Keywords

1 Introduction

High frequency reflection seismic data has been used for researching formations of young river sediment, lowland, and seabed visualization [1,2,3]. Seismic waves emitted from a sub-bottom profiler can propagate through its surrounding environment and reflect to its receiver after hitting seismic boundaries of different acoustic impedance zones [4]. With its smaller energy rather than other explosion traditional sources (i.e., air guns), the interest depth of this high-resolution seismic method can only go to hundreds of meters below the water surface. Data processing/analyzing stages can convert the raw measured data into interpretable data [5].

Our work focusses on collecting, processing, and interpreting ten 2D high-resolution seismic profiles to image seabed and young sediment at the fork area of Soai Rap River, Can Gio coast, Ho Chi Minh City, Vietnam. We have applied 3D visualization of their processed data in supporting analyzing stages. Interpreted seismic results are supported with information of two nearby drill holes. The drill holes’ locations are in Thanh An island, Can Gio District [4].

2 Study Area

The acquisition area is the fork of Soai Rap River, Can Gio district, Ho Chi Minh city, Vietnam shown in Fig. 1. Known as a coastal area, Can Gio district plays a vital role in the economic development of the city when it has many important rivers connecting to the outside world. Its Soai Rap river is the second maritime route (the other is Long Tau river) for domestic and international large ships to transport imported and imported products to seaports in Ho Chi Minh city [6].

Fig. 1
A map displays three views of the Soai Rap River between Tien Giang and Can Gio in Vietnam with hand-drawn lines that depict T1, T3, T5, T7, T43, T45, T16, T17, T23, and T5M.

Location of the survey area in Soai Rap river, Ho Chi Minh City, Vietnam [7]

Full size image

Recognized by UNESCO in 2000 as the first Mangrove Biosphere Reserve in Vietnam for having the highest diversity of mangrove living species, Can Gio also plays as “green lungs” of the city. Can Gio formed around 7000–6000 cal BP [8, 9] thanks to the Saigon-Dong Nai river basin. The factors such as semi-diurnal regime tidal from the estuary and average water discharge from Soai Rap and Long Tau rivers in the rainy season can form sand bars locating in front of their estuaries and Can Gio foreshore [4, 10].

Soai Rap river, an outlet of the Sai Gon–Dong Nai river, received a much smaller discharge of sediment compared to the Mekong River basin [11]. It is also close to the Mekong River delta. Therefore, knowledge of sediment of research area can be beneficial from geology of the Mekong delta basin and Ho Chi Minh city [4, 10, 12,13,14,15,16,17]. In Ho Chi Minh City, its northern part gets the higher elevation which are occupied by the pre–Holocene Cenozoic materials [18, 19]. Holocene sediment formations are seen in its southern areas as Nha Be and Can Gio districts [18, 19]. Investigations show that Holocene sediment patterns are linked to Sai Gon–Dong Nai river systems [18, 19]. Different acoustics impedances of types of young sediments can inspire application of high-resolution seismic method to image seabed and lowland topography [4, 5].

3 Methodology

High-resolution seismic method is applied in searching subsurface image of maritime areas and young sediment structures thanks to wave reflections from boundaries of different acoustic impedance layers [2, 4, 20, 21].

In our research, the sub-bottom profiler, namely, SB-216S [22], typed with fixed transmitter–receiver offset is carried in a ship for data collection. Being laid 1.5 m underwater, it emits signals of the frequency range from 2 to 16 kHz in each 20 ms length which propagate through water and sediment environments [22]. During data collection stage, the ship can be detected by the GPS locator equipment while the professional software, Edge Tech Discover [23], helps to visualize the reflection seismic waves seen as electric signals in the computer monitor. In 2017, ten 2D profiles were collected in the research area, ranging from 5500 m to around 12000 m. In Fig. 2, one example shows that how the equipment is laid in the ship and seismic waves propagate through surrounding environment.

Fig. 2
An illustration of data collection using a sub-bottom profiler immersed in seawater that transmits signals to the two layers of seafloor separated by lines.

Deployment of the sub-bottom profiler (SBS-216S) [4, 22, 24]. Boundary of two different acoustic impedance layers (i.e., water and sand) can be interpreted as green lines

Full size image

Data formats depend on data analysis/processing stages. For data collection, its seismic data format is binary with the extension *.jsf [25]. For processing/ analyzing the uninterpretable raw data to interpretable ones, we have used three professional software, Reflexw [26], OpendTect [27], and MATLAB codes [4].

3.1 Data Processing

Interpretable seismic data can be achieved via processing stages (Fig. 3). The first stage focuses on using processing tools in Reflexw for having interpretable seismic data in 2D view while the other stage utilizes the application of OpendTect for interpreting all the 2D data in 3D visualization.

Fig. 3
An illustration of raw data, step 1 subtract D C shift, and step 2 divide compensation from top to bottom with dimensions 5600 meters by 50 meters.

Processing steps for the raw seismic data (top image) in the profile T1 [4]. Suitable amplitude gains in Divergent Compensation applied to the Step 1 processed data (middle image) can make the Step 2 processed data (bottom image) visible in the larger travel time

Full size image

In the first stage, we applied two steps as Substract-DC-shift and Divergent (div.) Compensation Gain for all the ten seismic lines [4, 26]. A time constant shift is removed for each trace in Substract-DC-Shift filter. The Compensation Gain filter plays a vital role in reimbursing spherical energy loss while the waves propagate deeper [26].

Figure 3 illustrates one example of our first seismic processing stage for the profile T1. In the deeper part, a visible amplitude difference can be seen between two images, the raw data (top image) and the final processed image (bottom image).

In the second stage, with the help of OpendTech software, three main works are listed:

  1. (i)

    3D visualization of all the 2D processed seismic lines resulted from the first stage (Fig. 4)

    Fig. 4
    A three-dimensional view of Soai Rap river between Tien Giang and Can Gio with a grid-like structure in the river surrounded by cross lines.

    Survey setting for the area. All the ten seismic profiles are put in 3D visualization. They will be input for further seismic attributes calculation

    Full size image
  2. (ii)

    validating the data quality (Fig. 5)

    Fig. 5
    A three-dimensional view of a grid-like structure with cross lines and arrows that point eastward along all parts of the surface of the grid in the eastern direction.

    Validity check of seismic data in depth analysis of the meeting points in their seismic profiles. In the meeting points, their depths of any two cross profiles should be similar. Two images (top and bottom) show the checking in different 3D view. Blue arrows refer locations of meeting points

    Full size image
  3. (iii)

    computation of their different seismic attributes, 3D interpolated horizons, and the Holocene layer thickness.

Coordinates of all the seismic data profiles are used to form survey settings in OpendTect platform (Fig. 4). Validating the seismic quality could be done using the technique of checking a meeting point of any two different seismic profiles [4]. In Fig. 5, all the meeting points defined from couples of two cross different seismic lines are illustrated. That is, their similar depths of their meeting points can prove high reliability of the seismic quality in terms of data measurement and analysis.

Seismic Textural Attributes: Reflectors can be recognized from extreme values of processed seismic data and its seismic attributes [4, 5, 28, 29]. We have used application of entropy and energy textural attributes to help interpretation of the main seismic reflection.

Definition of textural attributes as energy, entropy, contrast, and homogeneity are well-known relating to detecting “zones of common signal character” [30]. The gray-level co-occurrence Matrix (GLCM) [31, 32] is established from seismic processed input data, then seismic textural attribute output such as energy or entropy are calculated from the GLCM data by the equations below[5]:

$${\text{Energy}} = \sqrt {\mathop \sum \limits_{i,j = 0}^{N - 1} P_{i,j}^{2} }$$
(1)
$${\text{Entropy}} = \mathop \sum \limits_{i,j = 0}^{N - 1} P_{i,j} \left( { - {\text{ln}}\left( {P_{i,j} } \right)} \right)$$
(2)

where Pi,j, shows the ith row and jth column of the GLCM matrix P.

The energy one brightens areas of high textural stability in each calculation zone which can be continuous features. The term, “Entropy” express how large randomness in the calculation zone is.

A calculation zone, namely a sliding window is characteristic through horizontal width and time length (see Fig. 6). For our research, we use time length as 2 × 27 ms and horizontal width as 2 × 9 traces.

Fig. 6
An illustration depicts time along the vertical axis against a horizontal stepout and displays a timegate within a dashed rectangle.

Representation of 2D sliding window for computing seismic textural attributes [4, 27]. The sliding window is seen as a dashed rectangular characterizing horizontal width and time length

Full size image

4 Results and Discussion

We have interpreted some reflectors in the interest area. Using distinguished waveforms of processed seismic data and its seismic textural attributes (i.e., energy and entropy), three manually extracted boundaries can form three shallow layers as (i) water body, (ii) mixture of mud, sand, and silk clays layers (Present), (iii) sand silk clay sediments (Holocene), and hard gray sediment (Pleistocene).

The water body is easily recognized from strong seismic data and extreme values of its textural seismic attributes (see Fig. 7). In the Fig. 7, processed seismic data and its seismic attributes are shown for providing different view of a geology object. For example, water body can be seen as the space bounded by the surface to strong reflection boundary (see wheat color line in Fig. 7) while zone of low value for energy texture and zone of high value for entropy are represented for the water body.

Fig. 7
A set of 3, three-dimensional grid-like structures with cross lines. The shades on the top surface depict amplitude, energy, and entropy.

Processed seismic data (top image) and its texture attributes, energy (middle image) and entropy (below image). Wheat, green, and pink lines are interpreted as seabed, top, and bottom of Holocene layer, respectively

Full size image

Thanks to the seismic data attributes, we have interpreted different 2D seismic boundaries to form the three shallow layers. For illustrating all the 2D boundaries in 3D visualization, Fig. 8 refers seabed, top and bottom of Holocene layer as green, wheat, and pink colors, respectively. The well match in depth between of any two intersect lines can prove high correctness in seismic interpretation.

Fig. 8
An illustration of a three-dimensional vertically sliced layer with different shades surrounded by cross lines, a perpendicular sliced layer, and lines in different shades that represent entropy and amplitude.

Representation of all 2D interpreted seismic boundary. Processed seismic data and its textural attribute, entropy. Green, wheat color, and pink lines are interpreted as seabed, top and bottom of Holocene layer, respectively

Full size image

3D surfaces are interpolated from the 2D seismic boundaries (Figs. 9 and 10). We have used the Matlab built-in function, scatteredInterpolant.m [33] to build up the surfaces. 3D seabed is firstly interpreted and interpolated from strong visible reflection seismic amplitude events (Fig. 9). For the Top and Bottom interfaces of the Holocene layer, it is not consistent for strong reflection to interpret. Then, we used the visible patterns of the Holocene layer in textural seismic attributes (see Figs. 7 and 8). The boundaries between Holocene and Pleistocene layers are hardly seen through the processed seismic data but visible in their attributes helping its interpretation in different seismic profiles.

Fig. 9
An illustration of a three-dimensional grid-like structure with cross lines in different shades that represent entropy, amplitude, and energy.

3D seabed and Bottom Holocene made by 10 seismic profiles. Linear interpolation image with green surface for Seabed and pink one for Bottom Holocene

Full size image
Fig. 10
An illustration of four, three-dimensional figures that represent seabed depth between Tien Giang and Can Gio, top depth, bottom depth, and thickness of holocene.

3D seabed, top, bottom and thickness of Holocene layer made by ten seismic profiles. Holocene layer thickness can range from 1.5 m to around 4.5 m

Full size image

Thickness of the Holocene layer can be calculated from the zone bounded by two 3D top and bottom surfaces. We have used propagation velocities as 1500 m/s and 1550 m/s for sea water and underground sediments, respectively [2] to convert two-way travel time (TWT) to depth. Note that multiple noise can interfere interpretation of the two 3D horizons. We have recognized seismic multiples by checking their travel time as double or triple of the original seismic events (See Fig. 7).

In the Fig. 10, 3D surfaces of seabed, the top, bottom, and thickness of Holocene layer are illustrated. There is distinguished river channel flowing northeast. The channel looks like to divide the interest area in to three different sub-areas with different depths in seabed, or top and bottom of Holocene layer. Figure 11 also shows a clear 3D visualization of the channel as seen as the red stripe.

Fig. 11
An illustration of a three-dimensional grid-like structure in the Soai Rap river channel in various shades.

3D view of small the Soai Rap river channel

Full size image

According to the drill holes information [4] and the seismic interpretation result, we can determine three layers as follow:

  1. (i)

    the first layer having mud, silk clay, sand, organic matter from Present time,

  2. (ii)

    the Holocene layer having clay sediment, and

  3. (iii)

    the hard clay Pleistocene matters.

5 Conclusion

3D image of seabed and sedimentation in Soai Rap River fork can be presented using interpretation of 2D high-resolution seismic data. Suitable interpolation of different 2D boundaries which express zones of contrast of acoustic impedance can bring the 3D visualization come true with the help of useful professional software Matlab and OpendTect. Better interpreting the boundaries needs combination of different textural seismic attributes rather than just utilizing only the seismic processed data itself. Great match between seismic results and drill hole information as lithology can prove existence of Holocene layer and Pleistocene matters which helps to government officials in deepening the riverbed and canals for large ship transportation, especially, in Can Gio, Ho Chi Minh city. The distinguished Soai Rap channel appears in the seismic data can prove the fact that the high-resolution seismic data can detect small but meaningful geology features (Fig. 12).

Fig. 12
An illustration of the seabed line, top of Holocene layer, and bottom of Holocene layer in drill 1, 40 meters, and drill 2, 31 meters with mud, silk clay, and organic matter present. The depth of nearby seismic data displays data.

Representations of two drill holes [4] with their two nearby seismic data profiles

Full size image