: 49-55. ProMAX@’ Users Manual, 1993, Advance Geophysical Corporation. Quinn, R., Bull, J. M. & Dix, J. K., 1996, Shipwreck surveying and Chirp (sub-bottom profiler) technology. Paper given at Society for Underwater Technology, Third Underwater Science Sympo- sium: ‘Underwater Research and Discovery’, Bristol, March 1996. Rao, T. C. S.. 1988, Geophysical techniques to locate pre- historic sites and artefacts on the continental shelf. In Rao, S. R. (Ed.), Murine Archaeology of Iiidiun Ocean Coiintries, Nat. Inst. Oceanog., 73-77. Goa. India. Redknap, M., 1990, Surveying for underwater archaeological sites: signs in the sands. The Hidrographic Journal, 58: 11-16. Rule, M. H., 1982, The Murv Rose. The Excrivation and Raising of Henry VIIIs Flagship. London. Schock, S. G. & LeBlanc. L. R., 1990, Chirp Sonar: a new technology for sub-bottom profiling. Seri Technolog,: 31: 35-39. Sheriff, R. E. & Geldart, L. P., 1995. E*xploration Seisinology (2nd edn.). Cambridge. Yilmaz, O., 1987, Seismic Data Processing. Society of Exploration Geophysicists, Tulsa.