Simplified Engineering Geomorphic Unit-based Seismic Site Characterization of Detailed Area Plan of Dhaka City

doi:10.21203/rs.3.rs-2220228/v1

Download PDF

Article

Simplified Engineering Geomorphic Unit-based Seismic Site Characterization of Detailed Area Plan of Dhaka City

https://doi.org/10.21203/rs.3.rs-2220228/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 09 Jul, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

Strong ground motion can amplify and prolong due to local site conditions; consequently, it can cause severe damage in poor non-engineered structures. Usually, site response analysis is used to predict strong motion characteristics at the surface. In this study, one-dimensional nonlinear site response analysis is performed based on the prepared simplified engineering geomorphic unit of the Detailed Area Plan (DAP) of Dhaka City. The engineering geomorphic unit-based map was prepared from image analysis and verified with the collected boreholes data and surface geology map. For response analysis, the earthquake time histories are selected from the PEER NGA WEST2 data set, and BNBC 2020 uniform hazard spectrum (UHS) is considered as the target spectrum. In addition, two synthetic time histories are also generated and used for the site response analysis. For the selected earthquake ground motions, the near-surface soil response of the DAP of the Dhaka City area shows de-amplification of acceleration in the short period and amplified acceleration in the long period. The amplified long-period acceleration can cause severe damage in non-engineered long-period structures. The outcome of this study can be used to prepare a seismic risk-sensitive land use plan for the future development of the DAP area of Dhaka City.

Earth and environmental sciences/Natural hazards

Physical sciences/Engineering/Civil engineering

Bangladesh, a low-lying deltaic country, is one of the most vulnerable countries to natural and anthropogenic hazards due to its geographical location, topography, monsoon climate, and dense population. It experienced over 200 disasters between 1980 to 2018¹. Floods, cyclones, river bank erosion, storm surge, drought, salinity intrusion, fire, and earthquake are common hazards for this country. Even though floods and cyclones regularly occur in Bangladesh, the Himalayan system in the north and the Arakan subduction-collision system in the east are the two major tectonic systems that can produce large-magnitude earthquakes in and around Bangladesh². The history of earthquakes in the recent past and the present rate of tectonic plate movement suggest that Bangladesh and its surrounding area are seismically active³. However, seismic vulnerability mainly depends on strong ground motion and the quality of the built-in environment. The strong ground motion characteristics of engineering bedrock (shear wave velocity (Vs) > 760 m/s) can be significantly amplified or de-amplified due to local site conditions, i.e., subsurface soil conditions. The earthquakes of 1995 Kobe, 1989 Loma Prieta, 1985 Mexico, 1964 Alaska, and 1964 Niigata earthquake witnessed the damage associated with effects of site conditions^4,5. Therefore, it is a prerequisite to analyze the site response to know the local site effects for performance-based seismic design to make a seismic resilient society.

Site response analysis aims to estimate the earthquake ground motions by determining how the local soil properties of any site influence the expected rock motions⁶. The most simplified site response analysis procedure is based on the site amplification factor calculated using the average of top 30 m soil (Vs30) based site amplification factor^7,8. However, this simplified procedure is unsuitable when the engineering bedrock depth is more than 30 m⁹. Therefore, to get an acceptable ground response, the use of the dynamic properties of the soil column along with bedrock ground motion as input is preferred^10,11,12,13. Researchers performed 1D site response analysis with linear, equivalent-linear, and nonlinear models and obtained the acceptable surface ground motion^14,15. Since soil behavior is nonlinear, some adjustments to the linear technique are required for estimating ground response. The equivalent-linear approach is an improved site response strategy that provides better results than the linear approach. A computer-based program termed "SHAKE" for equivalent-linear approximation demonstrated the actual hysteresis behavior of cyclic loading of soil¹⁶. However, when soil shear forces are more than 10^− 5 kN/m^2, the soil takes on a nonlinear behavior¹⁷. The nonlinear model is more specific than the equivalent-linear method in representing the hysteresis stress-strain reaction of soil. Frequency and time domain analyses are used to determine the site response's nonlinear effects¹⁸. Site response analysis can be carried out using various computer programs, for example; SHAKE2000¹⁹, SHAKE91²⁰, and SHAKE²¹ are for 1D equivalent-linear, and DEEPSOIL²², OpenSees²³, and SUMDES²⁴ are for nonlinear site response analysis.

However, a few studies have been conducted in Bangladesh to estimate surface ground motion using site response analysis. For example, Rahman et al.⁹ performed site response analysis based on Vs30-based site coefficient, linear, equivalent-linear, and nonlinear approaches at ten locations of Dhaka city, considering the surface geological units of Dhaka⁹. In addition, Ansary & Jahan²⁵ performed an almost similar site response analysis for a solar power plant site in the northern part of Bangladesh²⁵. However, to the best of our knowledge, no published research considered simplified engineering geomorphic units of DAP of Dhaka city for seismic site characterization by nonlinear site response analysis.

Dhaka City is Bangladesh's administrative, political, and economic capital, with nearly 20–30% GDP of the country from Dhaka City. The city has ranked as one of the highest densely populated cities in the world in 2022, and the built-up area changed from 1989 to 2020 by around 92% ²⁶. The vigorous expansion of Dhaka city has recently created a large built-in environment over soft soil. The seismic vulnerability of Dhaka city is very high due to its geographic location, population density, unplanned urbanization, and non-engineered construction practice²⁷. If any earthquake occurs with a magnitude greater than 7.0, it is approximated that more than 72,326 buildings will be damaged²⁸. Therefore, to reduce vulnerability and make a beautiful habitable city, the Government of Bangladesh planned for detailed area planning in Dhaka City^29,30. Different geotechnical, environmental, network analysis, etc. analyses are needed for detailed area planning. The Sendai Framework for DRR (2015–2030) sets seven vibrant targets and four priorities for action to prevent new and reduce existing disaster risks. One of the priorities of the Sendai Framework is understanding disaster risk. Therefore, a detailed consideration of geomorphic unit-wise seismic site conditions is needed to minimize the future seismic risk to make a seismic resilient build-in environment.

The graphical representation of geomorphic features for an engineering application is known as engineering geomorphic mapping³¹. Geomorphic features can be identified, categorized, quantified, and visualized for site characterization using mapping. As part of the pre-development planning to reduce risk after development, geomorphic mapping is helpful for site-specific projects to detect and describe existing or potential geotechnical and hydro-technical hazards³¹. Only for the Dhaka city area Kamal & Midorikawa³² prepared a GIS-based geomorphological map with landfill sites. Further, they performed surface soil response analysis using boreholes and horizontal-to-vertical (H/V) spectral ratio (microtremor data) based on predominant periods and corresponding amplification factors³². Later, Rahman et al.^33,34 prepared a simplified geomorphic map of Dhaka city based on Vs30 and the liquefaction hazard map.

Currently, Dhaka city is considered one of the top seismic vulnerable cities; there is a scope and need for a seismic risk-sensitive land use plan for the proposed DAP of Dhaka City to reduce seismic risk. Therefore, the objective of this paper is to prepare a simplified engineering geomorphic unit map of the DAP of Dhaka City, followed by delineating the engineering properties of each unit and performing a nonlinear site response analysis of each unit considering the similar response spectra to the Bangladesh National Building Code (BNBC) 2020 for maximum credible earthquake (MCE) and SB type soil condition (Vs. 360 to 800 m/s). The outcome of this study can be used to prepare a seismic risk-sensitive land use plan (RSLUP) for the future development of the DAP of Dhaka City to reduce seismic risk.

The area of DAP of Dhaka City is 1528.436102 sq. km; DAP is included from Gazipur to Narayanganj, covering Savar, Old Dhaka, and Keraniganj (Fig. 1). According to Bangladesh's seismic zonation map, the DAP area is in Zone II and has a seismic zone coefficient of 0.15³⁵. Uplifted Pleistocene Madhupur terraces and Holocene floodplains ^32,34 are the two major surface geological units of the DAP area of Dhaka. The Pleistocene terrace consists of yellowish brown medium stiff to very stiff clayey silt (Madhupur clay) and is underlain by medium dense to dense silty sand (Dupitila Sand). However, the Holocene alluvium, or floodplains, are made of gray, very soft to medium stiff clayey silt, and very loose to medium dense clayey silt with silty sand beneath^36,33.

DAP of Dhaka City is located almost at the center of the Bengal basin. Bengal basin, formed by the Ganges-Brahmaputra-Meghna River system, is one of the world's largest and most active deltas. It is unique because of its colossal sediment thickness (approximately 22 km³⁷). The Himalayan Ranges between the Indian Plate and Eurasian Plate, as well as the Bengal Basin in the eastern part of the Indian plate, were both produced by the northward collision of the Indian plate with the Eurasian Plate^38,39,40,41. The Indian plate bounds the Bengal basin to the west, the Eurasian Plate to the north, the Burmese Plate, and the Bay of Bengal to the south and includes Bangladesh and parts of West Bengal, Tripura, and Assam of Indian states. The Himalayan system (product of the Indian-Eurasian plate) in the north and the Arakan subduction-collision system (product of the Indian-Burma plate) are the two major tectonic systems that have the potential to produce large earthquakes in the Bengal Basin area (Supplementary Fig. S1)⁴².

The large and destructive earthquakes in Bangladesh, northern and northeastern India, Nepal, Bhutan, and Myanmar were caused by two significant active tectonic belts: the Indian-Eurasian plate and the Indian-Burma plate⁴³. The 1897 Great Indian Earthquake significantly damaged buildings in Dhaka. Supplementary Fig S1 shows the earthquakes (with a magnitude of Mw ≥ 5) located in and around Bangladesh between 1908 and 2015. The tectonic belt of the Himalayan system is expected to experience large earthquakes between M 7.5 and M 8.5 as the Indian plate moves northward at a rate of 4 cm/year and northeastward at a rate of 6 cm/year⁴⁴. The Dauki Fault, which is thought to have caused the 1897 Great Indian Earthquake⁴⁵, was activated three times in the last thousand years, according to recent paleo-seismological research^46,47,48. As a result, the rate of plate movement and regular occurrence of large-magnitude earthquakes along this plate boundary suggests that Bangladesh, Nepal, Bhutan, and Myanmar, as well as the northern and northeastern parts of India, are seismically active regions⁴².

The Chittagong-Tripura fold belt has a noticeably different landscape due to the N-S trending anticlines, which resulted from the collision of the Indian and the Burmese plates, resting on the décollement megathrust of the arc-trench system of the Indian plate and the Burma Arc⁴⁹. This zone is a consequence of the ongoing deformation along the subduction zone. In addition, the Neogene deformation system of the Chittagong-Tripura fold belt has minor seismicity indicating a state of locked condition of the plates⁴³.

Engineering Geomorphic Map

The engineering geomorphic unit map of the DAP of the Dhaka City area was prepared and verified with the boreholes and surface geological data. The region has been divided into three simplified geomorphic units. They are the Holocene deposit, Pleistocene deposit, and Landfill. The Holocene and Pleistocene deposits are the two main geomorphic units (Fig. 2). Nearly 50% of the area comprises Holocene deposits, and the remaining 33% includes Pleistocene deposits. The sticky, stiff to very stiff silty clay or clayey silt and reddish brown to yellowish brown color is the unique characteristics of Pleistocene deposit (overconsolidated clay) underlain by medium dense to dense silty sand³³. Holocene deposits are mainly composed of light grey to dark grey color, very soft to soft clayey silt or silty clay, and gray to brown, loose to dense sandy silt or silty sand³³. The common mineral components of the Holocene unit are quartz, feldspar, and mica^36,50,34. The landfill unit is characterized by grey loose silty sand to sand and soft clayey silt. River dredging sand-filling sites are very vulnerable to liquefaction hazards; for example, the 1995 Kobe earthquake caused massive liquefaction in Port Island's reclaimed soil deposits.

Engineering Geomorphic Unit-wise Soil Properties

We have identified seven types of soil profiles for the study area's major categorized geomorphic units. Holocene deposits cover most of the DAP region, followed by Pleistocene deposits (Fig. 2). Among the seven types of soil profiles ( geomorphic sub-units), two types included Holocene alluvium deposit (HA), two types as Landfill (LAN) and three classes as a Pleistocene deposit (overconsolidated clay (OC)) (Fig. 3).

As mentioned earlier, we have identified three prominent varieties of overconsolidated clay geomorphic units: OC 1, OC 2, and OC 3. OC 1 consists of three primary soil layers from top to bottom; Madhupur clay, Madhupur silt, and Dupitila formation (Fig. 3). OC 2 consists of Madhupur clay underlain by Dupitila formation, and OC 3 consists of Madhupur silt and Dupitila formation (Fig. 3). The identified two prominent verities of Holocene alluvium deposit are HA 1 and HA 2. HA 1 consists of, from top to bottom, Holocene clay, Holocene silt, and Holocene sand. However, HA 2 consists of HA 1 and OC 1, where OC 1 is overlain by HA 1 (Fig. 3). In addition, there are two types of landfill soil units where overconsolidated clay is overlain by Landfill (LAN OC) and Landfill underlain by Holocene alluvium (LAN HA). The average filling depth in these two types is 3–4 meters, and most of the Landfill was composed of river sand. Engineering properties of the identified seven soil profiles of three geomorphic units are given in Table 1.

Table 1

Engineering properties (Average) of Pleistocene Deposit (OC), Holocene Deposit (HA), and Landfill (LAN)
Unit Name	Layer type	% of sand	% of Fine	Thickness (m)	SPT N Value Range	Unit Weight (kN/m3)	Vs. (m/s)	C (kN/m2)	Friction ($^\circ$)	PI.	Density (gm/cc)	Vs30 (m/s)
OC 1	Madhupur clay	< 10	> 90	≈ 7.5	5–25	≈ 19	200–250	50–200	< 10	25–35	≈ 2.00	220–350
	Madhupur silt	< 20	> 80	≈ 6	20–40	≈ 18	220–300	100–300	< 15	25–35	≈ 1.9
	Dupitila	> 60	< 40	≈ 17	30–60	≈ 20	300–400	< 15	30–40	-	≈ 2.1
OC 2	Madhupur clay	< 10	> 90	≈ 8	5–25	≈ 19	200–250	50–200	< 10	25–35	≈ 2.0	250–350
OC 2	Dupitila	> 60	< 40	≈ 22	30–60	≈ 20	300–400	< 10	35–40	-	≈ 2.1	250–350
OC 3	Madhupur silt	< 10	> 90	≈ 7	20–40	≈ 18	220–300	100–300	< 15	25–35	≈ 1.9	250–350
OC 3	Dupitila	> 60	< 40	≈ 23	30–60	≈ 20	300–400	< 10	35–40	-	≈ 2.1	250–350
HA 1	Holocene clay	< 10	> 90	≈ 12	0–10	≈ 16	120–180	40–100	< 10	20–25	≈ 1.6	150–180
	Holocene silt	< 20	> 80	≈ 7	10–25	≈ 17	150–200	20–50	< 20	15–25	≈ 1.7
	Holocene sand	> 60	< 40	≈ 11	20–35	≈ 18	180–250	< 20	20–30	5–12	≈ 1.8
HA 2	Holocene clay	< 10	> 90	≈ 7.5	0–10	≈ 16	120–180	40–100	< 10	20–25	≈ 1.6	180–250
	Holocene silt	< 20	> 80	≈ 4.5	10–25	≈ 17	150–200	20–50	< 20	15–25	≈ 1.7
	Holocene sand	> 60	< 40	≈ 2	20–35	≈ 18	180–250	< 20	20–30	5–12	≈ 1.8
	Madhupur clay	< 10	> 90	≈ 5	5–25	≈ 17	200–250	50–200	< 10	20–35	≈ 2.00
	Madhupur silt	< 20	> 80	≈ 2	20–40	≈ 18	220–300	100–300	< 15	25–35	≈ 1.9
	Dupitila	> 60	< 40	≈ 9	30–60	≈ 19	300–400	< 15	30–40	-	≈ 2.0
LAN-OC	Landfill	> 90	< 10	≈ 3	0–10	≈ 16	120–180	< 10	25–30	-	≈ 1.7	200–300
	Madhupur Clay	< 10	> 90	≈ 7.5	5–25	≈ 19	250–300	100–200	< 10	25–35	≈ 2.00
	Madhupur Silt	< 20	> 80	≈ 1.5	20–40	≈ 18	250–300	100–300	< 15	25–35	≈ 1.9
	Dupitila	> 60	< 40	≈ 18	30–60	≈ 19	300–400	< 15	30–40	-	≈ 2.0
LAN-HA	Landfill	> 90	< 10	≈ 3	0–10	≈ 17	120–180	< 10	< 20	-	≈ 1.8	150–200
	Holocene Clay	< 10	> 90	≈ 9	0–10	≈ 16	120–180	40–100	< 10	20–25	≈ 1.6
	Holocene Silt	< 20	> 80	≈ 10	10–25	≈ 17	150–200	20–50	< 20	15–25	≈ 1.7
	Holocene Sand	> 60	< 40	≈ 2	20–35	≈ 18	180–250	< 20	20–30	5–12	≈ 1.8
	Dupitila	> 60	< 40	≈ 5	30–60	≈ 19	300–400	< 15	30–40	-	≈ 2.0

Site Response Analysis

a b

c
de
fg
Figure 4: a. Response spectra of OC 1; b. Response spectra of OC 2; c. Response spectra of OC 3; d. Response spectra of LAN-OC; e. Response spectra of LAN-HA; f. Response spectra of HA 1; g. Response spectra of HA 2 for selected earthquakes' strong motion input.

The geomorphic unit-based response spectra for nine selected input ground motions are illustrated in Fig. 4. The response spectrums indicate various outcomes depending on the geomorphic unit and input earthquakes ground motion; however, the overall observation is that spectral acceleration is de-amplified in the short period and amplified in the long period. All the responses are measured considering a 30m soil column where Vs is regarded as 360–450 m/s at the bottom of the soil column.

The overconsolidated clay soil (OC 1, OC 2 & OC 3) (Fig. 4a, b, c) represents that the responses are de-amplified in a short period (0.1s) while peak acceleration at around 0.7s. The three types of overconsolidated clay soil represent similar responses for all the selected earthquake ground motions. However, responses from landfill units (LAN-HA and LAN-OC) are not identical. The LAN-HA shows predominant peak acceleration at 0.2s, whereas the LAN-OC unit show peak response at around 0.5s. Figure 4d and 4e show the response spectrum for LAN-HA and LAN-OC, respectively. The spectral acceleration of LAN-OC shows two peaks; one is predominantly lower than the primary peak acceleration as the soil's response matches with the frequency of ground motion. The spectral acceleration of HA 1 and HA 2 is illustrated in Fig. 4f and 4g, respectively.

Kobe and Northridge's motion from the chosen earthquake ground motion has the highest response acceleration for all types of units (Fig. 4). However, the artificial and Northridge earthquake ground motion have been significantly magnified over a long period. In this analysis, response spectra are calculated concerning frequency-independent viscous damping ratio using MRDF curve fitting for Darendali's curve fitting factor⁵¹.

a		b

c

Figure 5: a. Soil profile for overconsolidated clay (OC1), b. Layer-wised response of OC 1 and input ground motion (Gorkha 2015 earthquake), and c. Response spectrum for OC1.

The layer-wise response for OC 1 unit is illustrated in Fig. 5. As expected, the top layer, Madhupur Clay, has a maximum response spectrum than Madhupur silt and Dupitila Formation (Fig. 5c) compared with the input earthquake motion (Gorkha 2015), the input motion has maximum acceleration at 0.17s with 0.66g. In contrast, the responses have maximum acceleration at 0.53s with 0.83g (Fig. 5b). The responses are de-amplified in a short period until 0.1s.

In this study, site responses are estimated for seven types of engineering geomorphic sub-units under three main units using nonlinear analysis. We observed that the response spectrum shifted towards the long period with maximum acceleration more significant than the input ground motion (Fig. 4 and Fig. 5). The soft sandy and clayey soil up to 30 m depth is responsible for amplifying acceleration over a long period and de-amplification of acceleration in short periods. However, according to the NEHRP site amplification factor, for 30m soil depth, the response is amplified in a short period⁵². On the other hand, the soft sandy and clayey soil of DAP of Dhaka city area do not follow the NHERP site amplification factor, and it shows de-amplification over a short period.

In a nonlinear analysis, the higher frequencies are amplified at high shear moduli, whereas the lower frequencies are amplified at low shear moduli (Supplementary Fig. S5). The de-amplification and amplification on the response spectrum depend not only on the shear wave velocity but also on the other soil properties. Hashash & Park⁵³ showed that because of the decreased shear wave velocity at the top 70 m of the profile, the Holocene alluvium profile exhibits less amplified acceleration in a short period than the Pleistocene profile and a slight shifting to longer periods.

From the engineering geomorphic unit map, the Holocene alluvium covers most of the study area. The rest of the site is covered by Pleistocene deposits and an artificial Landfill. The response spectrums are different according to soil type and shear wave velocity. We can see that Pleistocene deposits represent overconsolidated clay with maximum acceleration over a long period than Holocene and landfill deposits. On the contrary, Holocene deposits exhibit long-lasting amplification than overconsolidated clay. The artificial Landfill shows a mixture of them. High frequencies are attenuated more than low frequencies when passing through a column of soft, loose soil that consists primarily of sandy soil. High frequencies are amplified more in overconsolidated clay sediments than in thick and soft Holocene deposits due to the higher natural frequencies of overconsolidated clay sediments. Thicker Holocene sediments have lower natural frequencies, resulting in amplification in the lower frequencies than the overconsolidated clay sediments. Bangladesh is very susceptible to far-source low-frequency earthquakes. The low-frequency tremor can cause significant damage to low-frequency structures due to resonance effects.

In summary, for the considered earthquakes, the near-surface soil response of DAP of Dhaka city shows de-amplification of acceleration in the short period and amplified acceleration in the long period. This type of spectrum is controlled by the characteristics and properties of soft clayey and sandy soil, and response spectra differ according to the geologic unit. For example, the Holocene soil unit accelerates toward a long period more than the Pleistocene unit. However, the Pleistocene deposits exhibit higher acceleration values than those of the Holocene.

This paper has examined the engineering geomorphic unit-based nonlinear site response of DAP of Dhaka City. First, the simplified engineering geomorphic unit map was prepared by image analysis and then verified using borehole data and the surface geology map of GSB. The identified three major geomorphic units are Pleistocene over-consolidation clay/ Pleistocene Madhupur clay (OC), Holocene Alluvium (HA), and Landfill (LAN). Further, OC, HA, and LAN are divided into OC 1, OC 2, OC 3, HA 1, HA 2, LAN-OC, and LAN-HA sub-units considering the subsurface soil layer.

In total, nine input ground motions (seven from PEER NGA WEST2 data set and two synthetics) were used for one-dimensional nonlinear site response analysis of seven geomorphic sub-units of DAP of Dhaka city. The overall observation from the nonlinear soil site response analysis is that the responses are de-amplified in the short period while amplified in the long period.

OC's spectral acceleration amplitude is higher than HA in all units. However, HA units show longer peak spectral acceleration duration than OC units. This phenomenon might cause the resonance of tall buildings on HA units, consequently increasing the probability of damage. As expected, LAN-OC units show similar behavior of OC and LAN-HA to HA. The vigorous vertical and lateral expansion of Dhaka city has caused it to create a large built environment over soft soil. Therefore, the findings of this study can be used as an essential indicator to prepare a seismic risk-sensitive land use plan for the future development of the DAP of Dhaka City to reduce the seismic risk. A similar procedure can be adopted for other earthquake-vulnerable cities to understand the seismic risk.

In this study, a simplified engineering geomorphic map of the Dhaka DAP area is prepared. Then, the boreholes with various lab test results and downhole seismic test data are collected and used to verify the simplified engineering geomorphic unit and delineate unit-wise engineering properties. Later, nine strong ground motion data are selected to execute site response analysis. Supplementary Fig. S2 shows the methodological framework adopted for this study. Further details of each component are explained in the subsequent sub-sections.

Simplified Engineering Geomorphic Map

To reduce earthquake risk, an engineering geomorphic map is essential for zoning any area. It is a land classification map based on the geomorphic unit, which helps to know the soil type variation on the surface. We have used both boreholes' data and satellite images to produce the engineering geomorphic map. The Landsat 5 image of 1988 and the Landsat 8 image for 2020 from USGS Earth Explorer are used for this research and are corrected using radiometric and atmospheric correction. Using the image from 1988, we digitized the basic geographic unit through visual interpretation (using natural and false colors). We then verified the geomorphic unit map from boreholes data and surface geology data from GSB (Geological Survey of Bangladesh). Later, NDWI (Normalized Difference Water Index) detected water bodies and Landfill for both images of 1988 and 2020. We extract the water bodies from images by NDWI, give values 0 to1, and then analyze the changes of water bodies in the images. From the studied images, landfill areas and water bodies are detected for the present time. Then these two units are added to the previously prepared map.

Engineering Soil properties

Primary and secondary data have been collected to delineate subsurface engineering properties. Primary data includes Dhaka University Master Plan and Development of Open Space Management System to Response Scenario Earthquake in Dhaka Metropolitan Area projects data as some authors were directly involved in the acquisition, processing, and interpretation of SPT borehole, geotechnical lab test, and downhole seismic test.

The secondary data are collected from the Geological Survey of Bangladesh (GSB), Creative Soil Investigations, Prosoil, Comprehensive Disaster Management Programme (CDMP) (Fig. 2). All data contain information about soil geotechnical properties, SPT N value, lithological properties, grain size, liquid limit, plastic limit, shear wave velocity, cohesion, friction angle, etc.

We have identified three major types of soil profiles from the collected borehole data considering the geomorphic units. They are Pleistocene over-consolidation clay/ Pleistocene Madhupur clay (OC), Holocene Alluvium (HA), and Landfill (LAN). Further, Pleistocene over-consolidation clay or Pleistocene Madhupur clay (OC) is divided into OC 1, OC 2, and OC 3 based on the subsurface soil layer. Similarly, Holocene Alluvium (HA) and Landfill (LAN) are divided as HA 1, HA 2, and LAN-OC LAN-HA, respectively. LAN-OC and LAN-HA mean landfill underlay by over-consolidation clay and Holocene alluvium, respectively. Figure 3 shows the simplified soil profiles of all geomorphic units of the study area.

Site Response Analysis

Soil properties and shear wave velocity

Shear wave velocity (Vs) of the near-surface soils is critical in seismic site response analysis^54,52. Although it is costly and not simple like SPT, it has advantages over SPT-N value⁵⁵. For example, measurements can be taken in hard soils where the SPT and CPT are impossible or not permitted or in areas where undisturbed samples are difficult to collect. In addition, it is a fundamental mechanical property of soil. Therefore, it is directly related to the small-strain shear modulus, a required parameter in analytical procedures for estimating dynamic soil response, site response analysis, and soil-structure interaction analyses⁵⁶. The down-hole seismic (DS) test locations of the study are shown in Fig. 2. The downhole seismic (DS) test is a direct method to estimate the Vs. and is considered the most reliable and accurate⁵⁷. Supplementary Fig. S3 displays the Vs. profile obtained through the downhole seismic test (PS logging).

However, where downhole seismic test data are unavailable to estimate Vs., we have used the equation Rahman et al.³⁴. The proposed equation shows a correlation mostly with the equation of Fujiwara⁵⁸. The used equations for shear wave velocity are given below:

Vs = 92.1*N^0.337 (All soils) (1)

Vs = 97.3062*N^0.3393 (All soils) (2)

Finally, the average shear wave velocity of the top 30 m (Vs30) is calculated using the following equation:

$${V}_{s }30=\frac{30}{{\sum }_{i=1}^{N}{h}_{i}/{v}_{i}}$$

Where, ${h}_{i}$ and ${v}_{i}$ denote the thickness (m) and Vs (m/s) of the ${i}^{th}$ layer, respectively, and $N$ denotes the number of layers in the top 30 m depth.

The shear wave velocity profiles of all geomorphic units are given in Supplementary Fig. S4.

Shear modulus reduction and material damping curves

Shear modulus and soil material damping curves are prerequisites for nonlinear site response analysis ⁵⁹. As shear modulus and damping ratio reduction curves are not available for Dhaka city, it is a prevalent practice to use standard curves from the literature. The widely used curves for categorizing dynamic soil behavior are represented in many studies^60,61,62. In this study, to fit the damping ratio and shear modulus reduction curve, as this empirical model is one of the most acceptable among practitioners for determining the variability of soft soil nonlinear properties, are used from Darendeli⁶⁰. Using the soil plasticity index, over-consolidation ratio, and the number of loading cycles, this model can predict the baseline for the damping ratio and reduction of modulus curves. Supplementary Fig. S5 illustrated the reference⁶⁰, fit and current curves of shear modulus, damping ratio and shear strength.

Strong ground motion selection

The strong ground motion time history (acceleration, velocity, and displacement) data is a prerequisite for dynamic response analysis of soil and structure. However, selecting strong ground motion is challenging as it depends not only on epicentral distance and rupture length but also on other factors, such as subsurface geology, directivity, etc. In the case of a moderate seismicity-prone and developing country, Bangladesh, the lack of strong ground motion data (Mw > 7.0 earthquake did not occur for more than 100 years in Bangladesh) is becoming more challenging. Therefore, in this study, we have used the most widely used spectral matching technique to select the strong ground time history acceleration data. Firstly, we have downloaded strong motion raw data (horizontal component, H1) from the PEER NGA WEST2 database considering the similar response spectra to the Bangladesh National Building Code (BNBC) 2020 response spectra (target response spectra) for Maximum Credible Earthquake (MCE) and SB type soil condition (Vs 360 to 800 m/s; Supplementary Fig. S6)³⁵. The details of the downloaded earthquake accelerograms are given in Supplementary Table S1. Afterward, we adjusted earthquake accelerograms by matching them with BNBC 2020 SB_MCE response spectra³⁵ using the wavelets algorithm proposed by Atik & Abrahamson (2010). Note that we have used SeismoMatch for spectral matching. Supplementary Fig. S7 shows the matched and target response spectra (BNBC 2020 SB_MCE). In addition, for this study, we have generated two artificial earthquake accelerograms matched to target response spectra (BNBC 2020 SB_MCE) by using SeismoArtif software. Both far-field inter and intra-plate earthquakes have been considered to generate synthetic accelerograms. In summary, seven earthquake accelerograms from seven earthquakes and two synthetic accelerograms have been considered for site response analysis. Supplementary Fig. S8 illustrates the detailed (step-by-step) procedure of ground motion selection for the Kobe 1995 earthquake.

Computing one-dimensional surface site response analysis

When motion is propagated vertically and horizontally, one-dimensional site response analysis evaluates the soil column's impacts on earthquake time histories vertically. Frequency domain and time-domain analysis are the most often used techniques in one-dimensional analysis. However, time-domain response analysis is more realistic than frequency-domain analysis because it considers changes in soil characteristics over time and the propagation of input motion into the soil⁶³.

The dynamic response of multi-degree-of-freedom systems exposed to base excitation is determined by nonlinear site response. This study utilizes DEEPSOILv7 software²² for calculating nonlinear response analysis. It is possible to determine nonlinear, linear, and equivalent linear response analysis in DEEPSOIL. Several soil constitutive models are accessible in DEEPSOIL, including the MKZ model with pressure-dependent behavior⁶⁴ and the new General Quadratic/Hyperbolic (GQ/H) model with masing requirements⁶⁵. Numerous models for reference fit curves for shear modulus reduction and damping ratio curves are also available in this computer program. The GQ/H model attempts to become flexible with minor strain soil behavior while simultaneously introducing shear strength failure. In this study, the GQ/H model is used as it is updated and has the best fit for sandy and clayey soils. Shear modulus reduction damping data from references are used to fit the model. Although numerous models exist for normalized shear modulus reduction and material damping curves, the reduction factor approach⁶⁰ is utilized to fit with soil parameters. Hysteretic damping at large stresses is overstated when these criteria are used. As observed in laboratory tests, the damping behavior of soils can be better captured by a non-Masing hysteresis model that shrinks the size of hysteresis loops. An MDRF model was developed for non-masing rules in the cyclic behavior of soil⁶⁶. Masing-type behavior overestimates the hysteretic damping after a 0.001 percent shear strain, whereas the suggested model's MRDF-type activity accurately represents damping behavior using the MDRF model with the GQ/H model.

Data availability

Strong Ground Motion Data are available at https://ngawest2.berkeley.edu/. The SPT and DST data from mentioned organizations could be acquired by contact with them in a proper way.

Code Availability

Data analysis is performed using Excel file, DEEPSOIL, AgrcGIS and Seismosoft.

Declarations

Table 1

Engineering properties (Average) of Pleistocene Deposit (OC), Holocene Deposit (HA) and Landfill (LAN). The engineering soil properties of all the seven sub-units are presented in this table with their units. The values of properties are shown in ranges.

Declarations

Competing interests

The authors declare no competing interests.

Authors Contributions

Md. Shakhawat Hossain: Conceptualization, Methodology, Writing- Original draft preparation, review and editing, Visualization, Investigation, Formal analysis, Artwork, Funding acquisition

Acknowledgments

We thank the Ministry of Science and Technology, Bangladesh, for partial funding this project, grant number (GO: 39.00.0000.009.14.019.21–745 (SL. 358, Gr. SL. 358 ES), Dated: December 15, 2021). We are also indebted to Mehedi Hasan, Masaba Tasneem Juhi, and Abul Kashem Faruki Fahim for helping to prepare borehole lithologies and image analysis. In addition, we thank Prof Raquib Ahsan and Prof Jahangir Alam for their valuable suggestions. Finally, we are indebted to the Geological Survey of Bangladesh (GSB), Creative Soil Investigations, Pro-Soil, Compressive Disaster Management Program (CDMP), Department of Disaster Science and Climate Resilience, University of Dhaka for providing us boreholes, downhole seismic tests, and geotechnical lab test data.

Boccard, N. Analysis of trends in disaster risk. Int. J. Disaster Risk Reduct. 53, 101989 (2021).
Steckler, M. S., Akhter, S. H. & Seeber, L. Collision of the Ganges-Brahmaputra Delta with the Burma Arc: Implications for earthquake hazard. Earth Planet. Sci. Lett. 273, 367–378 (2008).
Rahman, M. Z. & Siddiqua, S. Evaluation of liquefaction-resistance of soils using standard penetration test, cone penetration test, and shear-wave velocity data for Dhaka, Chittagong, and Sylhet cities in Bangladesh. Environ. Earth Sci. 76, (2017).
Tamura, K. Seismic design of highway bridge foundations with the effects of liquefaction since the 1995 Kobe earthquake. Soils Found. 54, 874–882 (2014).
Cruz-Atienza, V. M. et al. Long Duration of Ground Motion in the Paradigmatic Valley of Mexico. Sci. Rep. 6, 1–9 (2016).
Ansal, A., Tönük, G. & Kurtulu\cs, A. Implications of Site Specific Response Analysis. in Recent Advances in Earthquake Engineering in Europe: 16th European Conference on Earthquake Engineering-Thessaloniki 2018 (ed. Pitilakis, K.) 51–68 (Springer International Publishing, 2018). doi:10.1007/978-3-319-75741-4_2.
BSSC. NEHRP Recommended seismic provisions for new buildings and other structures. Build. Seism. Saf. Counc. I, 515 (2015).
USGS. M 7.8–67 km NNE of Bharatpur, Nepal. https://earthquake.usgs.gov/earthquakes/eventpage/us20002926/executive#executive (2015).
Rahman, M. Z., Siddiqua, S. & Kamal, A. S. M. M. Site response analysis for deep and soft sedimentary deposits of Dhaka City, Bangladesh. Nat. Hazards 106, 2279–2305 (2021).
Cramer, C. Site-Specific Seismic-Hazard Analysis that is Completely Probabilistic. Bull. Seismol. Soc. Am. - BULL Seism. SOC AMER 93, 1841–1846 (2003).
Bazzurro, P. & Cornell, C. A. Ground-Motion Amplification in Nonlinear Soil Sites with Uncertain Properties. Bull. Seismol. Soc. Am. 94, 2090–2109 (2004).
Bazzurro, P. & Cornell, C. A. Nonlinear soil-site effects in probabilistic seismic-hazard analysis. Bull. Seismol. Soc. Am. 94, 2110–2123 (2004).
Kaklamanos, J., Baise, L. G., Thompson, E. M. & Dorfmann, L. Comparison of 1D linear, equivalent-linear, and nonlinear site response models at six KiK-net validation sites. Soil Dyn. Earthq. Eng. 69, 207–219 (2015).
Hashash, Y. M. a & Groholski, D. R. Recent advances in nonlinear site response analysis. Fifth Interantional Conf. Recent Adv. Geotech. Earthq. Eng. Soil Dyn. Symp. Honor Profr. I.M. Idriss 29, 1–22 (2010).
Stewart, J. P., Afshari, K. & Goulet, C. A. Non-Ergodic Site Response in Seismic Hazard Analysis. Earthq. Spectra 33, 1385–1414 (2017).
Schnabel, P. B., Lysmer, J. & Seed, H. B. Shake-91. SHAKE a Comput. Progr. Earthq. response Anal. Horiz. Layer. sites. (1972).
Hardin, B. O. & Drnevich, V. P. Shear modulus and damping in soils: design equations and curves. J. Soil Mech. Found. Div 98, (1972).
Arslan, H. & Siyahi, B. A comparative study on linear and nonlinear site response analysis. Environ. Geol. 50, 1193–1200 (2006).
Ordóñez, G. a. SHAKE2000 for the 1-D Analysis of Geotechnical Earthquake Engineering Problems. Geomotions, LLC, Lacey, Washington, USA 310 (2012).
Idriss, I. M. & Sun, J. I. User’s manual for SHAKE91: a computer program for conducting equivalent linear seismic response analyses of horizontally layered soil deposits. University of California Davis, Davis, California (1992).
Schnabel, P. B. SHAKE: A computer program for earthquake response analysis of horizontally layered sites. EERC Rep. 72 – 12, Univ. California, Berkeley. (1972).
Hashash, Y. M. A. DEEPSOIL 7.0 User Manual. (2017).
Mazzoni, S., McKenna, F., Scott, M. H. & Fenves, G. L. OpenSees Command Language Manual. Pacific Earthquake Engineering Research Center. University of California, Berkeley (2006).
Li, X., ZL, W. & CK, S. SUMDES: A Nonlinear Procedure for Response Analysis of Horizontally-layered Sites Subjected to Multidirectional Earthquake Loading. Dep. Civ. Eng. Univ. California, Davis (1992).
Ansary, M. A. & Jahan, N. Site response and liquefaction susceptibility estimation of a site in northern part of Bangladesh. Environ. Earth Sci. 80, 1–18 (2021).
Szmigiera, M. Cities with the highest population density globally 2021. Statista https://www.statista.com/statistics/1237290/cities-highest-population-density/ (2021).
Rahman, M. Z., Siddiqua, S. & Kamal, A. S. M. M. Seismic source modeling and probabilistic seismic hazard analysis for Bangladesh. Nat. Hazards 103, 2489–2532 (2020).
CDMP. Earthquake Risk Assessment for Dhaka, Chittagong and Sylhet. (2009).
Seraj, T. M. & Islam, M. A. Detailed Area Plan: Proposals to Meet Housing Demand in Dhaka. Dhaka Metrop. Dev. Area Its Plan. Probl. Issues Policies; Jahan, S., Kalam, AKMA, Eds (2013).
Mowla, Q. A. Review of Dhaka Structure Plan 2016–2035. 1–7 (2016) doi:10.13140/RG.2.1.1065.8322.
Miller, B., Filatow, D., Dufresne, A., Geertsema, M. & Dinney, M. Engineering Geomorphological Mapping. Encycl. Eng. Geol. 278–292 (2017) doi:10.1007/978-3-319-73568-9_108.
Kamal, A. S. M. M. & Midorikawa, S. GIS-based geomorphological mapping using remote sensing data and supplementary geoinformation: A case study of the Dhaka city area, Bangladesh. Int. J. Appl. Earth Obs. Geoinf. 6, 111–125 (2004).
Rahman, M. Z., Siddiqua, S. & Kamal, A. S. M. M. Liquefaction hazard mapping by liquefaction potential index for Dhaka City, Bangladesh. Eng. Geol. 188, 137–147 (2015).
Rahman, M. Z., Kamal, A. S. M. M. & Siddiqua, S. Near-surface shear wave velocity estimation and V s30 mapping for Dhaka City, Bangladesh. Nat. Hazards 92, 1687–1715 (2018).
BNBC. Bangladesh National Building Code (BNBC) 2020. Ministry of Housing and Public Works (2020).
Karim, M. F. & Rahman, M. Z. Possible Effect of Moderate Earthquake on Existing Infrastructures of Dhaka City: A Geological-Geotechnical Overview. Bangladesh J. ofScience Technol. 4, 193–203 (2002).
Al., A. & Alam, Md. Khurshid., Shahidul Hasan, A. K. M., Khan, Mujibur Rahman,, Whitney, John W., Abdullah, S. K. M., Queen, James E., Geological Survey (US)., Office of Scientific Publications, G. S. of B. Geological map of Bangladesh. Geological Survey of Bangladesh (1989).
Ambraseys, N. N. & Douglas, J. Magnitude calibration of north Indian earthquakes. Geophys. J. Int. 159, 165–206 (2004).
Szeliga, W., Hough, S., Martin, S. & Bilham, R. Intensity, magnitude, location, and attenuation in India for felt earthquakes since 1762. Bull. Seismol. Soc. Am. 100, 570–584 (2010).
Atik, L. Al & Abrahamson, N. An Improved Method for Nonstationary Spectral Matching. Earthq. Spectra 26, 601–617 (2010).
Curray, J. R. & Moore, D. G. Sedimentary and Tectonic Processes in the Bengal Deep-Sea Fan and Geosyncline. Geol. Cont. Margins 617–627 (1982) doi:10.1007/978-3-662-01141-6_45.
Steckler, M. S. et al. Locked and loading megathrust linked to active subduction beneath the Indo-Burman Ranges. Nat. Geosci. 9, 615–618 (2016).
Wang, Y., Sieh, K., Tun, S. T., Lai, K.-Y. & Myint, T. Active tectonics and earthquake potential of the Myanmar region. J. Geophys. Res. Solid Earth 119, 3767–3822 (2014).
Bilham, R. & Hough, S. Future earthquakes on the Indian subcontinent: Inevitable hazard, preventable risk. South Asian J. 12, 1–9 (2006).
Oldham, R. D. Report on the great earthquake of 12th June 1897. Mem. Geol. Surv. India 29, 1–379 (1899).
Yeats, R. S. et al. The geology of earthquakes. (Oxford University Press, USA, 1997).
Morino, M. et al. A paleo-seismological study of the Dauki fault at Jaflong, Sylhet, Bangladesh: Historical seismic events and an attempted rupture segmentation model. J. Asian Earth Sci. 91, 218–226 (2014).
Singh, A. P., Mishra, O. P. & Singh, O. P. Seismic evidence of pop-up tectonics beneath the Shillong Plateau area of Northeast India. Sci. Rep. 12, 1–12 (2022).
Vorobieva, I., Gorshkov, A. & Mandal, P. Modelling the seismic potential of the Indo-Burman megathrust. Sci. Rep. 11, (2021).
Khatun, M., Ali, R. M. E., Karim, S. & Munsura Akther, K. Geomorphology and Geology of the Dhaka City Corporation Area-an Approach of Remote Sensing and GIS Technique. Int. J. Astron. 6, 7–16 (2019).
Phillips, C. & Hashash, Y. M. A. Damping formulation for nonlinear 1D site response analyses. Soil Dyn. Earthq. Eng. 29, 1143–1158 (2009).
Dobry, R. et al. New site coefficients and site classification system used in recent building seismic code provisions. Earthq. Spectra 16, 41–67 (2000).
Hashash, Y. M. A. & Park, D. Non-linear one-dimensional seismic ground motion propagation in the Mississippi embayment. Eng. Geol. 62, 185–206 (2001).
Borcherdt, R. D. Estimates of Site-Dependent Response Spectra for Design. Earthquake Spectra vol. 10 617–653 (1994).
Andrus, R. D. & Stokoe II, K. H. Liquefaction Resistance of Soils from Shear-Wave Velocity. J. Geotech. Geoenvironmental Eng. 126, 1015–1025 (2000).
Youd, T. L. & Idriss, I. M. Liquefaction Resistance of Soils: Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils. J. Geotech. Geoenvironmental Eng. 127, 297–313 (2001).
Boore, D. M. & Brown, L. T. Comparing Shear-Wave Velocity Profiles from Inversion of Surface-Wave Phase Velocities with Downhole Measurements: Systematic Differences between the CXW Method and Downhole Measurements at Six USC Strong-Motion Sites. Seismol. Res. Lett. 69, 222–229 (1998).
Fujiwara, T. Estimation of ground movements in actual destructive earthquakes. in Proceedings of the fourth European symposium on earthquake engineering, London vol. 125 132 (1972).
Rathje, E. M., Kottke, A. R. & Trent, W. L. Influence of Input Motion and Site Property Variabilities on Seismic Site Response Analysis. J. Geotech. Geoenvironmental Eng. 136, 607–619 (2010).
Darendali, M. Ba. Development of a new family of normalized modulus reduction and material damping curves. The University of Texas at Austin, Austin (2001).
Zhang, J., Andrus, R. & Juang, C. H. Model Uncertainty in Normalized Shear Modulus and Damping Relationships. J. Geotech. Geoenvironmental Eng. 134, (2008).
Zefeng, Y., Jiao, Y., Junwei, L. & Bo, H. Shear Modulus Degradation Curves of Gravelly and Clayey Soils Based on KiK-Net In Situ Seismic Observations. J. Geotech. Geoenvironmental Eng. 143, 6017008 (2017).
Tsai, C.-C. & Chen, C.-W. Comparison Study of One-Dimensional Site Response Analysis Methods. Earthq. Spectra 32, 1075–1095 (2016).
Matasović, N. & Vucetic, M. Cyclic characterization of liquefiable sands. J. Geotech. Eng. 119, 4279 (1993).
Groholski, D. R., Hashash, Y. M. A., Musgrove, M., Harmon, J. & Kim, B. Evaluation of 1-D Non-linear Site Response Analysis using a General Quadratic/Hyperbolic Strength-Controlled Constitutive Model. in 6ICEGE: 6th International Conference on Earthquake Geotechnical Engineering (2015).
Numanoglu, O. A., Musgrove, M., Harmon, J. A. & Hashash, Y. M. A. Generalized Non-Masing Hysteresis Model for Cyclic Loading. J. Geotech. Geoenvironmental Eng. 144, 1–6 (2018).

Muneyoshi Numada: Supervision, Conceptualization, Methodology, Writing- review and editing

Momtahina Mitu: Software, Data Analysis, Writing- Original draft preparation

Kishor Timsina: Software, Data analysis

Chaitaniya Krisna: Supervision, Writing- reviewand editing

A. S. M. Maksud Kamal: Supervision, Methodology

Md. Zillur Rahman: Data curation, Supervision, Methodology

Kimiro Meguro: Supervision, Methodology

No competing interests reported.

Download PDF

Journal Publication

published 09 Jul, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
09 Apr, 2023
Reviewers agreed at journal
10 Mar, 2023
Reviews received at journal
03 Mar, 2023
Reviewers agreed at journal
02 Mar, 2023
Reviewers agreed at journal
02 Dec, 2022
Reviewers invited by journal
02 Nov, 2022
Editor assigned by journal
02 Nov, 2022
Editor invited by journal
02 Nov, 2022
Submission checks completed at journal
02 Nov, 2022
First submitted to journal
31 Oct, 2022

You are reading this latest preprint version

Simplified Engineering Geomorphic Unit-based Seismic Site Characterization of Detailed Area Plan of Dhaka City

Status:

Journal Publication

Version 1

Abstract

Introduction

Geology and Seismotectonics of the Study Area

Results

Engineering Geomorphic Map

Engineering Geomorphic Unit-wise Soil Properties

Site Response Analysis

Discussion

Conclusion

Methods

Simplified Engineering Geomorphic Map

Engineering Soil properties

Site Response Analysis

Soil properties and shear wave velocity

Shear modulus reduction and material damping curves

Strong ground motion selection

Computing one-dimensional surface site response analysis

Data availability

Code Availability

Declarations

Declarations

Declarations

Authors Contributions

Acknowledgments

References

Unsectioned Paragraphs

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1