23. SEISMIC MODELING OF DIAGENETIC EFFECTS IN CENOZOIC MARINE SEDIMENTS ATDEEP SEA DRILLING PROJECT SITES 612 AND 6131

David Goldberg, Borehole Research Group, Lamont-Doherty Geological Observatory and Department ofGeological Sciences, Columbia University

Roy H. Wilkens, Earth Resources Laboratory, Massachusetts Institute of Technologyand

Daniel Moos, Borehole Research Group, Lamont-Doherty Geological Observatory2

ABSTRACT

Velocity and density measurements at DSDP Sites 612 (slope) and 613 (rise) were obtained in the laboratory and bygeophysical logging. Discrepancies between the laboratory and in situ measurements that can be attributed to porosityrebound vary more strongly with lithology than with depth.

The in situ velocity and density logs were used at both sites to generate synthetic vertical seismic profiles. Surfaceseismogram phases are traced to their origin depths and correlated with nearby common depth points from USGS seis-mic lines 25 and 35. The resulting seismic stratigraphy is generally consistent with published interpretations. The corre-lation is poorer at Site 613, however, because of significant localized variations in bathymetry and sediment thickness inthe 4-km offset from line 35.

Bulk density and compressional velocity logs show increases across a diagenetic boundary (porcellanite) in the mid-dle Eocene sediments. The change is sharp at Site 612 and gradational at Site 613; this dissimilarity generates amplitudeand phase differences between synthetic seismic reflections at the two sites. The correlation of this boundary with Re-flector Ac suggests that changes in physical properties may generate seismic reflections which cannot be simply relatedto lithologic changes or depositional times.

INTRODUCTION

Results of recent DSDP investigations along the sea-ward edge of the Baltimore Canyon Trough have sub-stantially improved the interpretation of seismic data avail-able to the research community. The depth correlationand lateral continuity of seismic reflectors in this regionare still ambiguous, however. DSDP Leg 95 was designedto further improve seismic stratigraphy by well controlin a transect across the continental margin. Sites 612and 613 were drilled during Leg 95 and reliable geophys-ical logs were recorded. These logs constrain interpreta-tions of the seismic results and aid in the delineation ofvertical and lateral changes in lithology. In this chapter,the Cenozoic seismic stratigraphy is reinvestigated at bothLeg 95 sites.

A quantitative geologic model of seismic reflectionsrequires accurate determination of the compressional ve-locity and bulk density of the rock. The validity of themodel is directly related to the reliability of these physi-cal property measurements. Both measurements can beobtained in the laboratory and in situ. Unfortunately,measurements in the laboratory contain large errors re-sulting from drilling disturbance and dilatancy of thesamples, and the depth control is limited by the percent-age of core recovery. The in situ log measurements do

1 Poag, C. W., Watts, A. B. ; et al., Init. Repts. DSDP, 95: Washington (US. Govt.Printing Office).

2 Addresses: (Goldberg and Moos) Borehole Research Group, Lamont-Doherty Geologi-cal Observatory, Columbia University, Palisades, NY 10964; (also for Goldberg) Departmentof Geological Sciences, Columbia University; (Wilkens) Earth Resources Laboratory, Massa-chusetts Institute of Technology, Cambridge, MA 02142.

not contain these uncertainties. Thus, the reliability ofthe physical property measurements is improved, andthe relationship between geology and seismic reflectionscan be more accurately modeled through the use of logmeasurements.

SEISMIC REFLECTION DATA

U.S. Geological Survey (USGS) 48-fold seismic lines25 and 35 were recorded by Geophysical Services, Inc.(GSI) in mid-1978 using a DFS IV recording system. Site612 is approximately 0.1 km north of line 25 (Fig. 1), in1414.3 m of water near shotpoint 3060. Total penetra-tion at this site was 675.3 m sub-bottom, which corre-sponds to about 2.6 s total two-way traveltime. Site 613is approximately 4 km northwest of line 35, in 2333 mof water near shotpoint 1445; total penetration was581.7 m, corresponding to about 3.8 s total two-way trav-eltime.

USGS line 25 is the most intensively studied dip linein the Baltimore Canyon Trough. The major reflectionsin line 25 seaward of the base of the slope typify theseismic-stratigraphic framework of the continental mar-gin. Common depth point (CDP) traces from seismicline 25 near Site 612 and the Continental Offshore Strati-graphic Test (COST) B3 well are displayed in Figure 2.The major reflections apparent in the section have beenthoroughly discussed in the literature (Tucholke, 1979;Tucholke and Mountain, 1979; Schlee, 1981). In gen-eral, the reflections correspond to lithologic changes andepisodes of nondeposition or erosion. A widely correlat-ed marker horizon (Reflector Ac) in the middle Eocenesediments is associated with the onset of silica diagene-

589

D. GOLDBERG, R. H. WILKENS, D. MOOS

39°N -

38°50' -

38°40' "

73°W 72°50' 72°40' 72°30'

Figure 1. Map of Baltimore Canyon Trough showing locations of USGSseismic lines 25 and 35 and DSDP Sites 612 and 613 (after Robb etal., 1981).

sis from opal-A (biogenic) to opal-CT (porcellanite) (Wea-ver and Wise, 1974; Riech and von Rad, 1979). The por-cellanite typically occurs near the lower Eocene/middleEocene boundary, although it is not necessarily concor-dant with biostratigraphic changes.

Line 35 is aligned with the strike of the margin sea-ward of both DSDP sites (Fig. 1). CDP traces from seis-

mic line 35 near Site 613 are displayed in Figure 3. Sev-eral seismic reflections, including Reflector Ac, can betraced to line 35, suggesting that the horizons are con-tinuous downslope. However, a simple correlation of re-flections with lithologic boundaries (and subsequentlywith biostratigraphic ages) may oversimplify the strati-graphic relationships. Complex stratigraphic geometriesmay significantly affect reflections and vary over lateraldistance. In this study, seismic modeling suggests thatchanges in Reflector Ac occur between Sites 612 and613.

PHYSICAL PROPERTY MEASUREMENTS

Laboratory Data

Porosity is measured in the laboratory by the weightdifference between saturated and dry sediment samples.Bulk density is determined by dividing the sample wetweight by its volume. A detailed description of the tech-nique and data is presented in the Site 612 and 613 chap-ters (this volume). The cross-plots of bulk density andporosity in Figures 4 and 5 summarize the results for 115samples from Site 612 and 70 samples from Site 613.Each of the five lithostratigraphic units identified in therecovered core (Site 612 and 613 chapters, this volume)is represented by a different symbol.

The average grain density for each lithostratigraphicunit can be obtained by the zero-porosity intercept of alinear fit to these data. Lines of constant grain densityspanning the range of values at each site are shown inFigures 4 and 5. Grain density calculations (pg) by least-squares regression in each lithostratigraphic unit are shownin Table 1. The grain density in these samples is deter-

0.5

1.0

1.5

o 2.0

2.5

3.0

1.0 km

Site 61

Figure 2. Site 612 and COST B3 projected to USGS seismic line 25. Penetration to 675 m sub-bottom is approximately2.6 s total two-way traveltime.

590

2.51—

3.0

3.5

4.0

4.5

5.0

SEISMIC MODELING OF DIAGENETIC EFFECTS, SITES 612 AND 613

Site 613

Figure 3. Site 613 projected to USGS seismic line 35. Penetration to 585 m sub-bottom is approximately 3.8 s total two-way traveltime.

k de

nsity

2.3

2.2

2.1

2.0

1.9

1.8

1.7

1.6

1.5

1.4

1.3

• Unit I• Unit IIx Unit III* Unit IV

Unit V

30 35 40 45 50 55 60Porosity (%)

65 70 75 80

Figure 4. Shipboard measurements of bulk density versus porosity,Site 612. Symbols refer to lithostratigraphic units. Lines of calcu-lated grain density are shown over the range of values for these da-ta. Least-squares fit to the data for Units II and III yields graindensities of about 2.45 g/cm3, corresponding to chemical compo-sition 40% carbonate and 60% biogenic silica (see site chapters).

mined essentially by the relative proportions of carbon-ate (pg = 2.11 g/cm3) and biogenic silica (pg = 2.20 g/cm3) in each lithostratigraphic unit. Note that the graindensities in corresponding units at Sites 612 and 613 agreeto within about 3%. However, the standard deviation ofthe grain density calculations varies between 3 and 8%.Therefore, small differences in grain density between units,

j/cm

"ilk

den

sity

(s

CO

2.3

2.2

2.1

2.0

1.9

1.8

1.7

1.6

1.5

1.4

1.3

o-

X

-

• >

i

s × × "

\ y

Xj1

i i

• Unit 1• Unit IIx Unit III

xX 2.7

2.3 ^Ns^N.

1 I rs. >v30 35 40 45 50 55 60

Porosity (%)

65 70 75 80

Figure 5. Shipboard measurements of bulk density versus porosity,Site 613. Symbols refer to lithostratigraphic units. Lines of calcu-lated grain density are shown over the range of values for thesedata. Least-squares fit to the data yields grain densities of about2.51 g/cm3 for Unit II and about 2.43 g/cm3 for Unit III. Thiscorresponds to a higher carbonate content in Unit II.

such as Units II and III, cannot be effectively distin-guished by this statistical technique.

Comparison of Laboratory and in Situ Porosity

The sediment porosity measured in the laboratory maybe higher than in situ values because of expansion of thesample. This dilatancy results primarily from elastic re-

591

D. GOLDBERG, R. H. WILKENS, D. MOOS

Table 1. Grain densities (pg) calculated by least-squares regression.

Log porosity - lab porosity

- 2 0 - 1 0 0 10 20

Unit

IIIIIIIVV

pg

2.6122.4422.4692.5502.648

Site 612

R

0.6990.7320.7860.9570.988

TV

263138146

pg

2.6792.5122.428

Site 613

R

0.9880.8920.872

N

163222

Note: R = coefficient of correlation; N = numberof samples.

bound after removal of the in situ overburden stress (Ham-ilton, 1976). Since the overburden stress increases withdepth, the elastic rebound similarly increases, offset on-ly by lithification of the sediment (Shipley, 1983). Thedepth at which lithification effects become importantdepends on the extent of diagenesis in the sediment col-umn. Changes in lithology, however, also have significanteffects on the elastic rebound (Hamilton, 1976). Mayer's(1984) calculated rebound correction factor for deep-seacarbonates predicts a 5.5% maximum porosity increasefor samples raised from 400 m depth of burial. A pre-vious comparison of in situ and laboratory measurementsby Shipley (1983) has substantiated a porosity correctionof 9% in different sediment types.

At Site 612, a formation density log was recorded. Inthis logging experiment, medium-energy gamma rays areemitted from a downhole source, collide with formationelectrons, and lose energy because of the Compton scat-tering effect. The energy of recorded gamma rays is re-lated directly to the electron density which, in turn, isrelated to the bulk density of the formation (Schlum-berger, Ltd., 1972). Porosity is then calculated using thebulk density log and the average grain density of the for-mation obtained in the laboratory (Table 1). Hence, theporosity calculation is improved over standard techniquesby using grain density measurements instead of an as-sumed matrix value.

The difference between laboratory and in situ porosi-ty values at Site 612 is displayed versus depth in Figure6. The log values are plotted at depths coincident withlaboratory measurements, and each lithostratigraphic unitis plotted using a different symbol. In situ and labora-tory values agree to within about 10% as both generallydecrease downhole. The difference between log and labdata does not continuously increase with depth of bur-ial, though, but varies differently with depth in eachlithostratigraphic unit. This suggests that elastic rebounddepends more on lithology than on depth in these sedi-ment samples.

Sonic Porosity

Porosity can also be empirically estimated using thecompressional-wave velocity. This independent approachis useful in the absence of direct measurements. Sonicporosity has been empirically related to compressional-wave slowness in compacted sediments by using a time-average equation (Wyllie et al., 1958). The relationship

100

200

300

400

500

600

• Unit IIo Unit III* Unit IV

Unit V

700 L

Figure 6. Calculated difference between in situ and laboratory poros-ity measurements in each lithostratigraphic unit, Site 612 (sym-bols). In general, in situ porosity is lower than laboratory porosity,owing to elastic rebound of the sediment samples (see text). How-ever, this difference varies more strongly between units than withdepth.

592

SEISMIC MODELING OF DIAGENETIC EFFECTS, SITES 612 AND 613

is theoretically nonlinear, however, particularly at highporosities, as in uncompacted sediments (Wood, 1930;Officer, 1958; Shumway, 1960). Empirical relationshipshave also been defined to describe the results of labora-tory experiments on various seafloor sediments (Hamil-ton, 1982). Figure 7 shows a plot of the compressional-wave slowness versus porosity for data from Site 612,together with plots of Hamilton and Bachman's (1982)results from continental shelf data, Wood's (1930) equa-tion, and the time-average equation (Wyllie et al., 1958).

The apparent grouping of lithologies in Figure 7 sug-gests that the variation in velocity (reciprocal slowness)is controlled by changes in composition and porosity.Since the composition in Units II and III is similar(Table 1), the increase in velocity from Unit II to III canbe attributed largely to increases in compaction and ce-mentation. The data also range between the empiricalcurves for uncompacted and lithified sediments, as mightbe expected for partial lithification. Since Hamilton'srelationship for carbonate shelf sediments is nearly lin-ear in the limited 30 to 70% porosity range, the Unit IIand Unit III data alone can be reasonably fit by a linearregression. The regression of slowness (At in µs/m) as afunction of porosity (Φ in decimal units) is given by

At = 360.50 + 363.4 (1)

Using this relationship and the compressional-waveslowness, sonic porosity was calculated for Units II andIII. A comparison of the Site 612 density and sonic po-rosity values is shown in Figure 8. The data illustratethat sonic porosity is systematically higher than densityporosity in Unit II. This is not surprising, since the as-sumption of a linear relationship between slowness andporosity breaks down at higher porosities. However, the

700

665

630

595

560

525

490

455

420

385

350

Fluid velocity

1.6

1.8 -2

2.0 *

2.2

2.4

2.6

2.825 30 35 40 45 50 55 60 65 70 75 80

Porosity (%)

Figure 7. Traveltime (expressed as compressional-wave slowness) plot-ted versus log porosity calculated from bulk density and grain den-sity for each unit, Site 612. Line R23 is a least-squares regression toUnit II and Unit III, given by At = 360.5</> + 363.4. Also includ-ed are time-average (Wyllie et al., 1958) and shelf carbonate data(Hamilton and Bachman, 1982), and Wood's (1930) equation.

25 -

2020 25 30 35 40 45 50 55

Density porosity (%)

60

Figure 8. Comparison of porosity measurements in Units II and III atSite 612 from sonic and density logs. Sonic porosity is calculatedfrom a least-squares fit (R23) to the data in Figure 7; density po-rosity is calculated from bulk density log and grain densities mea-sured in the laboratory. Line corresponds to x = y. The sonic po-rosity is generally higher than the density porosity in Unit II. InUnit III, the data are scattered about x = y.

higher sonic porosity is consistent with laboratory po-rosity measurements in this interval.

Compressional velocity (Vp) and bulk density (pö) pro-files at Site 612 are plotted in Figure 9. Plots of labora-tory results (points), log data (light lines), and modelvalues (heavy lines) are superimposed. In general, labo-ratory measurements are lower than the log values, be-cause of expansion and disturbance of the laboratorysamples. The most severe discrepancy is in Unit II. Themodel values are interval averages of the logs calculatedto best match the raw data by varying the averaging in-terval between 2 and 10 m. Averages were calculated tobest match the laboratory data in the unlogged (cased)intervals of each well.

Figure 10 shows the compressional velocity and bulkdensity profiles at Site 613. No density log was obtainedat Site 613. Instead, the compressional-wave slownesslog was used to estimate porosity by the empirical rela-tionship (Eq. 1) determined from the Site 612 data. Site613 bulk density was then calculated using this estimatedporosity and Site 613 laboratory grain-density measure-ments. The resulting bulk density log and the laborato-ry measurements at Site 613 are consistent with the dis-parity between laboratory and log bulk-density measure-ments at Site 612.

SEISMIC MODELING

To correlate surface seismic reflections with depth,a synthetic vertical seismic profile (VSP) was generatedfrom the modeled velocity and density profiles and aseismic source at each site. The source used was the im-

593

D. GOLDBERG, R. H. WILKENS, D. MOOS

Compressional velocity (km/s) Bulk density (g/cm )1.5 1.7 1.9 2.1 2.3 2.5 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1

100

200-

300

400-

500

600

700

Vp logVp modelVp lab (v, h)

100

200

300

400

500

600

700

Figure 9. Overlay of shipboard, log, and interval-averaged compressional velocity(Vp) and bulk density (pb) for Site 612.

594

SEISMIC MODELING OF DIAGENETIC EFFECTS, SITES 612 AND 613

Compressional velocity (km/s)1.5 1.7 1.9 2.1 2.3 2.5

Bulk density (g/cm )1.5 1.6 1.7 1.8 1.9 2.0 2.1

100-

200

300-

400-

500-

600 J

Vplog— Vp model" Vp lab (v)

100-

200-

300-

400-

500

600'

Pb log— Pb model

Figure 10. Overlay of shipboard, log, and interval-averaged compressional velocity (Vp)and bulk density (pb) for Site 613.

595

D. GOLDBERG, R. H. WILKENS, D. MOOS

pulse response of the DFS IV recording system filtershown in Figure 11. The recording filter is assumed todominate any bubble reverberations or other aberrationsin the air-gun signal and to be uniform at each shot-point. The low-cut filter setting was 8 Hz with cutoff at3 dB down on an 18-dB/octave slope. The high-cut anti-alias setting was 62 Hz with cutoff at 70 dB down atone-half the sampling frequency (125 Hz) on a -72-dB/octave slope. These frequency limitations are showngraphically by the filter amplitude spectrum in Figure 12(Texas Instruments, 1970). The filter has a nearly linearphase response within the usable bandwidth between 8and 62 Hz.

A synthetic VSP algorithm (Ganley, 1981; Moos, 1984)was used to calculate the seismic signal in the frequency

Time (ms)

50 100 150

Figure 11. Impulse response of the seismic amplifier system used byGeophysical Services, Inc., in recording USGS lines 25 and 35.This is assumed to be the source signature for the seismic modelingin this study. Discrepancy of the actual source from this filter isdue to nonuniform air-gun firings or bubble reverberations.

300

- 200

- 100

- -100

- -200

- -300

- 6 02 4 6 8 10 20 45 60 80100

Frequency (Hz)

-400200

Figure 12. Amplitude and phase spectra of the source signature shownin Figure 11. Low-cut filter setting is at 8 Hz with cutoff at 3 dBdown on an 18-dB/oct slope. High-cut alias filter is at 62 Hz and70 dB down on an 72-dB/oct slope at one-half the sampling fre-quency (125 Hz).

domain. Computation time and numerical accuracy ofthe synthetic is directly related to the number of depthintervals in the model. Models of less than 100 layers, ofvariable thickness, were chosen.

A reflection-coefficient series was first calculated ateach frequency, using the velocity and density values fromthe physical property models. Then the synthetic ampli-tude of the wave field was calculated at each interface,including the contributions from internal reflections. Themethod is one-dimensional and does not account for theeffects of nonvertical incidence, geometrical spreading,or mode conversions. Because the synthetic is generatedin the frequency domain, however, the effects of attenu-ation can be included. There is evidence for changes insonic attenuation with depth (Goldberg et al., 1985). Butthe uncertainty in the frequency dependence of intrinsicattenuation precludes the extrapolation of these changesto seismic data. An intermediate and constant value ofattenuation (Q = 200) was chosen for the synthetic overthe entire depth interval.

The depth resolution of the synthetic is limited by thewavelength of the seismic source (about 100 m). How-ever, the amplitude and phase of seismic reflections aresignificantly affected by variation in structure on a finerscale (Widess, 1973; Schoenberger and Levin, 1976). Inan idealized case, the total seismic response can be de-scribed by the sum of reflections from successive bound-aries, expressed as a series of increasing-order time de-rivatives of the incident pulse (Widess, 1973; Stephensand Sheng, 1985). Therefore, seismic reflections may varyowing to the number or thickness of thin layers across amultiple-layer boundary.

RESULTS

Synthetic VSPs are shown in Figures 13 and 14 forSites 612 and 613, respectively. Seismograms were calcu-lated at depth intervals corresponding to the boundariesbetween lithostratigraphic units. By including these traces,several phases of the surface trace can be extrapolatedback to their origin depths. Although reflection coeffi-cients at each depth are directly related to the geologicmodel, a simple correlation is often complicated by in-ternally reflected energy between intermediate bounda-ries. The depth to each seismic reflector can be deter-mined by comparison of the surface synthetic with thenearby seismic section in Figure 15 or 16. Hence, at eachsite, seismic phases can be attributed to particular chan-ges in physical properties in the sediments.

At Site 612, synthetic reflections match most phasesin the seismic section in both relative amplitude and ar-rival time. At Site 613, the match is poorer. For the mostpart, discrepancies can be attributed to variation in thebathymetry and sediment thickness between the sites andthe lines. The general upslope thinning of the sedimentwedge can change the traveltime to a particular reflectorby more than 0.1s over a lateral offset of only a kilome-ter. Also, ancient and active erosion has created local-ized variations in sediment thickness. Because of theseeffects, the unfortunately large lateral offset of Site 613from line 35 generates significant differences between syn-

596

SEISMIC MODELING OF DIAGENETIC EFFECTS, SITES 612 AND 613

200

Time (ms)

400 600 800 1000

£ 300Q.<D

600

700-

Figure 13. Synthetic vertical seismic profile (VSP) at Site 612. Aver-aged bulk density and compressional velocity logs with a constantvalue of seismic attenuation (Q = 200) were used. Surface seismo-gram phases correspond to earlier phases in subsurface seismo-grams, owing to shorter travel paths from source to receiver. Hence,surface reflections can be traced to their depths of origin.

200

Time (ms)

400 600 800 1000

600

Figure 14. Synthetic VSP at Site 613. Estimated bulk density and aver-aged compressional velocity logs with a constant value of seismicattenuation (Q = 200) were used.

1.6 η

1.8-

2.0-

*2 2.2 HCD

ε"cüCO

fz 2 .4-

2.6-

2.8-

r θ

-100

-200

-300 o.

Eo

-400 £

-500

-600

L 7 0 0

Figure 15. Correlation of surface synthetic seismograms with depth atSite 612. Model seismogram is correlated with depth using the syn-thetic VSP at lithostratigraphic unit boundaries observed in thecore. Approximate time correlation is shown next to the seismicsection from line 25. In general, reflectors at Site 612 are well cor-related with lithostratigraphic boundaries observed in the core.

thetic reflections and the seismic section. An approxi-mate alignment of the synthetic with the section in Fig-ure 16 is tied by the large impedance contrast at the UnitI/Unit II boundary (Eocene/Miocene) and the high-am-plitude reflection (Merlin/Au) at 3.53 s total traveltime(G. Mountain, personal communication, 1985). The mis-match of the synthetic with other events is then mini-mized.

The lithostratigraphic boundary between Units II andIII was identified by the occurrence of porcellanite inthe sediments (Site 612 and 613 chapters, this volume).Porcellanite results from the dissolution of siliceous mi-crofossils and their subsequent reprecipitation as inter-granular cement in the form of lepispheres (Wise andWeaver, 1974; Wilkens et al., this volume). The physicalproperties across this diagenetic boundary change be-tween Sites 612 and 613. At Site 612, the boundary wasmodeled as a discrete high-velocity and high-density lay-er at the top of the porcellanite (324 m sub-bottom).One explanation of this pronounced layer is that cemen-tation was intensified as a result of high pore-water sa-linity (Site 612 and 613 chapters, this volume). A high-amplitude synthetic reflection was generated. At Site 613,a gradational boundary (442 m sub-bottom) was mod-eled by incremental increases in velocity and density thatgenerate a moderate-amplitude synthetic reflection. Inaddition, the phase of the synthetic reflection is differ-ent at each site. This observation results in part from theseismic response of a single-layer boundary at Site 612,

597

D. GOLDBERG, R. H. WILKENS, D. MOOS

3.0-

3.2 H

3.4-

3. 3.6-0)

E

"ö>

I 3.8 H

4.0-

4.2-

4.4-

- 0

MOO

-200

-300

-400

-500

-600

Figure 16. Correlation of surface synthetic seismograms with depth atSite 613. Model seismogram is correlated with depth using the syn-thetic VSP at lithostratigraphic unit boundaries observed in thecore. Approximate time correlation is shown next to the seismicsection from line 35. In general, reflectors at Site 613 are later(deeper) than boundaries observed in the core. This is attributed todipping reflectors, offset from the line, variable sediment thick-ness, and uncertainties in the upper portion of the geologic model.

as against a multiple-layer boundary at Site 613. The ac-tual boundary at both sites is oversimplified by the mod-els, however, and the seismic differences may be due toeven finer-scale variations.

CONCLUSIONS

The difference between in situ and laboratory mea-surements of bulk density is more strongly dependenton lithology than on depth of burial. In lieu of a densitylog, a reliable estimate of in situ bulk density was calcu-lated for Site 613 by using a linear regression of com-pressional slowness and porosity data at Site 612. Syn-thetic VSP seismograms were calculated from intervalaverages of the logs at both sites, which made possibleaccurate determination of the origin depths of surfaceseismogram phases. Owing to variations in bathymetryand sediment thickness, the correlation is poor betweensynthetic seismograms at Site 613 and seismic line 35.

The contrast in physical properties at the diageneticboundary between Units II and III is sharp at Site 612and gradational at Site 613; this difference generates am-plitude and phase differences between seismic reflections(Reflector Ac) at the two sites. From this observation,

changes in physical properties over lateral distances maygenerate seismic reflections which cannot be simply re-lated to lithologic changes or depositional times.

ACKNOWLEDGMENTS

David Goldberg and Dan Moos were supported in part during thiswork by the National Science Foundation and the Joint Oceanograph-ic Institutions under contract JOI 66-84. Roy Wilkens was supportedin part by the National Science Foundation under grant 8408761-OCE.

REFERENCES

Ganley, D. C , 1981. A method for calculating synthetic seismogramswhich include the effects of absorption and dispersion. Geophys-ics, 46:1100-1107.

Goldberg, D. S., Moos, D., and Anderson, R. N., 1985. Attenuationchanges due to diagenesis in marine sediments. Trans. 26th Ann.SPWLA Symp., paper KK.

Hamilton, E. L., 1956. Low sound velocities in high-porosity sedi-ments. J. Acoust. Soc. Am., 28:16-19.

, 1976. Variations of density and porosity with depth in deep-sea sediments. J. Sediment. Petrol., 46:280-300.

Hamilton, E. L., and Bachman, R. T., 1982. Sound velocity and re-lated properties of marine sediments. J. Acoust. Soc. Am., 72(6):1891-1904.

Mayer, L. A., 1980. Deep-sea carbonates: Physical property relation-ships and the origin of high frequency acoustic reflectors. Mar.Geol., 38:165-183.

Moos, D., 1984. A case study of vertical seismic profiling in fracturedcrystalline rock. In Simaan, M. (Ed.), Advances in GeophysicalData Processing (Vol. 1): Greenwich, CT (JAI Press), 9-37.

Officer, C. B., 1958. Introduction to the Theory of Sound Transmis-sion: New York (McGraw-Hill), p. 257.

Riech, V, and von Rad, U , 1979. Eocene porcellanites and Early Cre-taceous cherts from the western North Atlantic basin. In Tucholke,B. E., Vogt, P. R., et al., Init. Repts. DSDP, 43: Washington (U.S.Govt. Printing Office), 437-448.

Robb, J. M., Hampson, J. C , Kirby, J. R., and Twitchell, D. C ,1981. Geology and potential hazards of the continental slope be-tween Lindenkohl and South Toms canyons, offshore mid-AtlanticUnited States. U.S. Geol. Surv. Open File Rept., 81-600:1-38.

Schlee, J. S., 1981. Seismic stratigraphy of the Baltimore CanyonTrough. Am. Assoc. Pet. Geol. Bull., 65:26-53.

Schlumberger, Ltd., 1972. The formation density log. SchlumbergerLog Interpretation (Vol. 1): Principles, 43-48.

Schoenberger, M., and Levin, R, 1976. Reflected and transmitted fil-ter functions for simple subsurface geometries. Geophysics, 41:1305-1317.

Shipley, T. H., 1983. Physical properties, synthetic seismograms, andseismic reflection correlations at DSDP Site 534, Blake-BahamaBasin. In Sheridan, R. E., Gradstein, F. M., et al., Init. Repts.DSDP, 76: Washington (U.S. Govt. Printing Office), 653-665.

Shumway, G., 1960. Sound speed and absorption studies of marinesediments by a resonance method. Geophysics, 25:451-467, 639-682.

Stephens, R. B., and Sheng, P., 1985. Acoustic reflections from com-plex strata. Geophysics, 50:1100-1107.

Texas Instruments, 1970. Record filter response. DFS PerformanceManual, pp. 1-17.

Tucholke, B. E., 1979. Relationships between acoustic stratigraphy andlithostratigraphy in the western North Atlantic basin. In Tucholke,B. E., Vogt, P. R., et al., Init. Repts. DSDP, 43: Washington (U.S.Govt. Printing Office), 827-846.

Tucholke, B. E., and Mountain, G. S., 1979. Seismic stratigraphy, litho-stratigraphy, and paleosedimentation patterns in the western NorthAtlantic Basin. In Talwani, M., Hay, W W , and Ryan, W. B. F.(Eds.), Deep Drilling Results in the Atlantic Ocean: ContinentalMargins and Paleo-environment: Washington (Am. Geophys. Un-ion), Maurice Ewing Series, 3:58-86.

Weaver, F. M., and Wise, S. W , 1974. Opaline sediments of the SEcoastal plain and Horizon A: Biogenic origin. Science, 184:899-901.

598

SEISMIC MODELING OF DIAGENETIC EFFECTS, SITES 612 AND 613

Widess, M. B., 1974. How thin is a thin bed? Geophysics, 38:1176- Wyllie, M. R. J., Gregory, A. R., and Gardner, G. H. E , 1958. An ex-1180. perimental investigation of factors affecting elastic wave velocities

Wise, S. W., and Weaver, F. M., 1974. Chertification of oceanic sedi- in porous media. Geophysics, 23:459-493.ments. In Hsü, K. J., and Jenkyns, H. C. (Eds.), Pelagic Sedi-ments: On Land and Under the Sea (Vol. 1): London (BlackwellSci. Publ.), 301-326. Date of Initial Receipt: 4 March 1985

Wood, A. B., 1930. A Textbook of Sound: London (G. Bell and Sons). Date of Acceptance: 16 September 1985

599