NHESSNatural Hazards and Earth System SciencesNHESSNat. Hazards Earth Syst. Sci.1684-9981Copernicus PublicationsGöttingen, Germany10.5194/nhess-16-1583-2016GPS-derived ground deformation (2005–2014) within the Gulf of Mexico region referred to a stable Gulf of Mexico reference frameYuJiangboyujb6@mail.sysu.edu.cnhttps://orcid.org/0000-0001-5353-8435WangGuoquanSchool of Earth Science and Geological Engineering, Sun Yat-sen University, Guangzhou, ChinaDepartment of Earth and Atmospheric Sciences, National Center for Airborne LiDAR Mapping, University of Houston, Houston, Texas, USAJiangbo Yu (yujb6@mail.sysu.edu.cn)7July20161671583160227August20153November201510June201621June2016This work is licensed under a Creative Commons Attribution 3.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/3.0/This article is available from https://nhess.copernicus.org/articles/16/1583/2016/nhess-16-1583-2016.htmlThe full text article is available as a PDF file from https://nhess.copernicus.org/articles/16/1583/2016/nhess-16-1583-2016.pdf
This study investigates current ground deformation derived from the GPS
geodesy infrastructure in the Gulf of Mexico region. The positions and
velocity vectors of 161 continuous GPS (CGPS) stations are presented with
respect to a newly established local reference frame, the Stable Gulf of
Mexico Reference Frame (SGOMRF). Thirteen long-term (> 5 years)
CGPS are used to realize the local reference frame. The root mean square (RMS)
of the velocities of the 13 SGOMRF reference stations achieves
0.2 mm yr-1 in the horizontal and 0.3 mm yr-1 in the vertical directions. GPS
observations presented in this study indicate significant land subsidence in
the coastal area of southeastern Louisiana, the greater Houston metropolitan
area, and two cities in Mexico (Aguascalientes and Mexico City). The most
rapid subsidence is recorded at the Mexico City International airport, which
is up to 26.6 cm yr-1 (2008–2014). Significant spatial variation of
subsidence rates is observed in both Mexico City and the Houston area. The
overall subsidence rate in the Houston area is decreasing. The subsidence
rate in southeastern Louisiana is relatively smaller (4.0–6.0 mm yr-1)
but tends to be steady over time. This poses a potential threat to the
safety of coastal infrastructure in the long-term.
Introduction
The Gulf of Mexico (GOM) region has been the heart of the U.S. energy
industry because of substantial oil and gas deposits along the coast and
offshore of the GOM. It is heavily populated and vulnerable to local ground
deformation (faulting, subsidence, uplift) and relative sea-level rise
(e.g., Day et al., 1995; Kolker et al., 2011; Thatcher et al., 2013). Land
subsidence and faulting problems in the GOM region have been frequently
investigated by different research groups using GPS observations
(e.g., Dokka, 2011; Engelkemeir et al., 2010; Kearns et al., 2015; Khan et al.,
2014; Osmanoǧlu et al., 2011; Wang and Soler, 2013). However, it is
difficult to align the results from these research groups because they used
different data sets collected by different organizations during different
time periods. Furthermore, they focused on localized ground deformation and
applied different reference points or frames. This study aims to establish a
unified local geodetic reference frame, the Stable Gulf of Mexico Reference
Frame (SGOMRF), to investigate the current ground deformation within the
whole GOM region during the past decade (2005–2014). Observations of
161 high-quality continuous GPS stations among 450 active long-term GPS stations
(Fig. 1) are investigated in this study. Land subsidence and faulting in the
Houston region, Mexico City, and the southeastern Louisiana region are
discussed and compared.
Map showing current CGPS stations in the GOM region. Blue
triangles represent current CGPS installed in and after 2009. Red circles
represent current CGPS installed before 2009. Grey squares represent
decommissioned CGPS that have data spanned for more than 5 years. Data are
available through NGS and UNAVCO.
GPS data processing
This study applies the precise point positioning (PPP) method, for solving
the 24-hour average position of a GPS antenna. PPP is based on the
processing of the following ionosphere-free combinations of the
undifferenced code and phase observations (Zumberge et al., 1997):
PIF=f12⋅P(L1)-f22⋅P(L2)f12-f22=ρ-c⋅dT+dtropΦIF=f12⋅Φ(L1)-f22⋅Φ(L2)f12-f22=ρ-c⋅dT+dtrop+cf1N1′-cf2N2′f12-f22,
where f1 and f2 are the GPS L1 and L2 frequencies; P(Li) and
Φ(Li) are the code and phase observations at the corresponding
frequency; ρ is the true range; c is the speed of light; dT is the
receiver clock offset; dtrop is the tropospheric delays;
Ni′ is the phase ambiguity term in Φ(Li). Therefore, the unknown
parameters estimated in PPP include position coordinates, phase ambiguity,
receiver clock offset and the tropospheric delays. GNSS-Inferred Positioning
System and Orbit Analysis Simulation Software (GIPSY-OASIS) package (V6.3)
developed at the Jet Propulsion Laboratory (JPL) is applied for calculating
daily positions. The GIPSY-OASIS package provides single receiver phase
ambiguity fixed PPP solutions. The single receiver phase ambiguity method
uses the wide lane and phase bias estimates obtained from a global network
of ground GPS stations to perform ambiguity-resolved PPP resolution
(Bertiger et al., 2010). The major parameters estimated and
key models applied in the PPP processing include: the VMF1 troposphere
mapping model (Boehm et al., 2006), second-order
ionospheric delay (Kedar, 2003), the ocean tidal
loading model FES2004 (Lyard et al., 2006) calculated through
the free online service operated by Onsala Space Observatory, Sweden
(Free ocean tide loading provider, 2015), tropospheric
gradient (Bar-Sever et al., 1998), zenith
troposphere delay as a random walk with variance of 5 × 10-8 km s-1,
gradient troposphere wet delay as
a random walk with variance of 5 × 10-9 km S-1, and receiver clock as white noise with updates every measurement
epoch. Station coordinates are initially provided in the loose frame of the
JPL's fiducial-free GPS orbits. The coordinates are then transformed into
the International GNSS Service Reference Frame of 2008 (IGS08) using the
daily seven transformation parameters that are delivered with the JPL's
orbit products.
The PPP processing conducts all calculations within an
Earth-centered–Earth-fixed (ECEF) geocentric coordinate system (x, y, and
z). In order to track land surface deformation, the ECEF geocentric
coordinates are converted to cartographic (northing, easting, and
ellipsoidal height) coordinates, which are referred to the GRS-80 ellipsoid.
The three-component daily positional time series indicates the change of
ground surface over time at different directions. The three components
represent: north, east, and up. The up component (subsidence/uplift)
measurements used in this study are obtained by differencing GPS measured
ellipsoidal heights referred to the local reference frame described in the
next section. Detection of regional-scale land subsidence has historically
depended on surveying benchmarks periodically. This has traditionally been
accomplished by differencing orthometric heights obtained from spirit
leveling. A recent investigation conducted by Wang and Soler (2014)
has indicated that using ellipsoid and
orthometric heights would result in the same practical subsidence
measurements. Accordingly, the subsidence values used in this study can be
regarded as having the same “physical meaning” as the conventional
subsidence measurements obtained from leveling surveys. For each time series,
the outliers, defined as the days for which the uncertainty was greater than
2.0 times of the average uncertainty of the entire measurement, were removed
(Firuzabadì and King, 2011; Wang, 2011). The
uncertainty of each measurement was directly output by the GIPSY-OASIS
program. On average, 5 % of the total samples are removed as
outliers. The daily positional time series applied in this article are the
“cleaned” time series.
Stable Gulf of Mexico Reference Frame (SGMRF)
In general, a global or a continental-scale reference frame is realized with
an approach of minimizing the least square residual velocities of a large
number of selected reference stations. In the case of IGS08,
232 globally distributed, and well-performing GPS stations are used
(Rebischung et al., 2012). The velocities at
GPS sites referred to IGS08 are dominated by tectonic drift. For a regional
study, a stable local reference frame is often established through Helmert
transformation to exclude tectonic drift (Wang et al., 2013,
2014, 2015a). It facilitates the precise physical interpretation of local
ground deformation over time and space. The transformation involves seven
parameters including a rotation vector, a translation vector and one scale
factor. These seven transformation parameters can be estimated by comparing
the positions of a group of selected reference stations referred to the new
reference frame with those referred to a well-established reference frame.
Map showing the locations and velocity vectors with 95 %
confidence ellipses of the 13 reference stations used to define SGOMRF.
Black vectors are referred to NAD83; blue vectors are referred to IGS08; red
vectors are referred to SGOMRF.
In practice, at least three reference stations are needed to obtain the
transformation parameters. More reference stations often result in a more
reliable coordinate transformation. However, a reference station that is not
locally stable will degrade the overall performance of the frame
transformation. A stable site is defined as retaining zero velocities (three
components) with respect to a specified reference frame. The stability
(precision) of a local reference frame is therefore affected by the
velocities of stable sites with respect to the reference frame. Thus, the
selection of reference stations is critical for establishing a stable local
reference frame. In general, there is not a fixed criterion for selecting
reference stations. The selection mostly is based on the availability of
long-term CGPS stations in the study area. There are over 780 CGPS stations
in the GOM region (Fig. 1). As this study uses a
secular frame, the linearity of daily positional time series is a critical
criteria for selecting reference stations (Blewitt and
Lavallée, 2002). Additionally, the geographic distribution of reference
stations is also considered. The following specific criteria are initially
applied for selecting reference stations:
Having segments of data spanning at least 5 years (installed in 2009 or
earlier) with no steps (a sharp change of the mean in positional time series
caused by an earthquake, equipment change, or other unknown reasons).
No considerable subsidence or uplifting (the linear velocity rate of
vertical positional time series referred to IGS08 is less than 0.5 mm yr-1);
having less than 0.1 mm yr-1 “standard error” (σ) of the
slope (Vcal, Vcal) of the geocentric
coordinate time series (x, y, and z) referred to IGS08. σ is a
measure of the error in the precision with which Vcal has been
estimated by a linear regression. A smaller σ indicates a small
margin of error. Approximately 95 % of the time, the true velocity will be
contained in the interval between Vcal- 1.96 ×σ
and Vcal+ 1.96 ×σ.
The near-coast areas could be affected by subsidence and coastal erosion
problems (Simms et al., 2013; Williams et al., 1997; Yu et al., 2014). Accordingly, it is
preferred that reference stations be located inland rather than within
near-coast areas. However, in order to balance the overall coverage and
geometrical distribution of reference stations, one near-coast station in
Florida (RMND) and one near-coast station in Mexico (TAM1) were selected as
reference stations (Fig. 2). Uneven distribution of
reference sites could lead to biases in frame transformation
(Collilieux et al., 2010; Wang et al., 2014).
Positional time series affected by steps are identified by an automated edge
detection program based on the derivative of Gaussian kernel
(Canny, 1986). Initially, 30 CGPS stations were selected as
reference stations and the reference frame transformation were calculated.
Any station that had a horizontal velocity larger than 0.5 mm yr-1 with
respect to the resulted local reference frame was removed from the group of
reference stations and the transformation was recalculated again. Finally,
13 CGPS stations are selected as reference stations for realizing the
SGOMRF (Fig. 2).
Two different approaches are often used in geodesy to transform positional
time series from one reference frame to another: the daily 7-parameter
Helmert transformation (Blewitt et al., 2013) and the
14-parameter similarity transformation. This study applies a 14-parameter
transformation approach that has been frequently applied in the geodesy
surveying community (e.g., Pearson and Snay, 2012; Wang et al., 2014). The geocentric coordinates of a station
with respect to SGOMRF are calculated by the following formulas:
X(t)SGOMRF=TX(t)+[1+s(t)]⋅X(t)IGS08+RZ(t)⋅Y(t)IGS08-RY(t)⋅Z(t)IGS08Y(t)SGOMRF=TY(t)-RZ(t)⋅X(t)IGS08+[1+s(t)]⋅Y(t)IGS08+RX(t)⋅Z(t)IGS08Z(t)SGOMRF=TZ(t)+RY(t)⋅X(t)IGS08-RX(t)⋅Y(t)IGS08+[1+s(t)]⋅Z(t)IGS08.
Here, Tx(t), TY(t) and TZ(t) are translations along x, y and z axis; RX(t),
RY(t) and RZ(t) are counterclockwise rotations about three axes; s(t) is a
differential scale factor between IGS08 and SGOMRF. These seven parameters
at any point in time are specified relative to a reference epoch by the
following linear relationships:
TX(t)=TXt0+dTX⋅t-t0TY(t)=TYt0+dTY⋅t-t0TZ(t)=TZt0+dTZ⋅t-t0RX(t)=RXt0+dRX⋅t-t0RY(t)=RYt0+dRY⋅t-t0RZ(t)=RZt0+dRZ⋅t-t0s(t)=st0+ds⋅t-t0.
Here, t0 denotes a specific epoch (e.g., 2013.0). TX(t0),
TY(t0), TZ(t0), RX(t0), RY(t0), RZ(t0)
and s(t0) are the seven transformation parameters at epoch t0.
The two reference frames are aligned at epoch 2013.0. That means
the positional coordinates of a site with respect to both reference frames
are identical at this epoch. Thus, the seven transformation parameters at
epoch t0 are all zeros. dTX, dTY, dTZ, dRX,
dRY, dRZ, and ds are the first time derivatives of corresponding
parameters, which are constant over time. The units of these parameters are
meters for translational components, radians for rotational components,
m yr-1 for the rate of translational movement, and radian/year for the rate
of rotational components. s(t) is a unitless scale factor.
The unit of ds is 1/year. We use the same procedure as described in Wang
et al. (2014) to obtain the 14 parameters. The two
reference frames were aligned at the epoch 2013.0. Thus the coordinates of a
station at epoch 2013.0 with respect to the two reference frames are
identical. As a result, the seven parameters at the epoch 2013.0 for
reference frame transformation are all zeros. The seven transformation
parameters at epoch 2000.0 were calculated by comparing the coordinates with
respect to the two reference frame and solving the parameters in Eq. (3)
through least squares estimation. The coordinates with respect to SGOMRF at
epoch 2000.0 are assumed to be equal to those at epoch 2013.0 since the
sites are considered to have zero velocities with respect to SGOMRF. The
rates of these seven parameters over time were calculated using the
following formula:
dTX=TX(2013.0)-TX(2000.0)/13.0dRX=RX(2013.0)-RX(2000.0)/13.0ds=(s(2013.0)-s(2000.0))/13.0.
Figure 3 depicts the three-component positional
time series of two stations (OKAN and SG05) referred to SGOMRF, North
American Datum of 1983 (NAD 83) (2011), and IGS08. The NAD 83 is the
horizontal control datum for the US, Canada, Mexico, and Central America
(Schwarz, 1989; Snay and Soler, 2000). It is widely used as a
North American plate-fixed reference frame in the practice of surveying. The
continental-scale reference frame has been updated for several times. The
most recent realization is referred as NAD83(2011) at epoch 2010.0. The
positional coordinates referred to NAD83(2011) are transformed from IGS08
with Eqs. (3) and (4) and 14 parameters provided by Pearson and Snay (2012) (Table 1). Both sites retain near-zero
velocities (< 0.5 mm yr-1) with respect to the local reference frame
SGOMRF. The three-component velocities derived from the 10-year (2004–2014)
continuous observations at OKAN are -2.3 mm yr-1 (north),
-13.4 mm yr-1 (east), and 0.2 mm yr-1 (up) with respect to IGS08 and 0.6 mm yr-1 (north),
1.6 mm yr-1 (east), and -0.2 mm yr-1 (up) with respect to NAD83. The
velocities at SG05 are generally the same (Fig. 3).
The horizontal velocity (1.5 cm yr-1) referred to IGS08 can be explained by
the tectonic movement of the GOM region with respect to the global reference
frame. However the minor horizontal movements (∼ 2 mm yr-1)
with respect to the supposedly continent-fixed reference frame (NAD83) can
be misleading. The same minor horizontal movements with respect to
NAD83(2011) can also be observed at the 13 reference stations, even though
the near-zero velocities with respect to SGOMRF indicate they are stable
(Fig. 2). The coherent horizontal movement referred
to the NAD83(2011) could be incorrectly interpreted as the result of local
active faulting. In fact, the 2 mm yr-1 horizontal velocity does not
represent any local ground movement. It indicates the low precision of the
NAD83 reference frame within the GOM region. The instability of the
continental-scale reference frame will overlook or bias minor local
horizontal ground deformation signals.
Fourteen parameters for reference frame transformations from IGS08
to SGOMRF and from IGS08 to NAD83(2011).
a Source: Pearson and Snay (2012), Table 7;
b counterclockwise rotations of axes are positive;
c mas = milliarc second, radians to mas coefficient: 206264806.24709636;
mas to radians coefficient: 4.848137 × 10-9;
d ppb = parts per billion.
Comparisons of time series at two CGPS sites, OKAN and SG05,
referred to three reference frames: IGS08 (black), NAD83 (blue), and SGOMRF
(red). OKAN is located at Antlers, Oklahoma. SG05 is located at Melbourne, Florida.
Geodetic coordinates (longitude, latitude, ellipsoidal height) of
the 13 reference sites with respect to SGOMRF and their RMS accuracy (repeatability).
The root-mean-square (RMS) value of the linear velocities of these
13 reference stations are 0.15 mm yr-1 for the east component, 0.19 mm yr-1 for
the north component, and 0.25 mm yr-1 for vertical component with respect to
the local reference frame (SGOMRF). Table 2 lists the average coordinates
of these 13 reference stations referred to SGOMRF at epoch 2013.0 and the
RMS values of corresponding residual time series. These RMS values are often
regarded as the precision (repeatability) of the daily positions. The
results illustrated in Table 2 suggest that the PPP solutions obtained in
this study achieve 2 mm horizontal precision and 7 mm vertical precision,
which is comparable with the overall precision of GIPSY-OASIS PPP solutions.
Bertiger et al. (2010) reported that the single receiver
ambiguity fixed PPP solutions achieved overall 2 mm horizontal precision and
6 mm vertical precision for the 106 worldwide IGS reference stations. It
should be noticed that the 14-parameter transformation processing did not
improve the precision of velocity estimates since a linear regression model
was applied to the changes of the scale, three translational, and three
rotational motions in time.
Horizontal ground deformation
This study investigated GPS data from 148 stations that have step-free time
spans of longer than 4 years, and transform the daily positional time
series to the local reference frame (Fig. 4). The
average horizontal velocity of these 148 stations is below 1 mm yr-1, which
implies that the interior of the GOM region is rigid at the level of
sub-millimeter per year. Five long-term GPS stations have been moving
horizontally with relatively larger velocities (> 2 mm yr-1).
These stations are MMX1 (9.7 mm yr-1), FSHS (3.4 mm yr-1), UNIP
(2.9 mm yr-1), TXPR (2.4 mm yr-1) and ROD1 (2.4 mm yr-1). Four of these stations
are located in well-known subsidence areas. MMX1 is located in Mexico City,
Mexico. FSHS is located in Franklin, Louisiana. UNIP is located in
Aguascalientes, Mexico. ROD1 is located in Houston, Texas.
In the case of groundwater withdrawal induced subsidence, a subsidence bowl
can be formed by localized aquifer compaction. In such an event, it is
possible that GPS stations are pulled horizontally towards the center of the
subsidence bowl (e.g., Allis, 2000; Bawden et al., 2001, 2012). Depending on the position of a
station relative to the subsidence center, the ratio of horizontal to
vertical velocities varies. A station at the edge of the subsidence center
will show a relatively large horizontal velocity. A station located closer
to the center of subsidence could display large velocities in both the
horizontal and vertical components. Around the center of the subsidence
feature, the stations will mostly exhibit vertical movement.
Figure 5 depicts the three-component positional
time series at ROD1, TXCN, and TXLI referred to the local reference frame.
TXLI is a stable station located in Liberty County, Texas. ROD1 shows a
2.3 mm yr-1 horizontal movement towards the northeast. TXCN shows no movement in
the horizontal direction. Both ROD1 and TXCN indicate long-term subsidence.
According to our previous studies on the current subsidence in the greater
Houston metropolitan area (Kearns et al., 2015; Yu
et al., 2014), a rapid-subsidence bowl is forming around The Woodlands area.
The Woodlands is a vibrant and fast-growing business and entertainment
suburban area located 43 km north of downtown Houston (Fig. 6). Groundwater is the sole water source for
residential and business use in this area as of 2014. The subsidence rate in
the center of the subsidence bowl is about 25 mm yr-1 (Kearns et al., 2015). ROD1 is located in
the city of Spring, Texas, northern Harris County. The station is 12 km
southwest to The Woodlands area (Fig. 6). The
subsidence rate at ROD1 is 17 mm yr-1 derived from the whole time series
from 2007 to 2014. The closely spaced contour lines represent greater
spatial variation of subsidence rates in this developing subsidence bowl.
The horizontal velocity vector indicates that ROD1 is affected by the
differential subsidence. TXCN is located in the city of Conroe, Texas,
which is 21 km north of The Woodlands. The positional time series (2008–2014)
of TXCN does not indicate any considerable horizontal movement
(< 1 mm yr-1). However, steady land subsidence with a rate of
16 mm yr-` has been recorded at this site. The contour lines in this area are
spaced far apart. The different vertical-to-horizontal velocity ratio
suggests that TXCN is located at a more uniformly subsiding, therefore, flat
area. whereas ROD1 is more likely located along the steep sidewall of the
developing subsidence bowl. The comparison of the three-component positional
time series of ROD1 with those of TXCN illustrates a good example of
horizontal velocity variations around a subsidence bowl.
(a) Horizontal velocity with 95 % confident ellipses and
(b) vertical velocity vectors of 148 CGPS stations (> 4 years) in the
Gulf of Mexico region. Green dots represent the 13 reference stations. The
blue vectors represent the average velocities referred to SGOMRF.
Displacement time series of two rapidly subsiding CGPS sites (ROD1
and TXCN) in northern Houston. The displacement time series of a stable site (TXLI)
are plotted for comparative purposes. Locations of these three stations are
plotted in Fig. 6. The reference frame is SGOMRF.
A similar movement pattern at two CGPS sites around a subsidence bowl is
also observed in Mexico City (Fig. 7). MMX1 is
located in the Mexico International Airport, eastern Mexico City. UNIP is
located at the Universidad Nacional Autónoma de México, southwestern
Mexico City. UNIP records a smaller rate of 2.9 mm yr-1 towards the
northeast. Observations from InSAR indicated that subsidence rates in Mexico
City increased eastwards towards the center of the Basin of Mexico
(Chaussard et al., 2014). The horizontal movements of MMX1 and UNIP agree well with the subsidence bowl
illustrated by the InSAR data: both UNIP and MMX1 are moving toward the
center of the subsidence bowl. MMX1 is located much closer to the center and
therefore has demonstrated higher rates of horizontal motion (Fig. 7).
Locations of three CGPS stations and cities within the greater
Houston area. Black and blue velocity vectors represent horizontal and
vertical velocities with respect to SGOMRF, respectively. Red lines are
growth faults and yellow dots are salt domes (Garrity and Soller, 2009).
Contours lines are subsidence rate from 2005 to 2012 (Kearns et al., 2015).
Horizontal (red) and vertical (green) velocity vectors at MMX1 and
UNIP with respect to the local reference fame (SGOMRF). The color patterns
represent the average subsidence rate derived from InSAR analysis (Chaussard
et al., 2014). The three-component positional time series of UNIP and MMXI are
plotted in Fig. 10.
Three-component displacement time series from two CGPS sites with
considerable horizontal movements. FSHS is located at Franklin, Louisiana
(Fig. 9a). TXPR is located at Pharr, Texas (Fig. 4). The reference frame is SGOMRF.
Figure 8 depicts that FSHS, a permanent GPS station
(2010–2014) located at Franklin, Louisiana, has been moving toward the
southeast at a rate of 3.4 mm yr-1. The antenna of FSHS is mounted on a
reinforced concrete building located at Franklin High School. Subsidence at
this site is consistent and could lead to more serious problem over the long
term. There is no known excessive groundwater withdrawal issue in this
area. It is not likely that the horizontal movement is associated with an
on-going subsidence bowl. Our other GPS sites (AWES, DSTR, HOUM, GRIS, BVHS,
LMCN) in southeastern Louisiana also demonstrate movements southward with
rates smaller than 1 mm yr-1. Dokka et al. (2006) observed a similar southward
displacement and proposed the detaching of the South Louisiana Allochthon
which contributed to both the southward motion and subsidence in this area.
Latter studies, however, suggested the major driving mechanism of subsidence
in this area is the compaction of shallow strata with minor or no
contribution from active faulting and deep crustal processes (Blum
et al., 2008; Edrington et al., 2008; Törnqvist et al., 2008;
Wolstencroft et al., 2014; Yu et al., 2012). Our data show a lower rate of
southward displacement, which may indicate that the horizontal motion is
mainly caused by differential compaction and minor contribution from
downslope movement of listric faults. The large horizontal velocity at FSHS
may be a site-specific feature related to the obvious seasonal motion in the
EW component. TXPR, a permanent GPS site (2005–2014) located at Pharr,
Texas, has a horizontal southwest movement of 2.4 mm yr-1 (Fig. 8).
The antenna pole is anchored on a wall of an
office building owned by the Texas Department of Transportation. GPS
observations also show steady subsidence (5.8 mm yr-1) at this site. The
horizontal movement could be associated with local subsidence. Further study
is needed to verify the cause of the horizontal motion.
Vertical ground deformation
The overall spatial variation of vertical velocities in the GOM region is
much greater compared to that of horizontal velocities. There are certain
stations showing extremely large downward vertical velocities in this
region. Figure 4b indicates four rapid subsidence
zones in the GOM region – the southeastern Louisiana area, the Houston
metropolitan area, Aguascalientes, and Mexico City. The drivers of
subsidence vary from place to place. Different drivers would result in
different subsidence patterns – the spatial and temporal variability of
subsidence rates. In this section, we discuss subsidence in southeastern
Louisiana, Houston, Aguascalientes and Mexico City.
Subsidence in the southeastern Louisiana
The causes behind present subsidence in southeastern Louisiana have been
controversial and heavily studied. Ramsey and Moslow (1987)
attributed 80% of the present subsidence on the coast of Louisiana to
“compactional subsidence”. Roberts et al. (1994) studied
relationships between subsidence rates and faulting, land loss, thickness
and characteristics of Holocene sediment in the Louisiana coastal area, and
they concluded that sediment compaction was a primary cause of subsidence. A
number of studies proposed that the present-day subsidence in the
Mississippi Delta is mostly caused by the isostatic response to the delta
load (e.g., Ivins et al., 2007; Jurkowski et al., 1984). However, Dokka (2006) argued the conventional
opinions; using a case study conducted in the Michoud area of Orleans
Parish, Louisiana, Dokka (2006) concluded that 73 % (16.9 mm yr-1) of subsidence during the period
1969–1971 and 50 % (7.1 mm yr-1) of subsidence during the period 1971–1977
was attributed to tectonism (fault movements). Dokka and his colleagues
further addressed tectonic-induced subsidence in their other publications
(Dixon et al., 2006; Dokka, 2011; Dokka et
al., 2006). Wolstencroft et al. (2014) investigated the
cause of subsidence in the Mississippi Delta through geophysical modeling
and concluded that present-day basement subsidence rate due to sediment
loading was less than ∼ 0.5 mm yr-1 and the glacial isostatic
adjustment was likely to be the major driver of deep-seated subsidence.
Top maps show vertical and horizontal velocity vectors in
(a) southeastern Louisiana and (b) Houston. Bottom plots show vertical
positional time series of subsiding stations in (c) southeastern Louisiana
(> 1.5 mm yr-1) and (d) Houston-Galveston (> 4.5 mm yr-1).
The reference frame is SGOMRF.
Figure 9a and c illustrate the velocity vectors
and positional time series at long-term GPS stations across the southeastern
Louisiana area. Considerable subsidence rates are recorded at two near-coast
stations: LMCN (foundation depth 36.5 m) and GRIS (foundation depth unknown)
(< 10 km to the coastline). Both sites indicate steady subsidence of
approximately 6 mm yr-1 over 10 years. Four inland stations FSHS (foundation
depth > 5 m), AWES (foundation depth 1 m), HOUM (foundation depth > 15 m)
and DSTR (foundation depth unknown) show smaller
subsidence rates (2–4 mm yr-1). The seaward increase in the rate of
subsidence may be a combined result of shallow sediment compaction and deep
basement subsidence. Wolstencroft et al. (2014)
demonstrated that present-day Pleistocene basement subsidence (deep
subsidence) in the Mississippi Delta produced by viscoelastic deformation
mechanisms increased seaward. The natural compaction of young sediments
could occur in new infill and recently drained mashes (Törnqvist et al., 2008). GPS stations
underlain by shallow (Holocene) sediments in this area will be subject to
the ongoing compaction. Most data used in this study are from building-based
stations. Therefore the monuments of GPS stations are building foundations
which are typically 5–15 m below the land surface. In this case, the
contribution from the upper-most Holocene compaction is minimum and our
subsidence rate estimates should be considered minimum estimates. The
foundation depth used in this study is collected from Dokka et al. (2006)
and GPS log files from NGS. Unfortunately, this information is not always
available in the GPS log files. The foundation depth of a GPS site directly
determines its measuring target; therefore, it is of great importance to
check this information while interpreting the data. And we hope more GPS
network maintainers will include this information in the log files.
Compared to near-coast sites LMCN (2005–2014) and GRIS (2005–2014),
another near-coast GPS site BVHS (2002–2014) (foundation depth > 20 m)
recorded a 50 % smaller subsidence rate (3 mm yr-1).
This rate is comparable to the rate (3.5 mm yr-1) reported by Dokka et
al. (2006) and substantially smaller than the rate (5.7 mm yr-1) reported by
Karegar et al. (2015). Note that, due to antenna
changes and two large gaps (13 months and 8 months), the data prior to 2010
were not used to calculate the subsidence rate at BVHS. The choice of record
length may explain the large difference with the rate reported by
Karegar et al. (2015) in which the full data span is used. The reason for the smaller subsidence rate at BVHS
compared to LMCN and GRIS is unclear. Morton and Bernier (2010) showed historical subsidence rates
calculated from repeat leveling surveys at benchmarks along state highway
LA23 between Chalmette and Venice (near the location of BVHS), state highway
LA1 between Raceland and Grand Isle (near the location of GRIS), and state
highway LA 56 between Houma and Cocodrie (near the location of LMCN). The
1965/1966 to 1993 average subsidence rates along LA1 and LA56 was 9.6 and
11 mm yr-1, respectively, whereas, the subsidence rates along LA23 near
the location of BVHS was much larger with greater than 25 mm yr-1 between 1964
and 1971 and decreased to about 18 mm yr-1 between 1971 to 1984. The leveling
lines encompass a much larger area compared to the spot measurements by GPS
stations. Therefore, the subsidence rate measured at the single spot BVHS
may not be able to represent the subsidence rate in a larger area.
Despite the different subsidence rates between coastal sites and inland
sites, the overall spatial variation of subsidence rates across southeastern
Louisiana is relatively smaller compared to that of the Houston metropolitan
area. The slight variation of subsidence rates in space and the steady
subsidence in time suggest that the subsidence in southeastern Louisiana is
not likely dominated by the compaction of shallow aquifers associated with
groundwater pumping. Groundwater pumping induced subsidence often shows
considerable spatial and/or temporal variations as illustrated by subsidence
in the metropolitan area of Houston and Mexico City (discussed in the
following sections). In southeast Louisiana, groundwater withdrawal is
minimal because groundwater quality is affected by saltwater encroachment
(Baumann et al., 2006; Meckel, 2008).
The exception is for the greater New Orleans area, where groundwater is
pumped from shallow upper Pleistocene aquifers. Spatial correlations between
areas of large subsidence and areas with high-yield groundwater wells in the
New Orleans area were reported by Dokka (2011).
Hydrocarbon production has been frequently discussed as one of the possible
drivers for subsidence in southeastern Louisiana (Chang et al., 2014; Kolker et
al., 2011; Meckel, 2008). The oil production in this region peaked at
∼ 446 million barrels in 1968 but has decreased consistently
to less than ∼ 55 million barrels after 2005 (Kolker et al., 2011; Meckel, 2008).
The data used in this study are mostly from 2005, therefore we consider the
possible contribution from hydrocarbon production to the subsidence observed
in our data is marginal.
Horizontal (Vh) and vertical velocities (Vv) of GPS
stations plotted in Fig. 9.
Houston Southeastern Louisiana Reference frame: SGOMRF Reference frame: SGOMRF StationVhVvStationVhVv(mm yr-1)(mm yr-1)(mm yr-1)(mm yr-1)ROD12.33-17.32LMCN0.94-6.30TXCN0.56-16.38GRIS0.46-5.93ZHU10.80-10.80AWES0.74-3.76COH61.72-8.10HOUM0.96-3.62TXHE0.94-7.51FSHS3.49-3.33TXLM1.90-5.03BVHS0.82-3.02DWI10.98-4.63DSTR0.42-1.74TXGA0.12-4.36ANG61.35-2.57TXGV0.66-1.34TXLI0.44-0.44Subsidence in the Houston area
The groundwater induced subsidence in the Houston area has been intensively
investigated by researchers from the US Geological Survey (USGS) (Bawden
et al., 2012; Galloway et al., 1999; Johnson et al., 2011; Kasmarek et al.,
2009, 2010, 2012, 2013), National Geodetic Survey (NGS) (Zilkoski et al., 2003), and local research institutions
(e.g., Engelkemeir et al., 2010; Kearns et al., 2015; Khan et al., 2014; Qu et al.,
2015; Wang and Soler, 2013; Wang et al., 2015b). A recent study conducted by
Yu et al. (2014) indicated that the subsidence
in the Houston metropolitan area is attributed by the compaction of aquifers
within about 500 m to the ground surface. Since groundwater usage changes
according to the local population and land usage, the subsidence resulting
from groundwater withdrawal will vary in space. The subsidence rate will
also change over time in accordance to the city development and policy
changes. Historically, subsidence in the Houston area had been primarily
occurring in the eastern and southeastern portions. A comparison of the
current subsidence and recent subsidence (1915–2001) mapped by the USGS
(Bawden et al., 2012) indicates that the subsidence in the
Houston area has been migrating to the western and northern areas since the 1990s.
The overall subsidence has also been reduced significantly as a
result of rigidly enforced groundwater regulation plans (Harris-Galveston Subsidence District, 2013).
Figure 9b and d illustrate the velocity vectors
and positional time series at long-term GPS stations within the Houston
area. The spatial variation of subsidence in the Houston area is more
significant than that in southeastern Louisiana. The highest subsidence
rates in the Houston area are recorded at two inland sites: TXCN and ROD1
(Table 3). These two stations are more than 100 km away from the coastline.
The subsidence rate is as high as 17 mm yr-1 at TXCN. Two coastal stations
TXGV and TXGA record subsidence rates of only 1.3 and 4.4 mm yr-1,
respectively. Considerable temporal variations in subsidence rates are also
identified at several sites. For instance, the subsidence rate at ROD1 was
25 mm yr-1 before 2010 and has reduced to 13 mm yr-1 after 2010 due to the
enforced groundwater regulation since 2010 (Fig. 5) (Harris-Galveston Subsidence District, 2013).
Salt tectonics
The Gulf Coast area is one of the world's largest salt dome regions. Over
500 salt domes have been discovered onshore and under the sea floor of the
GOM (Beckman and Williamson, 1990). Long-term accumulation
of the salt movements could exert an impact on surface morphological
features and cause fault growth. The vertical velocity vectors illustrated
in this study do not show any considerable vertical movement that can be
resulted from salt dome uplift. Jackson and Seni (1983) showed that the maximum net growth
rate of diapirs in East Texas is 150 to 230 m per million years, which
equals ∼ 0.2 mm yr-1. The ground deformation rate at this
level is below the limit that can be identified with the current GPS geodesy
infrastructure in this region.
Subsidence in central Mexico
The CGPS sites in two central Mexico cities – Aguascalientes and Mexico
City show extremely rapid subsidence. The causality between groundwater
extraction and land subsidence in Mexico City was first investigated in the 1930s
and then the 1940s (Carrillo, 1947; Cuevas, 1936).
Groundwater accounts for nearly half of Mexico City's water usage
(Sosa-Rodriguez, 2010). As of 2011, shallow aquifers in
this region had been seriously overexploited (Engel et al.,
2011). The ground water level has been declining at average rates ranging
from 0.1 to 1.5 m yr-1in different zones since 1983 (Joint Academies Committee on the Mexico City
Water Supply et al., 1995). Figure 10 illustrates
the three-component positional time series at three CGPS sites: INEG, MMX1,
and UNIP. INEG is located in the city of Aguascalientes, Mexico. This
station shows a steady subsidence rate of 25.7 mm yr-1. MMX1 records steady
land subsidence at a rate of 266.3 mm yr-1, which could be the most rapid
subsidence rate ever recorded by a CGPS station. The subsidence rate at the
center of the subsidence bowl could be even larger. In fact, a subsidence
rate as high as 370 mm yr-1 (1996–2005) was derived from InSAR studies in
Mexico City (Cabral-Cano et al., 2008). UNIP records a small subsidence rate of 2.7 mm yr-1. The distance between
MMX1 and UNIP is only about 17.8 km. The subsidence rate at MMX1 is about
100 times higher than that at UNIP. It demonstrates the significant spatial
variation of the subsidence rate in Mexico City.
Three-component displacement time series of three CGPS stations
in central Mexico. The reference frame is SGOMRF. The locations of UNIP and
MMX1 are marked in Fig. 7 and the location of INEG is marked in Fig. 4.
Conclusions
This study utilizes the current GPS geodesy infrastructure in the GOM region
to investigate the ground deformation associated with subsidence and
faulting. A sophisticated regional GPS geodesy infrastructure should include
three components: individual GPS stations, a stable local reference frame,
and sophisticated positioning software packages. Currently, a unified
“local reference frame” does not exist in the GOM region. This study
established a stable local reference frame (SGOMRF) to fill the gap. In the
first release of the SGOMRF, the 14 Helmert transformation parameters for
converting coordinates from the IGS08 to SGOMRF are provided (Table 1). The
SGOMRF will be incrementally improved and periodically updated to
synchronize with the updates of the IGS reference frame. The potential
applications of the local reference frame include providing a consistent
framework for precisely monitoring coastal hazards over space and time,
studying long-term coastal erosion and wetland loss, studying sea level
rise, and comparing research results from different research groups. The
stable reference frame will also be useful for long-term health monitoring
of dams, sea walls, high-rise buildings, long-span bridges and for planning
and designing coastal restoration projects.
GPS observations show significant land subsidence in the coastal area of
southeastern Louisiana, the Houston metropolitan area, Aguascalientes, and
Mexico City. Significant spatial variations of subsidence rates due to
differences in groundwater withdraw and clay layer thickness are observed in
the Houston area and Mexico City. The decrease of the subsidence rate over
time is also observed at GPS stations located in Houston. The GPS sites in
the southeastern Louisiana area show steady subsidence and a general
southward horizontal movement toward the GOM. This may suggest a deep
tectonic process associated with faulting. Subsidence resulting from
faulting would be difficult to stop through human efforts. As a result, the
smaller but steady subsidence (4–6 mm yr-1) would cause considerable
damage to the coastal protection infrastructure (e.g., sea walls, levees,
flood walls, storm surge barriers) in the long term.
GPS observations presented in this study do not show any considerable ground
movements that could be associated with salt tectonics or faults. The
magnitude of salt dome uplift and faulting may be below the level that can
be identified by the current GPS geodesy infrastructure with a time span
less than a decade. A denser CGPS network and a longer period of data
accumulation are crucial for a more comprehensive study of local ground
deformation within the GOM region.
The GPS velocities with respect to SGOMRF
Locations and velocities of 148 GPS stations within the Gulf of
Mexico region with respect to SGOMRF.
StationLatLongHeightDate beginDate endVeastVnorthVupname(deg)(deg)(m)(mm yr-1)(mm yr-1)(mm yr-1)1NSU31.7508266.902428.0584916 Jan 200430 May 20140.180.32-1.811ULM32.529267.924115.9749914 Jun 200330 May 20140.52-0.29-2.71AL4032.9627273.9937210.91131 Jan 200710 Oct 2012-0.19-0.01-0.32AL7031.7827274.0348141.90446 Aug 20065 Nov 2013-0.010.13-1.32ALDI30.2492271.922-19.205828 May 200930 May 20140.45-0.86-1.41ANG629.3016264.5151-9.1886816 Nov 200730 May 20140.75-1.13-2.57ARCM33.5424267.117325.469239 Aug 200530 May 2014-0.150.41-1.52ARHP33.6961266.399484.467479 Aug 200530 May 20140.07-0.151.15AWES30.1003269.017-10.315116 May 201030 May 20140.41-0.62-3.76BKVL28.4738277.5463-6.5639213 Aug 200330 May 20140.20-0.061.12BNFY30.8484274.3962-0.6219118 May 200530 May 20140.17-0.11-0.43BVHS29.3368270.5936-15.740721 Aug 200230 May 2014-0.09-0.82-3.02CCV528.4602279.4548-24.23524 Jan 200730 May 2014-0.45-0.270.80CCV628.46279.4545-24.234624 Jan 200730 May 20140.120.00-1.20CHME35.2767279.1109225.252812 Feb 20029 Feb 2012-0.240.24-0.97COH630.0397264.8152-10.28561 Jan 200930 May 20140.791.52-8.10COVG30.4759269.9045-5.9404917 Jul 200430 May 2014-0.020.20-0.63CRST30.7261273.493832.5677526 Jul 200530 May 2014-0.21-0.54-0.12DLND29.0564278.7368-1.2700114 Jan 200530 May 20140.120.01-0.26DSTR29.9646269.6178-20.03888 Mar 200630 May 20140.23-0.35-1.74DUNN29.0622277.6291-7.8489325 Feb 200430 May 2014-0.38-0.041.54DWI129.0136264.5963-20.037322 May 200930 May 20140.92-0.33-4.63EXU023.564284.1266-20.057729 Jun 200730 May 20140.600.330.07FSHS29.8053268.4978-15.868713 May 201030 May 20141.22-3.27-3.33GABR34.8644275.6729515.50191 Jun 200530 May 20140.28-0.10-0.21GACC33.5458277.866298.503427 Nov 200330 May 2014-0.340.540.76GAGR33.2278275.7221264.08479 Mar 200515 Sep 2013-0.020.030.30GANW33.3058275.2326260.01248 Mar 200530 May 20140.24-0.08-0.09GRIS29.2655270.0427-17.04033 Sep 200530 May 2014-0.20-0.44-6.18HAC634.2808272.1441252.430817 Nov 200730 May 2014-0.540.620.71HAMM30.5131269.53245.81860113 Feb 200130 May 2014-0.250.39-1.04HOUM29.5923269.2764-12.724322 Nov 200330 May 2014-0.20-0.94-3.62INEG21.8562257.71581887.76221 Jul 199930 May 20140.771.72-25.77JCT130.4794260.1989571.823125 Oct 200512 Apr 2014-0.140.140.78JTNT33.0172259.0229684.508822 May 199722 Dec 2009-0.120.66-2.37JXVL30.484278.2985-19.220721 May 200230 May 2014-0.18-0.30-1.13KNS533.482280.657-20.2667 Nov 200730 May 20140.10-0.37-1.23KVTX27.546262.1071-2.6103920 Mar 200730 May 2014-0.67-0.12-2.05KWST24.5537278.2457-11.80047 Dec 200215 Sep 20130.420.44-0.82KYW524.5823278.347-13.627911 Oct 200730 May 20140.420.091.11KYW624.5823278.3472-13.661311 Oct 200730 May 20140.10-0.022.43LAUD26.1962279.8269-19.744212 Apr 200530 May 2014-0.19-0.14-1.77LMCN29.255269.3387-16.149323 Apr 200330 May 20140.66-0.67-6.30LSUA31.1788267.58774.70204521 Aug 200330 May 2014-0.070.25-0.07MCD527.8498277.4677-15.700623 Feb 200730 May 20140.34-0.32-0.21MCN532.6953276.439460.495872 Nov 200730 May 2014-0.12-0.04-0.09MCN632.6954276.439760.815822 Nov 200730 May 2014-0.220.21-1.59MERI20.98270.37977.8521011 Mar 200330 May 20140.490.64-0.38MLF632.0905272.607829.3724513 Jun 200730 May 2014-0.040.12-0.35MMD120.9319270.337228.6704126 Apr 200830 May 20140.76-0.24-0.77MMX119.4317260.93162235.24126 Apr 200830 May 20143.479.01-265.93
Continued.
StationLatLongHeightDate beginDate endVeastVnorthVupname(deg)(deg)(m)(mm yr-1)(mm yr-1)(mm yr-1)MSB534.1145269.309724.2128225 Sep 200730 May 2014-0.460.240.24MSCL33.7467269.266934.0347718 Aug 200630 May 2014-0.150.48-0.08MSGA30.3946271.3549-9.2740716 Dec 200730 May 20140.06-0.20-0.33MSPK30.7791270.856722.944717 Mar 200830 May 20140.02-0.09-0.45MSSC30.3752270.3861-13.09181 May 200530 May 20140.24-0.48-1.08MTNT25.8658279.093-20.54473 Jul 200330 May 2014-0.030.45-0.08NAPL26.1486278.2237-19.05463 Jul 200330 May 20140.320.38-0.84NAS025.0525282.5377-21.253626 Jun 200730 May 20141.340.20-0.32NBR535.1752282.95-24.675621 Mar 200730 May 20140.76-0.66-0.44NBR635.175282.9502-24.904121 Mar 200730 May 20140.11-0.59-0.23NC7735.1226279.0838156.19238 May 200330 May 20140.190.14-0.87NCAL35.3381279.7865111.26616 May 200430 May 20140.200.080.10NCCA35.3417280.6152148.241818 May 200430 May 20140.04-0.04-0.58NCCO35.3765279.4358153.02124 Jul 200330 May 20140.28-0.520.01NCKN34.9418282.019514.16821 Jul 200630 May 2014-0.21-0.130.14NCLI35.4202281.188824.327396 May 200430 May 2014-0.19-0.72-0.52NCLU34.6268280.922314.429851 Jul 200630 May 2014-0.08-0.04-0.39NCMR34.9819279.4762142.96328 May 200330 May 2014-0.180.49-1.45NCSL33.9826281.6099-11.419226 Feb 200830 May 2014-0.21-0.08-0.67NCWH34.2804281.2835-3.75323 Dec 200730 May 20140.03-0.53-0.87OAKH30.8155267.34320.227647 Apr 200430 May 2014-0.010.06-0.49OKAD34.8003263.2617292.12173 Sep 20028 May 2014-0.34-0.210.90OKAL34.6323260.6706400.489413 Jan 200530 May 2014-0.370.140.59OKAO35.0764261.7541339.326611 Dec 200430 May 20140.31-0.07-0.27OKCB27.266279.1447-15.33417 Dec 200230 May 2014-0.08-0.040.09OKCL35.4832261.0285469.597817 Dec 200230 May 2014-0.360.370.05OKDN34.4793262.0334314.32113 Jan 200530 May 20140.050.08-0.37OKHV34.9132265.3819145.32736 Aug 200330 May 2014-0.350.23-0.41OKLW34.5728261.5901313.57541 Sep 200230 May 20140.44-0.060.58OKMA34.928264.2628200.81219 Aug 200230 May 2014-0.260.830.00OKTE35.2602263.1022301.933919 Aug 200230 May 2014-0.47-0.721.19ORMD29.2982278.8911-19.88073 Apr 200330 May 20140.190.090.35PATT31.7783264.281594.8457123 May 199719 Oct 20090.00-0.190.26PBCH26.8463279.7807-16.889726 Feb 200530 May 20140.25-0.251.59PCLA30.469272.81061.94505911 Feb 200430 May 20140.06-0.04-0.77PLTK29.6634278.3123-19.478621 May 200230 May 2014-0.080.200.60PNCY30.2046274.3218-18.643722 May 200217 Aug 2010-0.26-0.040.03ROD130.0724264.473217.713261 Jan 200730 May 20140.911.98-15.95SA3734.7241273.3534180.143524 Nov 200426 Sep 2012-0.600.430.38SAL635.3673265.1837129.8417 May 200730 May 20140.360.42-0.34SAV532.1386278.30378.58499121 Oct 200730 May 2014-0.13-0.06-1.48SCCC32.7829280.062-11.381818 Mar 20059 Apr 2014-0.18-0.04-0.39SCGP34.9377277.7674278.03918 Mar 200530 May 2014-0.300.27-1.42SCUB20.0121284.237720.911346 Jan 200030 May 20143.260.150.79SCWT32.9034279.3316-3.384078 Feb 200630 May 20140.23-0.10-1.07SG0528.0652279.3772-10.81975 Feb 200217 Dec 20130.320.08-0.09SG3435.2691263.2598275.846519 Jun 200330 Nov 2009-0.28-0.440.79SHRV32.4277266.295436.7383818 Aug 200430 May 20140.270.37-0.73SNFD35.4735280.841994.0822414 Apr 200230 May 2014-0.27-0.01-0.55TXAB32.5033260.2432488.59892 Feb 200530 May 20140.270.25-0.08TXAM35.1536258.12151098.1218 Jan 199630 May 2014-0.220.140.66TXAU30.3117262.2437192.633818 Jan 199630 May 2014-0.14-0.010.58TXBS30.1129262.7094140.073610 May 200430 May 20140.31-0.140.26TXBU30.7505261.8156438.025110 May 200430 May 20140.12-0.060.13
Continued.
StationLatLongHeightDate beginDate endVeastVnorthVupname(deg)(deg)(m)(mm yr-1)(mm yr-1)(mm yr-1)TXBW31.7376261.0332389.3222 Feb 200512 Jul 2011-0.160.50-0.68TXBY30.6858263.629587.877612 Feb 200517 May 20120.430.53-1.98TXCH34.4596259.7217563.982125 Jan 200530 May 20140.69-0.370.08TXCN30.349264.558848.8176730 Jul 200530 May 2014-0.580.08-16.31TXDE33.2105262.8372178.782929 Oct 200330 May 20140.04-0.23-1.61TXDR29.3644259.1005280.143616 Jul 200430 May 20140.06-0.73-0.08TXGA29.3279265.2274-10.673730 Jul 200530 May 2014-0.01-0.12-4.36TXGR32.2404262.2456177.475823 Sep 200530 May 20140.09-0.15-0.27TXGV29.2851265.2107-17.439816 Feb 200720 Jul 20110.650.13-1.34TXHE30.099263.936547.4411630 Jul 200530 May 2014-0.91-0.26-6.59TXJA33.1948261.8544326.061923 Sep 200530 May 2014-1.270.273.42TXJC30.2665261.6027347.609310 May 200430 May 20140.310.17-0.13TXKE32.4097262.6768227.962623 Sep 200530 May 20140.120.20-1.62TXLF31.3564265.281777.4669330 Jul 200530 May 20140.160.000.43TXLI30.0559265.229-11.103530 Jul 200530 May 2014-0.35-0.27-0.44TXLL30.7335261.3214306.350730 Jul 200530 May 2014-0.130.33-0.22TXLM29.3922264.9763-17.50330 Jul 200530 May 2014-1.83-0.52-5.03TXLR27.5139260.5521113.623628 Jan 200230 May 2014-0.260.26-0.12TXLU33.5354258.1572956.039318 Jan 199630 May 20140.30-0.24-0.94TXMA32.5353265.711479.1209930 Jul 200530 May 20140.030.83-0.79TXNA32.0418263.4613105.455529 Oct 200330 May 2014-0.240.280.97TXPA33.6742264.443145.060330 Jul 200530 May 2014-0.040.18-0.46TXPR26.2085261.810715.1881428 Feb 200230 May 2014-0.73-2.33-5.52TXPS28.8888260.918175.357513 Oct 200730 May 20140.12-1.04-3.85TXPV28.6382263.3815-16.577516 Apr 201030 May 20140.70-0.43-1.86TXSG32.8557262.6558181.695923 Sep 200530 May 2014-0.530.060.09TXSM29.8779262.0973157.378211 May 200430 May 2014-0.310.351.37TXTY32.2496264.6064120.143830 Mar 200430 May 20140.090.25-1.45TXWA31.5777262.8895101.71082 Feb 200530 May 20140.61-0.62-1.61TXWE32.7589262.1765337.378823 Sep 200530 May 2014-0.290.35-0.05TXWF33.8539261.4944280.17475 Aug 200330 May 2014-0.190.120.10UNIP19.3127260.81872308.249 Dec 200519 Oct 20131.262.62-2.66UNPM20.8685273.1318-0.460518 Aug 200730 May 20140.23-0.26-0.02WACH27.5142278.11769.15797612 Apr 200530 May 2014-0.060.30-0.78WNFL31.8972267.218168.0091524 May 19976 Aug 20100.090.19-2.03XCTY29.631276.8918-15.307111 Feb 200430 May 20140.47-0.101.40ZEFR28.2276277.8354-1.508763 Sep 200330 May 20140.050.161.08ZFW132.8306262.9335155.194915 Jan 200330 May 20140.57-0.180.43ZHU129.9619264.668610.4186715 Jan 200330 May 20140.320.73-10.80ZJX130.6989278.09181.693641 Jul 200230 May 20140.12-0.150.07ZMA125.8246279.6808-8.0309311 Apr 200330 May 20140.12-0.010.14ZME135.0674270.044668.1613513 Mar 200330 May 2014-0.100.02-0.24ZTL433.3797275.7033260.69375 Nov 200230 May 2014-0.27-0.13-0.28
Locations and velocities of 13 reference stations used to establish SGOMRF.
StationLatLongHeightDate beginDate endVeastVnorthVupname(deg)(deg)(m)(mm yr-1)(mm yr-1)(mm yr-1)AL2034.71032-87.6627131.88746 Aug 200630 May 20140.16-0.230.15ARLR34.67263-92.382673.187759 Aug 200530 May 2014-0.020.04-0.25GAMC32.70182-83.648691.354328 Mar 20055 Dec 20130.050.11-0.10GNVL29.687-82.27722.4352121 May 200230 May 20140.01-0.110.44MTY225.71551-100.313521.738622 Sep 200530 May 20140.000.330.08OKAN34.1952-95.6214140.289119 Aug 200217 Jun 2014-0.260.09-0.19OKAR34.16846-97.1693235.76224 Dec 200430 May 2014-0.30-0.15-0.15RMND25.61378-80.3839-15.69923 Sep 200314 Jan 20140.10-0.07-0.17TAM122.27832-97.86421.042789 Jan 200530 May 20140.230.15-0.36TXDC33.23622-97.6087255.278423 Sep 200530 May 20140.34-0.140.19TXSA31.41432-100.473566.05682 Aug 200330 May 2014-0.300.25-0.08TXSN30.15258-102.409850.8730 Jul 200530 May 20140.12-0.140.17TXST32.23259-98.1822376.575423 Sep 200530 May 2014-0.18-0.170.09Acknowledgements
The authors thank NGS for providing GPS data to the public. Some data are
provided by the UNAVCO Facility with support from the National Science
Foundation (NSF) and National Aeronautics and Space Administration (NASA)
under NSF Cooperative Agreement No. EAR-0735156. This study was supported by
an NSF CAREER award EAR-1229278, an NSF MRI award EAR-1242383, and an NSF TUES
award DUE-1243582.
Edited by: N. Kerle
Reviewed by: T. Törnqvist and two anonymous referees
ReferencesAllis, R. G.: Review of subsidence at Wairakei field, New Zealand, Geothermics,
29, 455–478, 10.1016/S0375-6505(00)00016-X, 2000.Bar-Sever, Y. E., Kroger, P. M., and Borjesson, J. A.: Estimating horizontal
gradients of tropospheric path delay with a single GPS receiver, J. Geophys.
Res., 103, 5019, 10.1029/97JB03534, 1998.
Baumann, T., Goree, B., Lovelace, W., Montgomery, P., Ross, G., Walters, D.,
and Ward, A.: Water Resources Data for Louisiana, Water Year 2005,
Water-Data Report LA-05-01, US Geological Survey, Baton Rouge, LA, 2006.Bawden, G. W., Thatcher, W., Stein, R. S., Hudnut, K. W., and Peltzer, G.:
Tectonic contraction across Los Angeles after removal of groundwater pumping
effects, Nature, 412, 812–815, 10.1038/35090558, 2001.
Bawden, G. W., Johnson, M. R., and Kasmarek, M. C.: Investigation of land
subsidence in the Houston-Galveston region of Texas by using the Global
Positioning System and Interferometric Synthetic Aperture Radar, 1993–2000,
USGS Scientific Investigation Report 2012-5211, US Geological Survey, Reston, VA, 2012.
Beckman, J. D. and Williamson, A. K.: Salt-dome locations in the Gulf
Coastal Plain, South-Central United States, Water-Resources Investigations
Report 90-4060, US Geological Survey, Austin, TX, 1990.Bertiger, W., Desai, S. D., Haines, B., Harvey, N., Moore, A. W., Owen, S.,
and Weiss, J. P.: Single receiver phase ambiguity resolution with GPS data,
J. Geodyn., 84, 327–337, 10.1007/s00190-010-0371-9, 2010.Blewitt, G. and Lavallée, D.: Effect of annual signals on geodetic velocity,
J. Geophys. Res.-Sol. Ea., 107, ETG9-1–ETG 9-11, 10.1029/2001JB000570, 2002.Blewitt, G., Kreemer, C., Hammond, W. C., and Goldfarb, J. M.: Terrestrial
reference frame NA12 for crustal deformation studies in North America, J.
Geodyn., 72, 11–24, 10.1016/j.jog.2013.08.004, 2013.Blum, M. D., Tomkin, J. H., Purcell, A., and Lancaster, R. R.: Ups and downs
of the Mississippi Delta, Geology, 36, 675–678, 10.1130/G24728A.1, 2008.Boehm, J., Niell, A., Tregoning, P., and Schuh, H.: Global Mapping Function (GMF):
A new empirical mapping function based on numerical weather model data, Geophys.
Res. Lett., 33, L07304, 10.1029/2005GL025546, 2006.Cabral-Cano, E., Dixon, T. H., Miralles-Wilhelm, F., Diaz-Molina, O.,
Sanchez-Zamora, O., and Carande, R. E.: Space geodetic imaging of rapid ground
subsidence in Mexico City, Geol. Soc. Am. Bull., 120, 1556–1566, 10.1130/B26001.1, 2008.Canny, J.: A Computational Approach to Edge Detection, IEEE T. Pattern Anal.
Mach. Intell., PAMI-8, 679–698, 10.1109/TPAMI.1986.4767851, 1986.
Carrillo, N.: Influence of Artesian Wells in the Sinking of Mexico City, in:
Comision Impulsora y Coordinadora de la Investigacion Cientifica, vol. 47,
Secretaria de Hacienda y Credito Publico, Mexico, DF, 7–14, 1947.Chang, C., Mallman, E., and Zoback, M.: Time-dependent subsidence associated
with drainage-induced compaction in Gulf of Mexico shales bounding a severely
depleted gas reservoir, Am. Assoc. Petrol. Geol. Bull., 98, 1145–1159, 10.1306/11111313009, 2014.Chaussard, E., Wdowinski, S., Cabral-Cano, E. and Amelung, F.: Land subsidence
in central Mexico detected by ALOS InSAR time-series, Remote Sens. Environ.,
140, 94–106, 10.1016/j.rse.2013.08.038, 2014.Collilieux, X., Métivier, L., Altamimi, Z., Dam, T., and Ray, J.: Quality
assessment of GPS reprocessed terrestrial reference frame, GPS Solut., 15,
219–231, 10.1007/s10291-010-0184-6, 2010.
Cuevas, J.: Foundation Conditions in Mexico City, in: Proceedings of International
Conference on Soil Mechanics, International Society for Soil Mechanics and
Geotechnical Engineering, Cambridge, Mass., 233–237, 1936.Day, J. W., Pont, D., Hensel, P. F., Ibañez, C., and Ibanez, C.: Impacts
of Sea-Level Rise on Deltas in the Gulf of Mexico and the Mediterranean: The
Importance of Pulsing Events to Sustainability, Estuaries, 18, 636–647,
10.2307/1352382, 1995.Dixon, T. H., Amelung, F., Ferretti, A., Novali, F., Rocca, F., Dokka, R.
K., Sella, G., Kim, S.-W., Wdowinski, S., and Whitman, D.: Space geodesy:
subsidence and flooding in New Orleans, Nature, 441, 587–588, 10.1038/441587a, 2006.Dokka, R. K.: Modern-day tectonic subsidence in coastal Louisiana, Geology,
34, 281–284, 10.1130/G22264.1, 2006.Dokka, R. K.: The role of deep processes in late 20th century subsidence of
New Orleans and coastal areas of southern Louisiana and Mississippi, J. Geophys.
Res., 116, B06403, 10.1029/2010JB008008, 2011.Dokka, R. K., Sella, G. F., and Dixon, T. H.: Tectonic control of subsidence
and southward displacement of southeast Louisiana with respect to stable North
America, Geophys. Res. Lett., 33, L23308, 10.1029/2006GL027250, 2006.
Edrington, C. H., Blum, M. D., Nunn, J. A., and Hanor, J. S.: Long-Term
Subsidence and Compaction Rates: A New Model for the Michoud Area, South
Louisiana, GCAGS Trans., 58, 261–272, 2008.
Engel, K., Jokiel, D., Kraljevic, A., and Geiger, M.: Big cities. Big water.
Big Challenges. Water in an urbanizing world, WWF Germany, Berlin, 2011.
Engelkemeir, R., Khan, S. D., and Burke, K.: Surface deformation in Houston,
Texas using GPS, Tectonophysics, 490, 47–54, 2010.Firuzabadì, D. and King, R. W.: GPS precision as a function of session
duration and reference frame using multi-point software, GPS Solut., 16,
191–196, 10.1007/s10291-011-0218-8, 2011.Free ocean tide loading provider: http://holt.oso.chalmers.se/loading/index.html,
last access: 15 January 2015.
Galloway, D., Jones, D., and Ingebritsen, S.: Land subsidence in the United
States, USGS Circular 1182, US Geological Survey, Reston, VA, 1999.
Garrity, C. and Soller, D.: Database of the Geologic Map of North America;
adapted from the map by J. C. Reed Jr. and others (2005), US Geological
Survey Data Series 424, Reston, VA, 2009.
Harris-Galveston Subsidence District: Harris-Galveston Subsidence District
Regulatory Plan 2013, adopted 9 January 2013, amended 8 May 2013,
Harris-Galveston Subsidence District, Friendswood, TX, 2013.Ivins, E. R., Dokka, R. K., and Blom, R. G.: Post-glacial sediment load and
subsidence in coastal Louisiana, Geophys. Res. Lett., 34, L16303, 10.1029/2007GL030003, 2007.Jackson, M. P. A. and Seni, S. J.: Geometry and evolution of salt structures
in a marginal rift basin of the Gulf of Mexico, east Texas, Geology, 11,
131–135, 10.1130/0091-7613(1983)11<131:GAEOSS>2.0.CO;2, 1983.
Johnson, M. R., Ramage, J. K., and Kasmarek, M. C.: Water-level altitudes 2011
and water-level changes in the Chicot, Evangeline, and Jasper aquifers and
compaction 1973–2010 in the Chicot and Evangeline aquifers, Houston–Galveston
region, Texas, USGS Sci. Invest. Map 3174, US Geological Survey, Reston, VA, 2011.
Joint Academies Committee on the Mexico City Water Supply, Environment
and Resources Commission on Geosciences, National Research Council,
A. C. Academia Nacional de la Investigacion Cientifica and A. C. Academia Nacional
de Ingenieria: Mexico City's Water Supply: Improving the Outlook for
Sustainability, The National Academies Press, Washington, D.C., 1995.Jurkowski, G., Ni, J., and Brown, L.: Modern uparching of the Gulf Coastal
Plain, J. Geophys. Res., 89, 6247–6255, 10.1029/JB089iB07p06247, 1984.Karegar, M. A., Dixon, T. H. and Malservisi, R.: A three-dimensional surface
velocity field for the Mississippi Delta: Implications for coastal restoration
and flood potential, Geology, 43, 519–522, 10.1130/G36598.1, 2015.
Kasmarek, M. C., Gabrysch, R. K., and Johnson, M. R.: Estimated land-surface
subsidence in Harris County, Texas, 1915–17 to 2001, USGS Sci. Invest. Map 3097,
US Geological Survey, Reston, VA, 2009.
Kasmarek, M. C., Johnson, M. R., and Ramage, J. K.: Water-level altitudes 2010
and water-level changes in the Chicot, Evangeline, and Jasper aquifers and
compaction 1973–2009 in the Chicot and Evangeline aquifers, Houston–Galveston
region, Texas, USGS Sci. Invest. Map 3138, US Geological Survey, Reston, VA, 2010.
Kasmarek, M. C., Johnson, M. R., and Ramage, J. K.: Water-level altitudes 2012
and water-level changes in the Chicot, Evangeline, and Jasper aquifers and
compaction 1973–2011 in the Chicot and Evangeline aquifers, Houston–Galveston
region, Texas, USGS Sci. Invest. Map 3230, US Geological Survey, Reston, VA, 2012.
Kasmarek, M. C., Johnson, M. R. and Ramage, J. K.: Water-level altitudes 2013
and water-level changes in the Chicot, Evangeline, and Jasper aquifers and
compaction 1973–2012 in the Chicot and Evangeline aquifers, Houston-Galveston
region, Texas, USGS Sci. Invest. Map 3263, US Geological Survey, Reston, VA, 2013.Kearns, T. J., Wang, G., Bao, Y., Jiang, J., and Lee, D.: Current Land
Subsidence and Groundwater Level Changes in the Houston Metropolitan Area
(2005–2012), J. Surv. Eng., 141, 05015002, 10.1061/(ASCE)SU.1943-5428.0000147, 2015.Kedar, S.: The effect of the second order GPS ionospheric correction on receiver
positions, Geophys. Res. Lett., 30, 1829, 10.1029/2003GL017639, 2003.Khan, S. D., Huang, Z., and Karacay, A.: Study of ground subsidence in northwest
Harris county using GPS, LiDAR, and InSAR techniques, Nat. Hazards, 73,
1143–1173, 10.1007/s11069-014-1067-x, 2014.Kolker, A. S., Allison, M. A., and Hameed, S.: An evaluation of subsidence
rates and sea-level variability in the northern Gulf of Mexico, Geophys.
Res. Lett., 38, L21404, 10.1029/2011GL049458, 2011.Lyard, F., Lefevre, F., Letellier, T., and Francis, O.: Modelling the global
ocean tides: modern insights from FES2004, Ocean Dynam., 56, 394–415,
10.1007/s10236-006-0086-x, 2006.Meckel, T. A.: An attempt to reconcile subsidence rates determined from various
techniques in southern Louisiana, Quaternary Sci. Rev., 27, 1517–1522,
10.1016/j.quascirev.2008.04.013, 2008.Morton, R. A. and Bernier, J. C.: Recent Subsidence-Rate Reductions in the
Mississippi Delta and Their Geological Implications, J. Coast. Res., 26,
555–561, 10.2112/JCOASTRES-D-09-00014R1.1, 2010.Osmanoglu, B., Dixon, T. H., Wdowinski, S., Cabral-Cano, E., and Jiang, Y.:
Mexico City subsidence observed with persistent scatterer InSAR, Int. J.
Appl. Earth Obs. Geoinf., 13, 1–12, 10.1016/j.jag.2010.05.009, 2011.Pearson, C. and Snay, R.: Introducing HTDP 3.1 to transform coordinates across
time and spatial reference frames, GPS Solut., 17, 1–15, 10.1007/s10291-012-0255-y, 2012.Qu, F., Lu, Z., Zhang, Q., Bawden, G. W., Kim, J.-W., Zhao, C., and Qu, W.:
Mapping ground deformation over Houston–Galveston, Texas using multi-temporal
InSAR, Remote Sens. Environ., 169, 290–306, 10.1016/j.rse.2015.08.027, 2015.
Ramsey, K. E. and Moslow, T. F.: A Numerical Analysis of Subsidence and Sea
Level Rise in Louisiana, in Proceedings of Coastal Sediments '87, American
Society of Civil Engineers, New Orlean, LA, 1673–1688, 1987.Rebischung, P., Griffiths, J., Ray, J., Schmid, R., Collilieux, X., and Garayt,
B.: IGS08: the IGS realization of ITRF2008, GPS Solut., 16, 483–494,
10.1007/s10291-011-0248-2, 2012.
Roberts, H. H., Bailey, A., and Kuecher, G. J.: Subsidence in the Mississippi
River Delta; important influences of valley filling by cyclic deposition,
primary consolidation phenomena, and early diagenesis, Trans. Gulf Coast
Assoc. Geol. Soc., 44, 619–629, 1994.
Schwarz, C.: The North American Datum of 1983, NOAA Professional Paper NOS 2,
National Geodetic Information Branch, NOAA, Rockville, Maryland, 1989.
Simms, A. R., Anderson, J. B., DeWitt, R., Lambeck, K., and Purcell, A.:
Quantifying rates of coastal subsidence since the last interglacial and the
role of sediment loading, Global Planet. Change, 111, 296–308, 2013.
Snay, R. and Soler, T.: Modern Terrestrial Reference Systems. Part 2: The
evolution of the NAD83, Prof. Surv., 20, 16–18, 2000.Sosa-Rodriguez, F.: Exploring the risks of ineffective water supply and sewage
disposal: A case study of Mexico City, Environ. Hazards, 9, 135–146,
10.3763/ehaz.2010.0016, 2010.Thatcher, C. A., Brock, J. C., and Pendleton, E. A.: Economic Vulnerability
to Sea-Level Rise Along the Northern U.S. Gulf Coast, J. Coast. Res., 63,
234–243, 10.2112/SI63-017.1, 2013.Törnqvist, T. E., Wallace, D. J., Storms, J. E. A., Wallinga, J., van
Dam, R. L., Blaauw, M., Derksen, M. S., Klerks, C. J. W., Meijneken, C., and
Snijders, E. M. A.: Mississippi Delta subsidence primarily caused by compaction
of Holocene strata, Nat. Geosci., 1, 173–176, 10.1038/ngeo129, 2008.Wang, G.: GPS Landslide Monitoring: Single Base vs. Network Solutions – A
case study based on the Puerto Rico and Virgin Islands Permanent GPS Network,
J. Geod. Sci., 1, 191–203, 10.2478/v10156-010-0022-3, 2011.Wang, G. and Soler, T.: Using OPUS for Measuring Vertical Displacements in Houston,
TX, J. Surv. Eng., 139, 126–134, 10.1061/(ASCE)SU.1943-5428.0000103, 2013.Wang, G. and Soler, T.: Measuring Land Subsidence Using GPS: Ellipsoid Height
versus Orthometric Height, J. Surv. Eng., 14, 05014004-1–05014004-12,
10.1061/(ASCE)SU.1943-5428.0000137, 2014.Wang, G., Yu, J., Ortega, J., Saenz, G., Burrough, T., and Neill, R.: A Stable
Reference Frame for the Study of Ground Deformation in the Houston Metropolitan
Area, Texas, J. Geod. Sci., 3, 188–202, 10.2478/jogs-2013-0021, 2013.Wang, G., Kearns, T. J., Yu, J., and Saenz, G.: A stable reference frame for
landslide monitoring using GPS in the Puerto Rico and Virgin Islands region,
Landslides, 11, 119–129, 10.1007/s10346-013-0428-y, 2014.Wang, G., Bao, Y., Cuddus, Y., Jia, X., Serna, J., and Jing, Q.: A methodology
to derive precise landslide displacement time series from continuous GPS
observations in tectonically active and cold regions: a case study in Alaska,
Nat. Hazards, 77, 1939–1961, 10.1007/s11069-015-1684-z, 2015a.
Wang, G., Welch, J., Kearns, T. J., Yang, L., and Serna, J.: Introduction to
GPS geodetic infrastructure for land subsidence monitoring in Houston, Texas,
U.S.A., in: Ninth International Symposium on Land Subsidence, UNESCO Working
Group on Land Subsidence, Nagoya, Japan, 2015b.
Williams, S., Stone, G., and Burruss, A.: A perspective on the Louisiana wetland
loss and coastal erosion problem, J. Coast. Res., 13, 593–594, 1997.Wolstencroft, M., Shen, Z., Törnqvist, T. E., Milne, G. A., and Kulp, M.:
Understanding subsidence in the Mississippi Delta region due to sediment,
ice, and ocean loading: Insights from geophysical modeling, J. Geophys. Res.-Solid
Ea., 119, 3838–3856, 10.1002/2013JB010928, 2014.Yu, J., Wang, G., Kearns, T. J., and Yang, L.: Is There Deep-Seated Subsidence
in the Houston-Galveston Area?, Int. J. Geophys., 2014, 1–11, 10.1155/2014/942834, 2014.Yu, S.-Y., Törnqvist, T. E., and Hu, P.: Quantifying Holocene lithospheric
subsidence rates underneath the Mississippi Delta, Earth Planet. Sc. Lett.,
331–332, 21–30, 10.1016/j.epsl.2012.02.021, 2012.
Zilkoski, D., Hall, L., Mitchell, G., Kammula, V., Singh, A., Chrismer, W.,
and Neighbors, R.: The Harris-Galveston Coastal Subsidence District/National
Geodetic Survey automated Global Positioning System subsidence monitoring
Project, in: Proceedings of the US Geological Survey Subsidence Interest Group
Conference, US Geological Survey, Galveston, TX, 13–28, 2003.Zumberge, J. F., Heflin, M. B., Jefferson, D. C., Watkins, M. M., and Webb,
F. H.: Precise point positioning for the efficient and robust analysis of
GPS data from large networks, J. Geophys. Res., 102, 5005–5017, 10.1029/96JB03860, 1997.