Effects of Land Use Change on Surface Runoff and Infiltration: The Case of Dhaka City

Abstract

This study presents an integrated field- and model-based assessment of how rapid urbanization is transforming water infiltration and storm runoff dynamics in Dhaka—a megacity facing escalating flood risks. Unlike conventional studies that rely solely on secondary or modeled datasets, this research combines extensive in situ field measurements of soil infiltration with scenario-based hydrological modeling to capture the localized impacts of land use change. Using the SCS Curve Number and Water Balance methods, the study quantifies how variations in land cover under different urban growth trajectories alter surface runoff behavior. Results show that Dhaka’s annual infiltration rates—measured at 2034 mm, 1546 mm, and 1074 mm during wet (2017), normal (2018), and dry (2020) years—could decline by nearly 50% if current urban expansion trends persist. Concurrently, surface runoff volumes are projected to nearly double, amplifying flood hazard potential across the city. By grounding scenario modeling in empirical local data, this work offers a context-specific understanding of the evolving hydrological regime of a rapidly urbanizing South Asian metropolis, providing a framework for flood resilience planning in other data-limited cities.

1. Introduction

Cities in developing countries are increasingly experiencing environmental degradation due to rapid population growth, industrial expansion, and unplanned urbanization [1]. Dhaka, one of the world’s most densely populated megacities with nearly 20 million inhabitants, consistently ranks among the least liveable cities globally [2]. The city’s unregulated expansion has led to the extensive conversion of green and permeable land into impervious built-up surfaces, heightening the risks of surface runoff, flooding, and waterlogging. Against this backdrop, the present study investigates the impacts of land use changes on (soil) water infiltration and surface runoff processes, with the objective of providing insights for improved urban environmental management in data-scarce contexts like Dhaka.

While numerous studies have explored Dhaka’s flood, drainage, and stormwater management challenges, this study makes a distinct contribution by integrating locally measured infiltration data with scenario-based hydrological modeling using the SCS-CN (Soil Conservation Service Curve Number) and Water Balance methods. This integration bridges the persistent gap between empirical field evidence and model-based estimations in a tropical, data-limited megacity context. By calibrating model parameters through field observations, the study enhances the reliability and contextual validity of runoff simulations that are often generalized from other regions. Such locally grounded modeling offers an improved basis for scenario testing, urban planning, and climate-resilient water management.

The pace of urbanization in Dhaka continues to pose formidable challenges for sustainable growth, especially in managing land use transitions and maintaining hydrological balance [3,4,5,6]. In highly dense urban settings, where natural landscapes are replaced by impermeable surfaces, urbanization has emerged as the dominant driver altering the physical and ecological character of the built environment [7]. Dhaka’s expansion—driven by intense rural-to-urban migration and unregulated real estate development—has led to rapid land use transformation [8,9]. This process severely disrupts natural infiltration, amplifies surface runoff, and weakens the resilience of local drainage and aquifer systems [8,10,11,12].

Effective surface runoff management is, thus, critical to mitigate recurring urban challenges such as flooding, waterlogging, and stream erosion [1,13]. However, most drainage systems in Dhaka remain under-designed to accommodate changing hydrological realities. The city’s groundwater recharge depends largely on rainfall and infiltration, yet the shrinkage of natural recharge zones due to unplanned construction has drastically reduced infiltration capacity [14]. The sealing of vertical recharge inlets by concrete and asphalt further intensifies monsoon waterlogging, compromising the hydrological sustainability of the urban system [15].

Existing research has largely concentrated on urban stormwater management, drainage performance, and groundwater recharge, often overlooking the fundamental hydrological processes of infiltration and runoff that underpin these challenges. This study addresses that critical gap by conducting field-based infiltration experiments and integrating the results with scenario-driven modeling to evaluate how land use and climate variability jointly shape surface hydrological responses. Using the Curve Number and Water Balance approaches, the study quantifies changes in infiltration and runoff dynamics across different land use and climatic scenarios. The findings offer empirical insights with direct implications for urban environmental planning, stormwater management, and sustainable land development policies in Bangladesh and other rapidly urbanizing tropical cities.

The organization of this paper is as follows. Section 2 provides a review of relevant literature. Section 3 describes the study area, data, and methods applied. Section 4 presents the results from field experiments and secondary data analyses. Section 5 discusses the implications of the findings, and the limitations. Finally, Section 6 concludes with key insights and implications for urban environmental management.

2. Literature Review

Urbanization alters the hydrologic cycle by transforming natural landscapes into impervious built environments, thereby changing the timing, magnitude, and pathways of water flow [16,17,18,19]. The increase in paved surfaces reduces infiltration, shortens the lag time between rainfall and runoff peaks, and amplifies surface runoff volumes, which in turn exacerbate flooding, waterlogging, and environmental degradation [17,18,20]. Imperviousness also impedes groundwater recharge and increases pollutant loads in stormwater runoff, further degrading urban water quality [13,17,21]. These hydrological alterations underscore the need to understand and manage surface–subsurface interactions in cities undergoing rapid and unplanned development.

2.1. Surface Water System

Surface water management is fundamental to urban sustainability, requiring approaches such as decentralized water treatment, green infrastructure, and source-level reuse [22,23,24]. In Dhaka City, rapid urban expansion has reduced infiltration and natural storage capacity, leading to increased surface runoff and heightened flood risk [25]. Within the Drainage Compartment (DC-3) of Greater Dhaka East, the proposal for a retarding pond (RP 7-1) illustrates the role of structural measures in mitigating flood peaks and enhancing stormwater retention [26,27]. However, beyond engineering interventions, there is an urgent need for land use sensitive planning and multi-functional water management solutions, such as the use of retarding areas for aquaculture, recreation, and water harvesting. These approaches would transform flood control infrastructure into resilient and productive urban assets.

2.2. Storm Runoff

The proliferation of impervious surfaces in cities intensifies stormwater discharge, accelerates runoff generation, and elevates the risk of urban flooding [28,29]. Sustainable stormwater management strategies, guided by ecological design principles and hydrological balance, are, therefore, essential to protect both infrastructure and ecosystems [30,31]. Studies have demonstrated that urbanization increases peak runoff volumes and eliminates natural retarding basins, particularly in low-lying areas of Dhaka [32,33,34]. Simulation-based analyses further reveal that full urbanization scenarios may increase more storm runoff compared to pre-urban conditions, emphasizing the strong hydrological sensitivity of land use transitions [35]. Yet, few studies have investigated these models locally with field data in tropical South Asian megacities, where intense monsoon rainfall and heterogeneous land use patterns present challenges for runoff estimation and management.

2.3. Urban Flooding

Waterlogging and urban flooding are now pervasive challenges in many cities, driven by rapid urbanization, insufficient drainage capacity, and encroachment upon natural waterways [36,37,38]. In Dhaka, heavy rainfall frequently overwhelms existing drainage systems, flooding low-lying neighborhoods [39,40]. Studies attribute these issues to unplanned development, which blocks, diverts, and fills natural drainage channels [41], and to urban densification, which increases peak flows and shortens lag times between rainfall and runoff [42]. These findings collectively stress the importance of hydrological understanding in urban planning to reduce the vulnerability of flood-prone areas.

2.4. Aquifer Recharge

Groundwater depletion is another growing concern in Dhaka, where excessive extraction and limited recharge have led to declining water tables and associated hazards such as land subsidence [43]. Rapid urbanization obstructs natural recharge processes by sealing permeable areas with impervious materials, thus reducing infiltration and percolation rates [44]. Population growth and encroachment on natural retention areas further constrain recharge potential [45]. Addressing these trends requires targeted strategies, such as artificial recharge systems and the preservation of infiltration zones, to restore balance to the urban groundwater budget.

2.5. Water Cycle

Urbanization’s cumulative effects extend across the entire water cycle. In natural systems, approximately 25% of rainfall infiltrates into aquifers, whereas in highly urbanized areas, more than half becomes surface runoff with minimal deep percolation [46,47,48,49]. Land use changes, particularly the conversion of agricultural or open spaces into built-up areas, are key drivers of increased runoff coefficients and altered hydrological regimes [18]. Studies from diverse contexts, such as the River Basin in Shenzhen, show that urbanization amplifies runoff, shortens concentration times, and raises flood peaks [50,51]. Similar challenges are reported in mountainous catchments, where predicting hydrological responses to land use change remains complex due to uncertainty [52]. Green infrastructure, such as permeable pavements, offers a potential mitigation pathway by enhancing infiltration, reducing runoff, and improving stormwater quality [52,53].

Collectively, the literature demonstrates the profound impacts of urbanization on infiltration, runoff, and flood dynamics. Yet, most studies rely heavily on modeling or secondary datasets, with limited incorporation of empirical field measurements for calibration and validation, particularly in tropical, data-scarce megacities like Dhaka. This gap constrains the reliability of hydrological predictions and weakens their applicability in local urban planning. Hence, an approach that combines field-based infiltration experiments with scenario-driven runoff modeling is essential to capture the nuanced hydrological responses to land use change under real-world urban conditions. This study directly addresses this gap by applying a field experiment, scenario-based hydrological assessment within the Dhaka context, providing much-needed empirical grounding for sustainable urban water and flood management in South Asian megacities.

3. Data and Method

3.1. Study Area Profile

The study area is located within the Badda jurisdiction in Dhaka, standing on the eastern side of Pragati Sarani. It covers an area of 5.092 km², consisting of the northern part of the Banasree New Town area. The Balu River passes along the eastern side of the study area, which is about 3 km away. Figure 1 below represents the study area with respect to Dhaka city.

The layout plan of the study area was obtained from the Project Office of Eastern Housing Limited, a private developer, while images were sourced from Google Earth and geo-referenced by overlaying them onto a pre-geo-referenced road map of Dhaka City. Seven Ground Control Points were derived from the reference image for accurate geo-referencing.

The proposed land use plan envisions transforming the area into a planned urban extension. The development area is located within Badda, adjacent to the Rampura and bordered to the south by the Begunbari Khal. According to the developer’s plan, the proposed land use development is scheduled to be implemented over a 10-year horizon (2025–2035), with phased expansion of residential, commercial, and mixed-use zones alongside essential drainage, road, and utility infrastructure. This planned development aims to accommodate Dhaka’s eastward urban growth, yet it raises questions about hydrological sustainability, particularly concerning surface runoff, infiltration, and flood risk in this low-lying and flood-prone landscape.

The landscape comprises land-filled sites in the west, while vast wetlands characterize the northern and eastern regions. The low-lying areas are prone to flooding during the rainy season, remaining submerged for over half of the year.

The area experiences a subtropical humid climate characterized by short, cool winters and long, hot summers with high rainfall. Rainfall originates from the western depression of winter, early summer thunderstorms, and the monsoon. The rainy season spans from June to October, with approximately 80% of the annual rainfall occurring during this period, averaging around 2000 mm annually.

The stratigraphy exhibits homogeneity, comprising layers of sands, silts, and clays. Seven hydro-stratigraphic units have been delineated based on lithological composition, presence of textured sediments, and stratigraphic depth up to 315 m (

Table 1. Hydro-stratigraphic Units of Dhaka Aquifer System.

Hydro-Stratigraphic Unit	Average Thickness (m)	Average Bottom Depth (m)
Top Soil	8.98	8.98
Aquitard	40.65	49.63
Aquifer-1	68.82	118.45
Aquitard/Aquiclude-1	15.72	134.17
Aquifer-2	32.22	166.39
Aquitard/Aquiclude-2	26.08	192.47
Aquifer-3	93	285.47

Source: [54].

).

The Deeper Aquifer, recently examined by DWASA production wells, extends from depths of 195.40 m to 353.28 m. Within this range, approximately 35% of the sediment column comprises the 3rd aquifer system, while 32% consists of Aquifers 1 and 2. Based on grain-size distribution and hydraulic properties, the aquifers are classified into three units: upper, middle, and lower. The lower unit serves as the primary aquifer, with an average transmissivity of 1600 m²/day. The upper clay cap thickness ranges from 10 to 15 m [55].

In the eastern part of the city, a 30 to 45 m thick layer of silty clay covers the surface. The first aquitard thickness ranges from 45 to 60 m in the Banasree area. Wetland filling, alteration of natural drainage, and urban expansion substantially decrease recharge areas and vertical recharge. Water table decline rates have reached up to 3.5 m/year in the city’s central areas, leading to the upper aquifer transitioning from semi-confined to unconfined due to excessive extraction.

The study area is predominantly green field, covering 70.47% of the total area (

Table 2. Existing and proposed land use.

Land Use	Existing		Proposed
	Area (km2)	Area (%)	Area (km2)	Area (%)
Built-up	0.053	1.04	3.521	69.19
Agriculture	3.588	70.47	-	-
Road Network	0.217	4.26	1.164	22.84
Fallow land	1.218	23.92	-	-
Water body	0.016	0.32	0.109	2.14
Open Space	-	-	0.297	5.83
Total	5.092	100.00	5.092	100.00

). Fallow land, comprising land-filled areas earmarked for residential development, occupies 23.92% of the area. Road networks cover 4.26% of the study area. Built-up areas, including residential and commercial structures, account for 1.04% of the total area. Only one water body spans 0.016 km² and serves as a drainage channel. Additionally, a proposed artificial lake covering 0.093 km² is planned as an on-site retention area. Figure 2 and Figure 3 below illustrate existing and planned land uses.

3.2. Land Use Classifications and Map Preparation

Land use and land cover (LULC) categories for the natural, existing, and proposed scenarios were defined and mapped through an integrated visual-interpretation and plan-based classification procedure. First, the official layout plan obtained from Eastern Housing Limited was used as the primary reference for delineating proposed land use classes such as residential, commercial, road networks, water bodies, and open spaces. For the existing condition, high-resolution Google Earth imagery (acquired for the year corresponding to the field survey) was visually interpreted to identify surface features based on their spectral signatures, texture, shape, and contextual attributes. Built-up areas were identified by the presence of permanent structures, paved surfaces, and dense rooftop patterns; agricultural land and open green fields were characterized by uniform vegetation tones and parcel boundaries; fallow land (land-filled but undeveloped) was distinguished by exposed sandy fill material, absence of vegetation, and the presence of earth-moving traces; and water bodies were mapped using their distinct dark spectral tone and continuous surface geometry.

To ensure spatial accuracy, all Google Earth images were geo-referenced to a pre-referenced road map using seven ground control points (GCPs), allowing the imagery to be aligned to the same coordinate system as the layout plan. Each LULC polygon was manually digitized in GIS software 10 and cross-validated through field observations, with particular attention to differentiating between natural clayey lowlands and recently filled sandy areas. The final LULC maps, therefore, represent a consistent, plan-verified classification where each category corresponds to a distinct land use function and hydrological behavior, forming the basis for assigning Curve Numbers (CN) and infiltration parameters under natural, existing, and proposed development scenarios.

The existing land use information was extracted from the geo-referenced image and the layout plan obtained from Eastern Housing Ltd. The data were converted to vector format through on-screen digitization using ArcGIS 10. The digitized features were saved as shapefiles and projected using the Lambert Conformal Conic (LCC) coordinate system. The line shapefile was subsequently converted to polygons, and each polygon was assigned its corresponding land use category. An existing land use map was then prepared to illustrate the spatial distribution of land-cover types and was verified through field observations. The area of each land use class was calculated accordingly. Future land use areas were determined from the polygon shapefile created from the geo-referenced layout plan.

3.3. Data and Field Experiment

Infiltration rates were determined through field experiments, supplemented by proposed land use data, rainfall, and evaporation data gathered from secondary sources. Runoff volume was estimated using both the SCS Curve Number and the water balance methods.

Rainfall and evaporation data were gathered from the Bangladesh Meteorological Department, with three-hourly rainfall data from Dhaka Station utilized to determine present and projected runoff volumes. Geomorphological data, specific to Location-10 of the Detailed Area Plan, were obtained from the Capital Development Authority. Additional relevant information, such as climatic conditions, was sourced from online sources.

The study emphasizes the use of field measurements to determine infiltration rates at different soil sites and employs the curve number method, which accounts for land cover characteristics affecting permeability. For infiltrated areas, the influence of surface conditions, such as compaction and paving, is reflected through adjusted curve numbers based on land use types.

• Infiltration rate

The infiltration rates were measured at two sites using a double-ring infiltrometer, with the inner and outer rings driven 15 cm into the soil to prevent lateral water movement. One of the sites had soil of sandy texture, and the other had soil of clayey texture. These two types of soil cover the majority of the soil types in the study areas under existing conditions. Water was poured into the inner ring, and measurements were taken of the water infiltrated per unit area of soil surface over time. Readings were recorded at regular intervals until the drop in water level stabilized. Initially, readings were taken every 1–2 min, with intervals gradually extended to 20–30 min as the test progressed.

Infiltration for the present and natural condition was determined from the measured infiltration rate of each three-hourly rainfall event. For the proposed land use condition, infiltration rates were derived from measured compacted fill and paved surfaces, assuming mostly impermeable cover with limited permeable paving in designated open spaces.

• Holding Capacity of soil

Soil samples were taken from two places to estimate the water-holding capacity of the soils in the study area. One soil sample was taken for sandy loam soil and another for clay soil. The samples taken from the field were carried to the laboratory, and their water-holding capacities were measured.

Samples were collected from a wetted area, covered with polythene to minimize evaporation, and allowed to drain for two days, considering the sandy soil’s drainage characteristics. Soil samples were obtained using a coring device, filled into a can, and weighed to determine wet soil weight. After oven-drying at 105 °C for 24 h, the dry soil weight was measured. The difference between wet and dry soil weights indicated moisture retention. Soil volume was calculated from the core sampler dimensions. Two samples were collected to represent water holding capacity, one from a sandy land-filled site and the other from a natural clay low-lying area.

• Soil texture

The sieve analysis was applied to understand the distribution of the coarser, larger-sized particles. Initially, the weights of each sieve and the bottom pan were recorded, followed by the measurement of the dry soil sample weight. The sieves were then arranged in ascending order, with the soil poured into the top sieve and capped. The stack underwent shaking for 10 min in a mechanical shaker, after which each sieve and the bottom pan were weighed to determine soil retention. The mass of soil retained on each sieve was calculated by subtracting the sieve’s empty weight, and the percentage retained on each sieve was determined by dividing the retained weight by the original sample mass. The percentage passing through each sieve was calculated cumulatively by subtracting the percentage retained from 100%.

3.4. Estimation Methods

• Surface Runoff

The runoff generated from rainfall was estimated by both the ‘SCS Curve Number’ and the supplementary ‘Water Balance Methods’. The water balance technique is used to cross-check the results obtained from the SCS Curve Number method. The estimation was carried out for high (2017), normal (2018), and low (2020) rainfall years.

• SCS Curve Number method

The steps involved can be grouped as follows: Rainfall data collection, deciding the Antecedent Moisture Condition (AMC), preparation of land use data, determining the hydrological soil groups, exploring the soil cover complex, assigning the actual curve number values, calculation of potential maximum retention (S), calculation of 3-hourly runoff for the study area, and calculation of monthly runoff for the study area.

The method is widely adopted for the estimation of runoff from rainfall depths and takes into account the important physical aspects of a catchment on which runoff depends, such as land use, land gradient, soil cover, and antecedent soil moisture condition [56].

The rainfall-runoff relation used in the SCS method of the USDA was applied for the estimation of runoff as per Equation (1) [56,57].

$$Q={\displaystyle \frac{{\left(P-0.3S\right)}^2}{\left(P+0.7S\right)}}(P>0.3S)$$

(1)

where S is the watershed storage in mm; Q is the actual direct runoff in mm; and P is the total rainfall in mm. S is related to curve number (CN), $S={\displaystyle \frac{25400}{CN}}-254$.

Where CN is a parameter and its value ranges from 1 to 100. It is determined based on hydrologic soil group, land use, land treatment, and hydrologic conditions.

The Curve Number required to use this method was determined from the relevant table [57] for each hydrological soil group and land use and listed in

Table 3. Curve number (CN) for natural, present, and future land use conditions.

Land Use	Natural				Present				Future
	Soil Group	CN	Area (km2)	Area (%)	Soil Group	CN	Area (km2)	Area (%)	Soil Group	CN	Area (km2)	Area (%)
Built-up	-	-	-	-	A	98	0.053	1.04	A	98	3.521	69.19
Agriculture	D	89	5.076	99.68	D	89	3.588	70.47	-	-	-	-
Road	-	-	-	-	A	98	0.217	4.26	A	98	1.164	22.84
Vacant land	-	-	-	-	A	77	1.218	23.92	-	-	-	-
Water	D	89	0.016	0.32	D	100	0.016	0.32	A	100	0.109	2.14
Open space	-	-	-	-	-	-	-	-	A	39	0.297	5.83
Total			5.092	100.00			5.092	100.00			5.092	100.00

. To determine the curve number values for each classified area, the hydrological soil group and the land use and land cover information for natural, existing and proposed conditions were used.

By assigning Curve Numbers to each land use type, complying with the corresponding soil groups, and calculating the weighted average value, composite curve numbers were determined. Based on the data, the composite curve number was found by using Equation (2).

$$\mathit{CN}={\displaystyle \frac{\sum A_i\ast {CN}_i}{\sum A_i}}$$

(2)

where CN is the composite curve number of the study area, A_i is the area of land parcels of identical land uses, and CN_i is the curve number of the corresponding land parcels.

The CN values are documented for the case of AMC-II. To adjust the CN for the cases of dry condition (AMC-I) and wet condition (AMC-III), Equations (3) and (4) are used:

$${CN}_{\left(I\right)}={\displaystyle \frac{4.2\ast {CN}_{\left(II\right)}}{10-(0.058\ast {CN}_{\left(II\right)})}}$$

(3)

$${CN}_{\left(III\right)}={\displaystyle \frac{23\ast {CN}_{\left(II\right)}}{10+(0.13\ast {CN}_{\left(II\right)})}}$$

(4)

where CN_(II) is the curve number for normal condition, CN_(I) is the curve number for dry condition, and CN_(III) for wet condition.Water Balance method

The method considers precipitation on the input side, and runoff, evapotranspiration, and infiltration on the output side [58]. The study considered only the natural water parameters, as the area is mostly greenfield. In addition, the focus of the study is to assess only the changes in natural hydrological factors. The water balance equation for the runoff estimation is:

$$P=Q+ET+I$$

(5)

where P is precipitation, Q is runoff, ET is evapotranspiration, and I is infiltration. Infiltration depends on the intake capacity of the soil, rainfall depth in a three-hourly time period, and soil moisture condition in the previous three-hourly time period.

The infiltration rate obtained from the field experiment was used for the three-hourly water balance calculation. Reference crop evapotranspiration (ET₀) for Dhaka was assumed to be the parameter for evapotranspiration (ET) since the study area was covered mostly with grass. Cropwat 8.0 software was used to estimate the reference evapotranspiration (ET₀) from the Penman-Monteith equation for Dhaka [59].

Surface runoff for the study area has been estimated first from the SCS Curve Number method for existing, natural, and planned land use conditions. In order to verify the runoff depth estimated from the SCS Curve Number method, runoff depth has also been estimated from the water balance method.

The Water Balance method was applied to validate the SCS-CN estimates. Differences between the two reflect model structure; SCS-CN is event-based and sensitive to Curve Number calibration, while the Water Balance method accounts for cumulative water fluxes and may yield higher runoff during average rainfall years due to evapotranspiration assumptions.

4. Results

4.1. Infiltration

(i) Infiltration curve

Infiltration rate was computed from the datasheets of the field tests are presented in graphical forms as infiltration curves in Figure 4a and Figure 4b for clayey and sandy soil, respectively. Figure 4a illustrates the infiltration characteristics of clayey soil, showing an initial infiltration rate of about 55 mm/h in dry conditions, which decreases to approximately 6 mm/h within one and a half hours. This suggests that continuous rainfall or standing water for approximately one and a half hours is needed to saturate the top soil layer and establish steady state infiltration. The basic infiltration rate for clayey soil is, thus, determined to be around 6 mm/h, representing the predominant soil type across the study areas as a natural geologic formation.

Figure 4b demonstrates that the initial infiltration rate for sandy soil is notably high at around 450 mm/h, decreasing to about 20 mm/h within approximately 50 min as the soil saturates. Comparatively, the infiltration rate of sandy soil in the study area is over three times that of clayey soil. Approximately 30% of the total area consists of sandy soil, primarily due to land filled with predominantly sandy material, exhibiting a basic infiltration rate of 20 mm/h. This condition reflects an intermediary state resulting from ongoing land development activities in the study area.

From the infiltration curves shown in Figure 4a,b, Horton’s infiltration equations for clayey and sandy soils, respectively, have been derived as follows:

$$F=6+50e^{-0.02t}\left(\mathrm{Clayey}\mathrm{Soil}\right);F=20+430e^{-0.04t}\left(\mathrm{Sandy}\mathrm{Soil}\right)$$

F is the infiltration rate in mm/h and t is the time in minutes since the beginning of a rainfall event.

(ii) Infiltration depth

Annual rainfall exhibits variation across different climatic conditions, including wet, normal, and dry years. According to the statistical data, the annual rainfall measured 3030 mm in the wet year (2017), with values of 2003 mm and 1413 mm recorded during normal (2018) and dry (2020) years, respectively. Monthly rainfall distribution indicates no recorded precipitation in January and December (except 2018). Generally, rainfall is reduced between February and May, as well as from September to November. Conversely, the highest levels of rainfall occurred between June and August during wet years, ranging from 471 mm to 815 mm, while corresponding values for the same period in dry years range from 175 mm to 315 mm.

Infiltration levels exhibited variability across the study years, aligning with the prevailing climatic conditions and seasonal fluctuations. Under current circumstances, annual infiltration depths ranged from 1000 mm to 1860 mm, contrasting with values of 1074 mm to 2024 mm observed under natural conditions. Projections indicate a further decrease in infiltration depths in the future, with estimates ranging from 560 mm to 961 mm.

In 2017, 2018, and 2020, the soil received approximately 2024 mm, 1547 mm, and 1074 mm of rainfall, respectively, under natural conditions. These intake depths were 1860 mm, 1440 mm, and 1000 mm, respectively, under the present condition of the corresponding years. However, under the future condition (as planned for the areas), the intake depths would significantly decrease to around 960 mm, 766 mm, and 560 mm for the corresponding climatic years. Future infiltration depths were estimated by applying reduced permeability coefficients corresponding to proposed built-up and paved fractions, calibrated against measured infiltration rates for compacted fill soil. Figure 5 provides a visual representation of the variability in infiltration depth across various land use scenarios and rainfall years.

4.2. Runoff

(i) SCS curve method

Under natural land use conditions, annual runoffs were observed to be approximately 948 mm, 373 mm, and 177 mm for the years 2017, 2018, and 2020, respectively. Conversely, under current land use conditions, runoffs would amount to about 1080 mm, 439 mm, and 218 mm for the same years. Projected runoffs under future land use conditions are estimated to be around 1674 mm, 802 mm, and 447 mm for the corresponding years. Consequently, runoff depth could potentially increase by approximately 77%, 115%, and 153% for the wet, normal, and dry years, respectively, if the area adheres to the proposed development plan.

Analysis reveals that runoff depth exhibits notable variations, particularly during the months of June to August across all years, with the highest increase observed during wet years. Under current conditions, runoff depth ranges from 195 mm to 281 mm in wet years, whereas it could potentially increase to 277 mm to 409 mm under future land use conditions. Figure 6 provides a visual representation of the variability in runoff depth across various land use scenarios and rainfall years.

(ii) Water Balance method

The monthly depth runoffs are estimated through the Water Balance Method. Analysis indicates that annual runoffs under natural land use conditions are approximately 806 mm, 269 mm, and 203 mm for the years 2017, 2018, and 2020, respectively. Conversely, runoffs would amount to about 968 mm, 376 mm, and 274 mm under present land use conditions. Projections suggest that runoffs under future land use conditions would increase to approximately 1870 mm, 1050 mm, and 720 mm in the corresponding years. Consequently, runoff is expected to escalate by approximately 50% for wet, normal, and dry years if the area undergoes urbanization in line with the proposed plan.

The data highlights substantial disparities in runoff depth, particularly observed from May to August across wet, normal, and dry years, with the most significant increase occurring during wet years. Currently, runoff depth ranges from 167 mm to 221 mm in wet years, while potentially escalating to 258 mm to 483 mm under future land use conditions. Figure 7 provides a visual representation of the variability in runoff depth across various land use scenarios and rainfall years.

5. Discussion

5.1. Infiltration

The data analysis provides a comprehensive overview of the annual rainfall and infiltration depths within the study area across varying land use and soil conditions. The findings reveal the dynamic nature of annual rainfall, which fluctuates in response to different climatic conditions, notably wet, normal, and dry years. For instance, the wet year experienced a substantial rainfall measurement of 3030 mm, whereas in normal and dry years, the precipitation decreased to 2003 mm and 1413 mm, respectively. This variability underscores the influence of climatic variability on rainfall patterns, an essential consideration for water resource management and environmental planning.

Monthly rainfall distribution further elucidates temporal patterns, with no recorded precipitation typically observed in January and December. Rainfall tends to be less between February and May, as well as from September to November, while peaking between June and August, particularly during wet years. This observed seasonal variability underscores the importance of understanding intra-annual rainfall patterns for effective water resource management and land use planning.

Infiltration levels also demonstrate variability across the study years, reflecting the influence of climatic conditions and seasonal fluctuations. Under current circumstances, annual infiltration depths ranged from 1000 mm to 1860 mm, with natural conditions exhibiting slightly higher values ranging from 1074 mm to 2024 mm. Projections for the future scenario suggest a notable decrease in infiltration depths, with estimates ranging from 560 mm to 961 mm. This anticipated decline in infiltration highlights the potential impact of urbanization on soil permeability and water retention capacities, emphasizing the need for sustainable land development practices to mitigate adverse environmental consequences.

Moreover, the comparison of intake depths under different conditions further underscores the potential implications of urbanization on water infiltration. While natural conditions facilitate higher intake depths, the transition to urbanized environments leads to a significant reduction in infiltration capacity. These findings underscore the importance of incorporating measures to mitigate the adverse impacts of urbanization on hydrological processes, including stormwater management strategies and green infrastructure initiatives. Overall, this provides valuable insights into the complex interactions between land use, soil conditions, and hydrological processes, informing sustainable land management practices and environmental policy decision-making.

5.2. Surface Runoff

The results demonstrate that land use transformation exerts a significant influence on surface runoff dynamics in the study area. Under natural land use conditions, annual runoff depths were relatively moderate, approximately 948 mm, 373 mm, and 177 mm for 2017, 2018, and 2020, respectively. However, under current land use conditions characterized by expanded impervious surfaces, runoff depths increased to about 1080 mm, 439 mm, and 218 mm. If the proposed 2025–2035 land development plan is implemented without adequate hydrological safeguards, the projected annual runoff depths could rise dramatically to approximately 1674 mm (wet year), 802 mm (normal year), and 447 mm (dry year). This represents a nearly 50% increase in total runoff, signaling a substantial escalation in hydrological stress during monsoon periods.

Monthly analyses further reveal that the peak runoff occurs between June and August, coinciding with Dhaka’s monsoon season. Under current conditions, monthly runoff depths during wet years range from 195 to 281 mm, but projections for future land use indicate peaks as high as 277–409 mm. This sharp increase during critical rainfall months suggests that existing drainage infrastructure, already operating beyond capacity during events of 50–80 mm/h rainfall, will be increasingly inadequate. Consequently, without mitigation measures, urban flooding and waterlogging are likely to intensify, consistent with prior findings linking urban expansion to elevated flood risk in Dhaka and other South Asian megacities [34,36,41].

The Water Balance Method corroborates these results, indicating comparable trends in annual runoff depths. Under natural conditions, annual runoff was estimated at 806 mm, 269 mm, and 203 mm for wet, normal, and dry years, respectively. These values increased to 968 mm, 376 mm, and 274 mm under current land use, and are projected to rise further under future development scenarios. Together, both methods converge on the same message that urban expansion and imperviousness are amplifying runoff volumes across all hydrological years.

These quantitative findings hold significant implications for urban environmental management. A projected 50% rise in runoff depth is not merely a hydrological statistic—it translates directly into greater flood frequency, drainage failure, and reduced groundwater recharge, undermining the livability of Dhaka’s rapidly urbanizing eastern fringe. This aligns with earlier research emphasizing how unplanned development and infill of low-lying lands exacerbate urban flood hazards [25,34,41,42]. Importantly, the findings highlight that runoff management cannot be an afterthought in land development planning. Instead, there is an urgent need to integrate green infrastructure (e.g., permeable pavements, bio-retention systems, urban wetlands) and decentralized stormwater control measures into the proposed land use plan.

In summary, this study provides empirical, locally calibrated evidence that rapid urbanization (if unregulated) will significantly worsen surface runoff and flooding in Dhaka. The results reinforce the necessity of proactive land use control, stormwater-sensitive urban design, and hydrologically informed planning policies to enhance infiltration, manage runoff, and safeguard the resilience of Dhaka’s future urban landscape.

5.3. Limitations of the Study

The study acknowledges several limitations that also point toward future research opportunities. First, the absence of continuous discharge and spatially detailed catchment data constrained the estimation of peak runoff rates and storm hydrographs, limiting insight into drainage and waterlogging dynamics; future work should integrate continuous flow monitoring and account for other recharge mechanisms such as leakage and lateral inflows. Second, as runoff was estimated using theoretical methods without direct field measurement, subsequent studies should include empirical runoff observations to strengthen model validation. Finally, the use of the project area as a proxy for catchment boundaries and the lack of an updated citywide catchment database highlight the need for refined boundary delineation and the development of integrated rainfall–runoff models linking future land use scenarios with spatially distributed hydrological simulations.

6. Conclusions and Implications

The study investigates the dynamic interplay between land use changes, and hydrological processes within the study area. The infiltration reveals significant fluctuations in annual rainfall patterns across climatic conditions, emphasizing the importance of intra-annual rainfall distribution. Additionally, the variability in infiltration levels underscores the potential impact of urbanization on soil permeability and water retention capacities. Furthermore, the study explores the impact of land use changes on runoff depth. The increased runoff depths, particularly in wet years, heighten vulnerability to flooding and waterlogging under future land use scenarios. The findings add value to the existing knowledge on the adverse impacts of urbanization on hydrological processes.

The study has great implications for adopting sustainable land development practices for effective water management and land use planning for megacities like Dhaka. It could also imply strategies for stormwater management and green infrastructure, imperative to mitigate the escalating risks of urban flooding and ensure the resilience and sustainability of the city’s hydrological systems in the face of increasing urbanization. By prioritizing sustainable land management strategies and adopting resilient infrastructure solutions, urban policies can effectively manage the challenges posed by land use changes and climatic variability, safeguarding the long-term health and functionality of urban hydrological systems. Permeable pavement, an alternative to traditional asphalt and concrete surfaces, can be an option to allow water to infiltrate, thus reducing runoff and acting as a filter. It allows storm water to infiltrate into either a storage basin or exfiltrate to the soil and recharge the groundwater table, while removing pollutants. For Dhaka, it designates areas for retention ponds and polders which should be reserved, and multiple use of it is encouraged.