GIS-Based Assessment of Submarine Mudflow
Hazard Offshore of the Mississippi Delta,
Gulf of Mexico
C. Hitchcock, R. Givler, M. Angell, and J. Hooper
Abstract Documented changes in bathymetry and geologic features have been
incorporated into a GIS-based approach to map hurricane-induced mudflow susceptibility on the seafloor offshore of the Mississippi Delta. Available bathymetric data
were interpreted to delineate areas of relative sea-floor stability over the past century,
areas of active mudflow transport, and areas of active mudflow lobe deposition,
incorporating mapping by Coleman et al. (1982). Our research is designed to provide
regionally-consistent hazard information for the siting and design of pipelines and
structures using readily available datasets. Based on interpretation of available datasets
for the Mississippi Delta region, and our results for a test area using post-Hurricane
Ivan (pre-Hurricane Katrina) bathymetric data, mudflow transport within the Delta
generally occurs within well-defined submarine channels or gullies, spreading out
onto the seafloor in deeper water to form over-lapping lobes of thick, viscous silty
clay. Semi-stable areas between the mudflow channels locally provide the least
hazardous locations for siting of future production facilities and pipeline routing.
The zone of overlapping mudflow lobes located downslope of the gullies is an area
of recent deposition vulnerable to mudflow overruns from upslope mudflows.
Keywords Submarine landslide • mudflow • hurricane • geohazard • seafloor
geomorphology • GIS • susceptibility
1
Introduction
High-rate accumulation of low shear strength, low permeability sediment
within the shallow Gulf of Mexico at the mouth of the Mississippi River Delta
coincides with an offshore area of significant oil and gas production (Fig. 1).
C. Hitchcock (), R. Givler, M. Angell, and J. Hooper
Fugro (William Lettis & Associates, Inc.), 1777 Botelho Drive, Suite 262, Walnut Creek,
CA 94598, USA
e-mail: hitch@lettis.com
D.C. Mosher et al. (eds.), Submarine Mass Movements and Their Consequences,
Advances in Natural and Technological Hazards Research, Vol 28,
© Springer Science + Business Media B.V. 2010
353
354
C. Hitchcock et al.
Fig. 1 Regional map showing Mississippi Delta study area
The potential for large mudflows damaging to production facilities and pipelines is
greatest within, and downslope of, these unstable deposits on the delta submarine
apron. Sediment transport generally occurs within well-defined submarine channels
or gullies, spreading out onto the seafloor in deeper water to form overlapping mud
lobes of thick, viscous silty clay (Hooper 1980; Coleman et al. 1982). Wave-induced
bottom pressures accompanying large hurricanes trigger large-scale failures of
these sediments, causing mudflow overruns of deep-water regions (Hooper
and Suhayda 2005). Slopes exhibiting marginal static stability, and areas of past
submarine failure, usually are the most susceptible to hurricane-induced failure.
Oversteepened slopes in recent mudflow lobes are particularly prone to failure.
In this paper, we present a GIS-based approach for mapping mudflow susceptibility using existing datasets. We digitized and analyzed available bathymetric
and geologic map data to de-lineate areas of relative seafloor stability over the past
century, areas of active mudflow transport, and areas of mudflow deposition
(Hitchcock et al. 2006). In the absence of comprehensive borehole and material
properties databases, comparison of bathymetric data over time is combined
with existing seafloor mapping by Coleman et al. (1982) to delineate likely areas
of elevated hazard for erosion and related mudflow failures. In addition, we conducted
new geomorphic mapping of the seafloor to classify the recency of mudslide movement and mudflow lobe deposition.
GIS-Based Assessment of Submarine Mudflow Hazard Offshore
1.1
355
Geologic Setting
The Balize delta, or ‘birdsfoot delta’, is an active depositional delta that began
prograding in the Gulf of Mexico over 1,000 years ago (Saucier 1963). The delta is
supplied with sediment by three major distributary channels (Southwest Pass,
South Pass, and Pass A Loutre). Although partial capture of the Mississippi River
flow by the Atchafalaya River has diminished sediment supply to the Balize delta
within the last several 100 years (Kesel 1988; Roberts 1997), the major river
distributaries have grown seaward at an average rate of 100 to 200 ft/yr over the
past 150 years (Morgan 1977). Deposition rates as high as 1 to 2 ft/yr at the mouths
of the distributary channels result in rapid accumulation of low shear strength, low
permeability sediment (Coleman et al. 1982; Hooper and Suhayda 2005).
The submerged delta apron therefore consists of thick, very weak sediments that
are inherently unstable and vulnerable to hurricane wave-induced failure (Bea et al.
1975; Hooper 1980). Rapid sedimentation can change the slope of the seafloor, causing
oversteepening and loading of the underlying sediments (Lindsay et al. 1984).
Associated increased pore-water pressure and buildup of methane gas increases the
susceptibility of accumulated sediment to failure (Wheland et al. 1978; Roberts 1997).
As a result, wave-induced bottom pressures accompanying large hurricanes can
cause significant failures of the accumulated sediments (Bea et al. 1975; Hooper
1980; Hooper and Suhayda 2005).
Geologic mapping by Coleman et al. (1982) has provided the context for regional
and site-specific studies over the past 27 years. However, with deposition rates of
1 to 2 ft/yr, much has changed in the submarine environment including several major
hurricanes. The impacts of large hurricanes (Ivan, Katrina, Rita, Ike, Gustav) to oil and
gas infrastructure have been significant but the locations of change to the seafloor
as a result of hurricane-induced waves and currents are poorly documented.
1.2
Mudflow Hazard
Mudflows along the submerged Mississippi delta are part of a dynamic system of
sediment transport and deposition (Fig. 2). Mudflow transport generally occurs
within well-defined submarine gullies (Shepard 1955; Hooper 1980; Coleman et al.
1982). Local accumulation of sediment coupled with scour during mudflow transport results in highly variable, and unstable, conditions within the gullies.
Semi-stable areas between the mudflow gullies locally are vulnerable to failure
as the gullies grow in length and width over time, cannibalizing the gully margins.
However, areas between mudflow gullies are believed to be generally stable with
low sediment accumulation rates (Coleman et al.1982). As a result, these areas have
long been targeted for siting of future production facilities and pipelines.
The mudflow gullies supply a complex zone of overlapping mudflow lobes in
deeper water. The mud-flow lobe deposition area, located downslope of the gullies, is
a region of recent to active deposition vulnerable to mudflow overruns from upslope
356
C. Hitchcock et al.
Fig. 2 Interpreted multi-beam bathymetry showing
mudflow gullies (blue), mudflows (orange), and inferred
transport (white arrows). White line is the 1982 gulfward
extent of gullies from Coleman et al. (1982) and yellow
line is interpreted 2005 extent
gullies and mudflows. Localized damage of pipelines and production facilities is
well documented in this zone (Coleman et al. 1982; Hooper and Suhayda 2005).
2
Approach
Geographic information systems (GIS) technology enables compilation and mapping
of susceptibility to failure of the land and seafloor (e.g. Hitchcock et al. 1999).
Using a criteria matrix, we ranked key factors that influence submarine stability by
GIS-Based Assessment of Submarine Mudflow Hazard Offshore
357
assigning point scores for map units (rate of change, slope, and geologic unit) in
each map layer (Fig. 3). The point scores within a series of derivative GIS map
layers are summed in a single interpretative map layer depicting mudflow susceptibility. This approach allows for inclusion of additional data layers and also enables the extrapolation of relatively limited site-specific, material properties data
derived from boreholes and laboratory testing.
The GIS data compilation and evaluation process involved the following steps:
(1) compilation of available bathymetric data and development of a composite
bathymetric base map; (2) development of derivative map layers that depict rates of
change of the seafloor (deposition and erosion) from historic bathymetric maps; (3)
development of a submarine slope gradient map from bathymetric data; (4) digitization of available geologic mapping and tectonic structures from Coleman et al.
(1982) to create a geologic map layer; (5) revision of the geologic map layer to
incorporate independent interpretation of multibeam bathymetric data; (6) development of a mud-flow susceptibility criteria matrix; (7) derivation of a point score
system for integration of map units for each map layer, and (8) summation of associated point scores for each layer to develop a derivative mudflow susceptibility
map. Below we discuss the data and methods used to develop the susceptibility map
and interpret the relative contribution of various criteria for each map layer.
The study area in the Mississippi Delta is unique in that an historic record, supported by detailed bathymetric data, exists of the seafloor over the past 140 years.
Bathymetry in 1874 was compiled and drafted by Coleman et al. (1982) from maps
that span the period 1872 to 1874 and incorporated copies of original soundings.
Coleman et al. (1982) also provide regionally consistent bathymetry collected
between 1977 and 1979.
Bathymetric data provides valuable information used to delineate submarine
landforms, including individual mudflows, and derive seafloor slope. Also, these data
are essential for interpretation of long-term patterns of sediment transport and
deposition. In particular, comparison of bathymetric surveys provides information on
changes in the seafloor bottom over time. Sediment accumulation and erosion rates
derived from comparison of bathymetric surveys allow for interpretation of locations,
amounts, and frequency of mass movement on the seafloor (Coleman et al. 1982).
However, available bathymetric datasets vary in how seafloor measurements were
collected with associated variations in resolution, accuracy, and sources of potential
error that make direct comparison of the seafloor in different time periods difficult.
Geologic mapping of seafloor features provides valuable information on the
processes associated with mudflow failure, transport, and deposition. Important
geologic factors influencing the susceptibility of mud deposits to failure include the
genesis (source), composition, and age of seafloor sediment. For this project, polygons
of map units derived from Coleman et al. (1982) were digitized into a geologic map
layer. Interpretation of multi-beam bathymetric data allowed revision and updating
of mudflow lobe features in the mudflow deposition zone (e.g. Fig. 4).
Relative ages of underwater landforms are estimated based on the relative positions of sediments (e.g. the law of superposition with younger sediments covering
older ones) and evaluation of relative geomorphic expression of landform surfaces.
For example, mud-flow channels with rough surfaces typically have moved recently
358
C. Hitchcock et al.
Fig. 3 Flow chart showing GIS-based mudflow susceptibility mapping procedure (adapted after
Bachhuber and Hengesh 2000)
GIS-Based Assessment of Submarine Mudflow Hazard Offshore
359
Fig. 4 Modification of mudlobe front from Coleman et al. (1982), shown as black line, to that based
on 1989 NOAA multi-beam data, shown as red line. Colored band shows difference between 1989
and 1979 bathymetry with warm areas (yellow through red) depicting net accumulation
Fig. 5 Mudflow susceptibility map of study area with higher hazard areas shown in warmer colors
(i.e. within a year or two). Thus, detailed mapping provides information on the age
and, indirectly, relative strength of stored sediment.
Based on relative age and activity, mudflow gullies typically are most active
areas of sediment transport on the seafloor and are assigned the highest hazard
point score (Fig. 5). Mudflow lobes are areas of recent deposition and vulnerable
360
C. Hitchcock et al.
to mudflow overruns from upslope mudflows and slope failures, and thus also
assigned high hazard. Older marine landslides are assigned lower hazard ratings, as
is slightly disturbed seafloor. Areas of undisturbed seafloor as mapped by Coleman
et al. (1982) have the lowest associated hazard and point score.
3
Results
Our final mudflow hazard map (Fig. 5) depicts areas of relative susceptibility to
mudflow failure and related change, i.e. local erosion or burial. Areas of ‘very high’
mudflow susceptibility may have extensive mudflows triggered during winter
storms or by hurricanes along with associated submarine slope failures on steep
slopes. Industry structures may be damaged by mudflow transport and deposition
of new mudflows. Based on documented regional pipeline damage from Hurricane
Ivan (Coyne and Dollar 2005; Mirza et al. 2006), our mapping appears to have
adequately captured areas of greatest net seafloor change and associated damage,
including locations of damaged platforms.
Comparison of post-Ivan bathymetric data with pre-Ivan data for a test area
allows us to compare locations where mudflows likely occurred with our GIS-based
mapping (Fig. 6). Our mapping adequately characterizes the largest mudflow lobes
Fig. 6 Comparison of areas of net seafloor accumulation and loss derived from post-Hurricane
Ivan multi-beam dataset with mudflow hazard map derived from 1977–1979 geologic mapping
and bathymetric data provided in Coleman et al. (1982). Arrows show inferred directions of mass
movement (from Hitchcock et al. 2006)
GIS-Based Assessment of Submarine Mudflow Hazard Offshore
361
that have moved since 1977–1979 and which present the largest hazard. Comparison
of bathymetric surveys performed in 1977–1979 and 2004 show that these large
flows are migrating downslope.
4
Discussion
Our approach focused on delineating and ranking areas of greatest historic change
for the Mississippi delta area as part of our map-based evaluation of portions of the
seafloor with the highest likelihood of future failure. However, equally important,
are areas of the seafloor that apparently have been relatively stable over time and
thus likely less susceptible to future failure.
Major areas of net sediment accumulation are located off the main distributaries.
These areas have progressed seaward but continue to be directly connected to
deposition at the mouths of the distributaries. Major and minor mudflow gullies
appear to have remained in relatively the same locations over the same time
periods, with relatively minor lateral shifts. The deposition zone, located downslope
of the gullies is an area of net change with recent to active deposition. A northeasttrending zone of sediment accumulation present offshore of South Pass from
1874 to 1977, coincides with a region of relatively steep slopes formed by the
front of young mudflow lobes. The high slopes and high-rates of deposition
suggest the mudflow lobe deposition zone should continue to be considered a
very high hazard area as existing and new infrastructure may be damaged by
future mudflows.
In particular, localized damage and burial of pipelines and production facilities
is largely due to mudflow overruns from upslope gullies and mudflows. The distal
portion of the mudflow lobe deposition zone coincides with the headward migration
of salt-related growth faults near the shelf edge. This association suggests a possible
relationship between the two processes and associated potential hazards in this
area. Although a distinctly separate process from that of sediment transport and
deposition, long-term displacement on these faults cause local steepening of
submarine slopes that may serve to intensify slope failures.
Although locally associated with sediment transport and associated high
mudflow hazard, major and minor mudflow gullies appear to have remained in
relatively the same locations over the same time period, with minor lateral shifts.
In particular, the residual map derived from the difference between the 1940 and
1874 bathymetric data sets highlighted: (1) a northeast-trending zone of mid-slope
deposition and (2) poorly expressed south-southeast-trending zones of net seafloor
loss (Fig. 7a). The seafloor loss areas roughly coincide with mudflow gullies (e.g.
regions of active mudflow transport to the mudflow lobe front) mapped by Coleman
et al. (1982). The residual map derived from the difference between the 1977–1979
and 1940 bathymetric data sets shows less deposition in the center of the zone
of gully transport and much better expression of the mudflow gullies (Fig. 7b).
362
C. Hitchcock et al.
Fig. 7 Residual maps of seafloor change produced by (a) the difference between 1940 and 1874
bathymetry and (b) between 1977–1979 and 1940 bathymetry
GIS-Based Assessment of Submarine Mudflow Hazard Offshore
363
This apparent evolution could reflect sediment bypass of the former depositional
areas during progressive mudflow transport events or may capture the seafloor
impact of Hurricane Camille in 1969.
Based on interpretation of the post-Ivan bathymetry, areas of net seafloor
erosion are most closely associated with mudflow gullies (Fig. 2). These gullies
appear to have been emptied between 1977–1979 and 2004, with the intervening
gully interfluves relatively unchanged during this same period. Net seafloor loss
also is closely associated with downslope failure of large mudflow lobes, in 300 to
400 feet of water, with net deposition within the mudflow lobe front. Failure of
the mudflow lobes appears to consist primarily of downslope slumping with net
loss in the uphill portion of the mudflow lobes and net deposition downslope.
This deposition consists of mudflow lobe overflow onto older lobes, causing oversteepended mudfront slopes.
Progression of the mudflow into the gulf appears to be accompanied by gulfward
development of gullies (e.g. limits of gully region shown on Fig. 4). Areas of relative stability located between active mudflow gullies likely primarily consist of
older sediment. Age of the sediment following its last failure episode is a major
material factor controlling future failure (Hooper 1980; Hooper and Suhayda
2005). The younger the deposit, the more likely it is to fail in the next large storm.
Remnants of seafloor between mudflow lobes that don’t fail as mudflows likely
serve to channel the gully flows and become increasing stable as part of the
gulfward evolution of the gully system into deeper water.
5
Conclusions
Using existing data, we have developed, and applied, a GIS-based approach
for delineating the relative susceptibility of underwater slopes to mudflows.
Our hazard mapping incorporates all available, existing datasets to produce updated
hazard maps that rank the seafloor susceptibility to mudflow failure and provide
a framework for inclusion and interpretation of recent and future bathymetric
data, geologic mapping, and compilation of geotechnical data. Based on comparison
with post Hurricane Ivan bathymetric data and published pipeline damage reports,
our map of mudflow hazards accurately characterized areas of relatively high hazard.
As previously identified and mapped by Coleman et al. (1982), these areas are
associated with active mudflow channels and the mudflow lobe deposition front.
Our mapping approach can be used on a regional basis to highlight relative mudflow
hazard or, by incorporating detailed site-specific data, to evaluate hazards for
pipeline corridors and siting of production facilities.
Acknowledgments Our study was funded by the Mineral Management Service under MMS
Order No. 0105PO39228. The authors would like to thank Mik Else of the MMS Engineering &
Research Branch, K.C. Gan of Fugro McClelland, and Mimi D’Iorio for their assistance and
support of this study.
364
C. Hitchcock et al.
References
Bachhuber JL, Hengesh JV (2000) Seismic hazard zonation of greater San Juan area, northern
Puerto Rico Coastal Plain: Proceedings Sixth International Conference on Seismic Zonation,
Earthquake Engineering Research Institute, November 2000, Palm Springs, California
Bea RG, Bernard HA, Arnold P, Doyle EH (1975) Soil movements and forces developed by waveinduced slides in the Mississippi Delta: Journal of Petroleum Technology 7:500–514
Coleman JM, Prior DB, Garrison LE (1982) Subaqueous sediment instabilities in the offshore
Mississippi River Delta: Environmental Information On Hurricanes Deep Water Technology,
and Mississippi Delta Mudslides in the Gulf of Mexico, Bureau of Land Management OpenFile Report 80:1–49
Coyne MJ, Dollar JJ (2005) Shell Pipeline’s Response and Repairs after Hurricane Ivan, OTC
Paper #17734
Hitchcock C, Givler R, Angell M, Hooper JR (2006) A Pilot Study for Regionally-Consistent
Hazard Susceptibility Mapping of Submarine Mudslides. OTC paper #18323
Hitchcock C-S, Loyd R-C, Haydon W-D (1999) Mapping liquefaction hazards in Simi Valley,
Ventura County, California: Environmental and Engineering Geoscience 5:441–458
Hooper JR (1980) Crustal layers in Mississippi Delta mudflows, OTC Paper #3370:277–287
Hooper JR (1996) Foundation soil motion in South Pass 47, Offshore Technology Conference,
OTC Paper #7953
Hooper JR, Suhayda JN (2005) Hurricane Ivan as a geologic force: Mississippi Delta Front
sea-floor failures: OTC Paper #17737
Kesel RH (1988) The decline in the suspended load of the lower Mississippi River and its
influence on adjacent wetlands: Environmental Geology and Water Science 11:271–281
Mirza S, Skinner J, Mathew A, Ekstrom L (2006) Hurricane Ivan-Pipeline Damage, Integrity
Assessment and on Bottom Stability, OTC Paper #18183
Morgan DJ (1977) Recent geologic history and geomorphology of the Mississippi River Delta:
Geoscience and Man, XVI, Louisiana State University, p. 34
Roberts HH (1997) Dynamic changes of the Holocene Mississippi River Delta plain: The delta
cycle: Journal of Coastal Research 13:605–627
Saucier R (1963) Recent geomorphic history of the Ponchartrain basin: Louisiana State University
Coastal Studies Series 9, 114 p
Shepard RP (1955) Delta-front valleys bordering the Mississippi distributaries: Geological
Society of America Bulletin 66:1489–1498
Whelan T, Ishmael JT, and Rainey GB (1978) Gas-sediment interactions in Mississippi delta
sediments: Offshore Technology Conference, OTC Paper #3166