A bottom-simulating reflection (BSR) is a seismic reflec-tivity phenomenon that is widely accepted as indicating thebase of the gas-hydrate stability zone. The acoustic imped-ance difference between sediments invaded with gas hydrateabove the BSR and sediments without gas hydrate, but com-monly with free gas below, are accepted as the conditionsthat create this reflection. The relationship between BSRs andmarine gas hydrate has become so well known since the1970s that investigators, when asked to define the mostimportant seismic attribute of marine gas-hydrate systems,usually reply, “a BSR event.” Research conducted over thelast decade has focused on calibrating seafloor seismic reflec-tivity across the geology of the northern Gulf of Mexico(GoM) continental slope surface to the seafloor. This researchindicates that the presence and character of seafloor brightspots (SBS) can be indicators of gas hydrates in surface andnear-surface sediments (Figure 1). It has become apparentthat SBSs on the continental slope generally are responsesto fluid and gas expulsion processes. Gas-hydrate forma-tion is, in turn, related to these processes. As gas-hydrateresearch expands around the world, it will be interesting tofind if SBS behavior in other deepwater settings is as use-ful for identifying gas-hydrate sites as in the GoM.

Research background. Joint research with the MineralsManagement Service (MMS), through a cooperativeagreement with the Coastal Marine Institute atLouisiana State University, has resultedin a study using the university’s GoM-wide 3D seismic data coverage. Seafloorreflectivity across the northern GoM’scontinental slope has been mapped, andnumerous SBS areas like those illustratedin Figure 1 have been identified. Throughresearch projects funded primarily byMMS and NOAA, manned submersibledives on many SBS sites have provideddirect observations and samplings to cal-ibrate the actual geologic character of theseafloor to seismic reflectivity. Theseinvestigations have led to the model illus-trated in Figure 2 which defines threequalitative ranges in the rates of fluid andgas expulsion coupled to geologic-biologicresponse at the seafloor. These reflector-anomaly sitesalso indicate three gas-hydrate domains that can be asso-ciated with these seafloor reflectivity behaviors.

The block diagram shown in Figure 2 illustrates that flu-ids and gases have various migration pathways to the mod-ern seafloor and that delivery rates, as well as fluid-gascomposition, impact geologic response on the continentalslope surface. Deep-cutting faults that intersect the over-pressured zone frequently form the migration routes forhydrocarbons, formation fluids, and sometimes fluidizedsediments. In addition, considerable heat can accompanythese products, which eliminates the gas-hydrate stability

Seafloor reflectivity—An important seismic property for interpreting fluid/gas expulsion geology and the presence of gas hydrateHARRY H. ROBERTS, Louisiana State University, Baton Rouge, USABOB A. HARDAGE, Bureau of Economic Geology, Austin, USAWILLIAM W. SHEDD and JESSE HUNT JR., Minerals Management Service, New Orleans, USA

620 THE LEADING EDGE MAY 2006

Figure 1. Computer-enhanced multibeam bathymetry image of the north-ern Gulf of Mexico continental slope showing a small sector of theMississippi Canyon lease area and its seafloor bright spots (SBS) or localareas of high seismic seafloor reflectivity across the continental slopesurface.

Figure 2. Thisblock diagram schemati-

cally illustrates the threedeepwater types of seafloor

responses to rapid-to-slow fluid andgas expulsion. Rapid-flux systems gener-

ally transport heat as well as fluidized sedimentto the seafloor, creating mudflows and mud volca-

noes. Heat eliminates stable conditions for gas-hydrateformation, but gas hydrates can form in mud vents that are inactive or inflank sediments. Slow seepage promotes seafloor lithification processesmediated by microbial utilization of hydrocarbons. Authigenic carbonatesare the most common lithification products. Gas hydrate may occur in thesubsurface, but never at the seafloor. Between rapid venting and slowseepage, conditions are compatible with creating and sustaining gashydrate at the modern seafloor. These areas exhibit highly variable seafloorfeatures, from productive chemosynthetic communities to localized areasof lithification (Modified from Roberts, 2001).

zone or dramatically shifts it toward the seafloor in the migra-tion pathways. Rapid flux or venting of these products in suf-ficient volumes can create mud flows or mud volcanoes.Because venting of hydrocarbon-laced mud is a sporadicprocess, mud flows and the flanks of mud volcanoes may par-tially cement if the interval between expulsion events is long.

In contrast, short migration routes are common to slow-flux, or seepage, of hydrocarbon to the continental slope sur-face, and cementation of surficial sediments is widespreadacross these systems. The process of cementation is associ-ated with microbial communities that utilize the hydro-carbons and then produce authigenic carbonates asby-products. The most common carbonate mineral is mag-nesium-calcite even though aragonite and dolomite alsooccur. These carbonate minerals share a common character-istic. They all are 13C-depleted, indicating that 12C, microbiallyseparated from the hydrocarbons, is incorporated into thecalcium-magnesium carbonate molecules. For the slow-seep-age case illustrated in Figure 2, migration pathways are asso-ciated with a shallow salt mass.

Between the extremes of rapid-flux and slow-flux set-tings, conditions are optimal for creating gas hydrates in thevery shallow subsurface, with some hydrate being exposedas localized outcrops. At these sites, gas is supplied at sucha rate that seafloor exposures of gas hydrate are maintained

and rebuilt after hydrate instability and dissociation relatedto thermal loading by the water column, e.g., the LoopCurrent. These intermediate flux areas generally display ahighly variable seafloor character, including exposed gashydrates, localized outcrops of authigenic carbonate, anddiverse and densely populated communities of chemosyn-thetic organisms. It has been suggested that exposed andshallow-subsurface gas-hydrate deposits in these settingsprovide the trophic resources to sustain chemosyntheticcommunities that live primarily on microbially mediatedhydrocarbon gas and hydrogen sulfide. Slow-flux and rapid-flux environments do not support well-populated anddiverse chemosynthetic communities.

Before illustrating seismic examples of seafloor reflec-tivity across these three fluid gas expulsion rate categories,we first show photographs in Figure 3 of seafloor sites rep-resenting each type of expulsion rate and gas-hydrate set-ting. These direct observations illustrate actual seafloorconditions associated with each of the three major types offluid-gas expulsion environments, rapid-to-slow flux, andthe associated settings for three different occurrences of gashydrate at the surface or in near-surface sediments.

Gas-hydrate domain 1—rapid fluid-gas expulsion. It haslong been recognized that during rapid venting of fluids and

MAY 2006 THE LEADING EDGE 621

Figure 3. This figure summarizes the response at the seafloor of venting-to-seepage rates of hydrocarbons, formation fluids, and fluidized sediment(modified from Roberts, 2001). Rapid-flux systems are generally accompanied by fluidized sediment resulting in mudflows and mud volcanoes of vari-ous dimensions. Sedimentation rates are typically too high to support complex communities of benthic organisms. Bacterial mats (Beggiatoa) andlucinid-vesycomyid clams occur on the surfaces of recently deposited sediments laced with hydrocarbons. Slow-flux systems are characterized by hard-grounds and mound-like buildups of authigenic carbonates. Intermediate-flux areas support large and densely populated communities of chemosyn-thetic mussels and tube worms. Localized authigenic carbonate hardgrounds and bacterial mats are also common to this setting. Each picture has a fieldof view of 2–4 m across.

gas, sediment can be entrained in a slurry-like state andextruded at the seafloor. Mud volcanoes and mud flowshave been observed onshore and offshore from many geo-logic settings ranging from passive margins with salt andshale tectonics like the Gulf of Mexico to accretionary prismsassociated with subduction zones. Although quantitativedata on the frequency and volume flux of fluidized sedimentin rapid flux settings is sparse, it is clear that the expulsionprocess is episodic. On 3D seismic seafloor reflectivity mapsof active vents with associated “fresh” mudflows, reflectiv-ity is low because of the “soft,” recently extruded sedimentthat frequently contains bubble-phase gas. The vent itself usu-ally exhibits the lowest reflectivity and often is accompaniedby a phase reversal of the seafloor reflection when comparedto surrounding areas. In profile view, the vent is the surfaceexpression of a vertical “gas chimney” characterized by

acoustic transparency in the subsurface.Figure 4 illustrates a 3D seismic seafloor reflectivity map

and a seismic profile acquired across Green Canyon Block204 (GC204). The reflectivity map illustrates several expul-sion centers of relatively low amplitude and long mudflowsof much higher amplitude. Submersible observations indi-cate that the venting sites (relatively low amplitude) are stillminimally active, but not forcing out large volumes of flu-idized sediment as has happened in the past. Well-mixed sed-iment along the expulsion pathway, as well as the presenceof gas, gives the expulsion site a low reflectivity at the seafloorand an acoustic transparency in the subsurface (Figure 4).The GC204 site represents a rapid-seafloor flux system thatis in the process of becoming dormant. The strong reflectiv-ity of the flows is related to the presence of nodular, car-bonate-cement crusts in the muds and seafloor clam shells.

622 THE LEADING EDGE MAY 2006

Figure 4. (a) Seafloorreflectivity map oftwo expulsion sitesin GC204. Note therelatively low reflec-tivity of the ventingcenters and the“bright” and highlyreflective flowsextending downslopefrom these centers.The highly reflectiveflows are related toauthigenic carbon-ates and clam shells.Gas in sediments ofthe venting sitesaccounts for theirlow reflectivity. (b) Seismic profilethrough the twoexpulsion sites illus-trated on the seafloorreflectivity map as aNNW-SSE red line.Note the shallowsubsurface, high-amplitude events ineach migration path-way (yellow). Also,note that these eventsare connected to aninterpreted BSR. Thehigh-amplitudeevents are interpretedas gas and perhapsgas hydrate in thethroats of the twovents profiled.

Both the diagenetic products and shell beds indicate that theflows are not new. As discussed above, the carbonates ariseas by-products from microbial oxidation of hydrocarbons inthe extruded muds. The lucinid-vesycomyid clams exploithydrogen sulfide produced as microbes metabolize hydro-carbons incorporated in the mudflows. Once this trophicresource is depleted, the clams die, leaving a surface “pave-ment” of clam shells like those illustrated in Figure 3 (lower-

left panel). So, the high surface reflectivity is related to boththe clam shells and diagenetic carbonate products.

The seismic profile of Figure 4 cuts through the two promi-nent expulsion centers shown in the accompanying seafloorreflectivity map. Note the high-amplitude events near theseafloor. Beneath these events is the acoustically transparentmigration pathway that is the vertical transport route for flu-ids, gases, and heat. These shallow, high-amplitude events

624 THE LEADING EDGE MAY 2006

Figure 5. (a) A 3D seis-mic seafloor reflectivitymap across Block GC232illustrating the variableseafloor reflectivity pat-terns typical of sites whereseafloor exposures of gashydrate are frequentlyfound. This site is highlyfaulted and was the firstsite where a surface expo-sure of gas hydrate wasrecognized. The highlyreflective area in GC233represents carbonatehardgrounds, while gashydrates are found in theextremely variable, butgenerally low-amplitudezones that display waveletinterference effects acrosslow-impedance gas expul-sion sites. (b) Seismicprofile across the gas-hydrate site identified bythe red line on the reflec-tivity map. A strongwavelet interferenceoccurs in the seafloorreflection at the left side ofthe seismic profile. Thisarea is invaded with mas-sive vein-filling and nodu-lar gas hydrates.

in the subsurface of the once highly active venting sites areconnected to reflections that slope away into the subsurfacecutting across stratigraphic horizons. These reflections areinterpreted as BSRs, and the high-amplitude event near theseafloor is interpreted as gas hydrate. Increased heat flow andperhaps high-salinity fluids are responsible for elevating thebase of the hydrate stability zone (BHSZ) to near the mod-ern seafloor. During active venting, gas hydrate would prob-ably not be present in the migration pathways. Because thissite is now inactive, gas hydrate has formed at the vent site,if our interpretations of the seismic data are correct. If thesevents remain inactive for long periods of time or they are aban-doned entirely, gas hydrate will likely form along the verti-cal migration pathway. Gas hydrate will not generally befound in the flanking surfacial flow deposits, but the BSR iden-tified in the seismic profile of Figure 4 suggests hydratesshould be found in the subsurface.

Gas-hydrate domain 2—moderate fluid–gas expulsion.The intermediate-flux case, as illustrated in Figures 2 and3, represents a set of conditions that promote gas-hydrateformation at or very close to the modern seafloor. In con-trast to the rapid-flux case, where venting of fluidized sed-iment and heat is not conducive to gas-hydrate formationor to the growth and development of complex chemosyn-thetic communities, slope-depth intermediate-flux settingspromote both. Gas is continually supplied to these sites, asdemonstrated by bubble plumes detected in the water col-umn on echo sounder records and other high-resolutionacoustic data sets. Mounds and outcrops of gas hydrate aregenerally observed at these sites. The association betweenshallow-subsurface or exposed gas hydrates and the pres-ence of diverse and densely populated chemosynthetic com-munities is not a chance happening. For such communities todevelop and persist, they must have a constant supply of hy-

drocarbon gas and hydrogensulfide.

Gas hydrate, which is stableat temperatures and pressuresbelow water depths of 400–500m in the GoM, functions as areservoir that can produce andstore the necessary products tosustain these communities. Ifgas supply from the subsurfacevaries with time, gas hydrateprobably functions as a low-pass process filter to stabilizetrophic resources to maintainchemosynthetic life. Anaerobicmicrobial utilization of hydro-carbons in these environmentsproduces bicarbonate, which isresponsible for triggering theprecipitation of authigenic car-bonates in the form of nodularmasses in the sediment, crusts,and mound-like buildups. It isalso common that in areaswhere gas hydrates are exposedat the surface that small mudvents extruding fluidized mudand hypersaline fluids are pre-sent. Expulsion of salt in thehydrate-forming process isinvolved in the creation of thesefeatures. So, intermediate-flux

settings are highly complex in terms of their seafloor geologyand biology. This complexity is reflected in both the patternsand strengths of seafloor reflectivity values derived from 3Dseismic data.

Figure 5 illustrates the seafloor reflectivity response in anarea of Green Canyon Block 232 (GC232) where the first expo-sure of gas hydrate on the seafloor was recognized while con-ducting manned submersible research in 1991. Gas hydratehas been discovered in cores from many sites across the upperslope since the early 1980s. After the discovery of outcroppinggas hydrate in GC232, many similar outcrops have beenobserved at SBS sites across the northern Gulf of Mexico’s con-tinental slope. These exposures of gas hydrate usually occurin localized mounds a few meters in diameter and in somesettings they align themselves with well defined faults. Figure3 (center panel) illustrates one of these mounds that is cov-ered with an orange bacterial mat (Beggiatoa). The gas hydratein this mound occurred as white layers exposed at the edgeof the mound. Beggiatoa mats act as semipermeable mem-branes that collect migrating hydrocarbons beneath. Anaerobicmicrobial utilization of hydrocarbon depletes oxygen andencourages sulfate reduction to produce hydrogen sulfideneeded for Beggiatoa growth. Therefore, because of this andother microbial interactions, hydrocarbons in moderate fluxsettings, like GC232, demonstrate the characteristics associ-ated with biodegradation.

The wide range of seafloor reflectivities and their com-plex plan-view patterns observed in Figure 5a are typical ofSBS sites that support gas hydrate at or very near the mod-ern seafloor. Phase reversals of the surface reflection (Figure5b) frequently occur where gas hydrates and associatedvent sites are found at the surface. Highly populated com-munities of chemosynthetic mussels and tube worms accom-panied by authigenic carbonate hardgrounds are scatteredthroughout these areas.

MAY 2006 THE LEADING EDGE 625

Gas-hydrate domain 3—slow fluid–gas expulsion. At theslow-seepage end of the fluid-gas expulsion spectrum,hydrocarbons and other fluids do not arrive at the seafloorin sufficient quantities to support chemosynthetic commu-nities of macroorganisms or to build topography throughexpulsion of large volumes of fluidized sediment. However,microbial communities in both reduced and oxidized, sur-face to shallow-subsurface environments utilize seep-deliv-ered hydrocarbons and produce conditions conducive toauthigenic carbonate precipitation. Hydrocarbon oxidationin aerobic environments produce CO2 and decrease pH,

which in sufficient quantities can favor dissolution of car-bonates rather than precipitation. However, anaerobic micro-bial sulfate reduction involving hydrocarbon substratesresults in sulfide depletion and both bicarbonate and hydro-gen sulfide enrichments. The increase in carbonate alkalin-ity of pore fluids resulting from these microbial reactionsproduces calcium-magnesium carbonate by-products. Thesecarbonates can take the forms of small nodular masses insediments, hardground slabs or pavements, and mound-likebuildups (Figure 3, right panel). Areas of seafloor wherethese processes are prevalent and widespread display high

626 THE LEADING EDGE MAY 2006

Figure 6. (a) A 3D seismicseafloor reflectivity map ofthe dome top in GC140.The extremely high reflec-tivity is related to the pres-ence of authigeniccarbonate mounds andhardgrounds scattered overthe top of this feature. Theinset is a side-scan sonarimage from the dome topillustrating the spatialarrangement and shapes ofthe carbonate mounds. (b)This NW-SE seismic profileacross the GC140 domeillustrates the highly reflec-tive surface of the dome,subsurface structure, and“Bush Hill,” a moundinvaded with gas hydrateon the flank of the GC140dome.

reflectivity on seismic records.Figure 6 illustrates the 3D seismic seafloor reflectivity

response and profile character of a site, Green Canyon Block140 (GC140), that is characterized by slow hydrocarbonseepage and an abundance of resulting authigenic carbon-ate mounds, plus seafloor hardgrounds. Analysis of both 3Dand higher-frequency seismic data as well as numerousmanned submersible dives at this site revealed no obviousgas plumes in the water column or sites of fluid-gas expul-sion on the seafloor. Only one localized site of a fewmacrochemosynthetic organisms was observed in a fissurethrough one of the authigenic carbonate mounds. However,bacterial mats on the seafloor and in fissures and aroundpores in the authigenic carbonate mounds were frequentlyobserved. This pattern of bacterial mat occurrence with thenear total absence of macrochemosynthetic organisms andan abundance of authigenic carbonates suggests long-term,slow seepage of hydrocarbons. The carbonates are primar-ily magnesium-calcite and dolomite. Both mineral phasesare 13C-depleted with δ13C values as negative as -55 permille with respect to the PDB standard belemnite.

The 3D seismic data reveal a high reflectivity of the GC140mound surface. The positive polarity and strength of the sur-face reflection suggests a hard bottom which certainly is thecase. The absence of areas of negative polarity in the surfacereflection, lack of obvious gas plumes in the water column,and the lack of macrochemosynthetic organisms collectivelysuggest that gas-hydrate deposits are probably not presentin the very shallow subsurface as is the case in intermediate-delivery systems. In slow-seepage systems, the sulfate reduc-ing zone which limits gas-hydrate occurrence can be quitethick compared to that of moderate-to-rapid-flux settings.

Conclusions. Nearly two decades of research have been con-ducted across the northern Gulf of Mexico’s continental slopeusing seismic data, coring, and manned submersible divesto study the impacts of fluid-gas expulsion on the geologyand biology of the modern seafloor. These investigationshave been made primarily in water depths of 1000 m or shal-lower. They clearly establish that areas of seafloor reflectiv-ity (seafloor bright spots, or SBS), easily identified on 3Dseismic data, are generally areas impacted by hydrocarbonventing-to-seepage. A few exceptions, such as recentlydeposited sand-rich slope fans, do occur. At the shelf edge,biogenic carbonate veneers on shelf edge knolls are highlyreflective, but by far the greatest number of SBS sites in thezone of gas-hydrate stability across the middle-to-upper slopeare related to the fluid-gas expulsion process. Seafloorresponse to the expulsion process is qualitatively related todelivery rate of hydrocarbons, formation fluids, and some-times fluidized sediment. Patterns of seafloor reflectivity areclues to the delivery rate, seafloor responses, and ways gashydrates may be associated with various sites representingdifferent fluid-gas expulsion rates. Table 1 summarizes thegeneral relationships between gas-hydrate occurrence andfluid-gas expulsion domain. The authors acknowledge thehigh degree of variability in seafloor reflectivity associatedwith seep and vent sites. Relationships between reflectivityand gas hydrate are offered as general cases based on bothseismic data and direct seafloor observations.

Rapid-delivery expulsion systems generally buildseafloor topography such as mud volcanoes and displaymud flows that transport fluidized sediment down slopefrom an expulsion site or vent. Seafloor reflectivity patternsclearly distinguish these features as shown in Figure 4. Highseafloor reflectivity can be caused by precipitation of authi-

genic carbonates in the hydrocarbon-laced sediment flowsand clam shells (lucinid-vesycomyid clams) left after exploit-ing the limited supply of hydrogen sulfide microbially gen-erated in these deposits. Where a BSR can confidently beidentified in association with a rapid-delivery system, thisreflector rises to the seafloor, or near it, in the fluid-gasmigration pathway, as shown in Figure 4. Heat and perhapshigh-salinity fluids transported during expulsion eventschange the boundary conditions for gas-hydrate stability andforce the boundary toward the seafloor as illustrated in thisGC204 example. Therefore, if gas hydrate occurs in themigration pathway of a recently waning system like GC204,it is likely to be a very thin shallow deposit. In active vent-ing situations, it is probable that no gas hydrate will occurin association with the vent. As the interpreted BSR of Figure4 suggests, if gas-hydrate deposits are to be found in rapid-delivery settings, they occur lateral to the vent sites and theirsubsurface migration pathways.

Moderate delivery rates of hydrocarbons to the modernseafloor provide the optimal set of conditions for gas-hydrateformation at or near the seafloor. Under these conditions thesulfate reducing zone is very thin or essentially at the seabed.Therefore shallow-to-exposed gas-hydrate deposits can beexpected under these conditions. Seafloor reflectivity, asdetermined from 3D seismic data, is highly variable formost moderate-delivery settings (Figure 5). These sites sup-port diverse and densely populated communities ofchemosynthetic organisms and display scattered areas ofseafloor lithification and may even display small areas offluidized mud venting.

Because pervasive seafloor lithification is a product ofmicrobial oxidation of hydrocarbon in slow-delivery settings,seafloor reflectivity is very high, as displayed on 3D seis-mic seafloor reflectivity maps like the one illustrated inFigure 6. These areas can have relatively thick sulfate reduc-ing zones, limiting gas-hydrate formation near the seafloor.Heat flow from the deep subsurface is low in comparisonto rapid and moderate flux settings. However, if low-deliv-ery-rate settings occur above a shallow salt body, the effi-cient heat transfer from the deep subsurface through the saltmay limit the thickness of the hydrate stability zone, asschematically illustrated in Figure 2.

Suggested reading. Several summaries of gas hydrate andhydrocarbon seep research have been published recently.“Natural Gas Hydrates: Occurrence, Distribution, and De-tection” by Paull and Dillon (American Geophysical Union,Geophysical Monograph 124, 2001). Methane Hydrate in QuaternaryClimate Change by Kennett et al. (American Geophysical Union,2003). Submarine Gas Hydrates by Ginsburg and Soloviev (NormaPublishers, 1998). Clathrate Hydrate of Natural Gases by Sloan(Marcel Dekker, 1998). “Hydrocarbon Migration and Its Near-Surface Expression” by Schumacher and Abrams (AAPGMemoir 66, 1996). Seabed Pockmarks and Seepages: Impacts onGeology, Biology, and the Marine Environment by Hovland andJudd (Graham & Trotman, 1989). TLE

Acknowledgment: The authors thank the Minerals Management Servicefor support of several research projects related to the use of 3D seismicdata for studying impacts of fluid and gas expulsion on the northern Gulfof Mexico continental slope. Manned submersible time has been supportedby NOAA-NURP, MMS, Louisiana Education Quality Support funds,and the Gulf of Mexico Hydrates Research Consortium at the Universityof Mississippi.

Corresponding author: hrober3@lsu.edu

628 THE LEADING EDGE MAY 2006