Interdisciplinary integration of land use and drainage planning for urban adaptation under the climate change scenario

In the past few decades, the impact of urbanization on watershed hydrology has been investigated, and the increase in urban storm runoff is directly proportional to the impervious area in the watershed (Parkinson 2003; Shuster et al. 2005; Dietz & Clausen 2008; Qin et al. 2013; Liu et al. 2015; Leimgruber et al. 2018; Bai et al. 2019; Feng et al. 2020; Andualem et al. 2023). Flood wave movements become more concentrated after replacing the natural drainage network with man-made streets and sewers (Guo 2003). The runoff volumes and peak flows increase because soil infiltration is significantly reduced due to more pavements (Elliott & Trowsdale 2007). To mitigate the negative impacts of developments on the urban water environment, urban planners need to conduct a regional master drainage plan in which the future land use plan can include an effective stormwater drainage system to mitigate potential flooding problems from extreme storm events (Brabec et al. 2002).

The above descriptions address a significant problem not only globally but also locally. An island country, Taiwan issue, located in East Asia, is threatened by flash floods (Chang et al. 2021; Guo et al. 2023) and needs to formulate a suitable method to prevent urban flooding. The overloading concept in stormwater management has two strategies: First, the distribution of flow loading according to watershed area size should be controlled downstream at the junction, and the excessive runoff should be allocated to avoid flood disasters (Guo & MacKenzie 2014); second, the control of flow releases expecting to recover the function of water resource retention, to increase permeation, and to reduce runoff, water conservation areas like depressions and natural lakes were constructed (Emerson et al. 2005). The previous studies aimed to determine available master drainage planning (MDP) in cities. Conventional studies that aim to solve the on-site drainage system associated with a small sub-area are often designed using the rational method. The network of MDP is organized according to three major process categories: (1) river-based MDP, (2) city-based MDP, and (3) site-based MDP. These categories represent impervious area, peak flow, infiltration, and runoff volume processes or mechanisms.

A regional MDP consists of three elements, including I-design using LID devices at the upstream points of runoff sources and Q-design using streets-sewers-ditches to collect and convey runoff flows to the watershed outlet effectively. At the outfall point, V-design uses wetland, retention, and detention facilities to reduce the release of flows into the downstream receiving water bodies. An MDP provides an overall assessment of various alternatives to use the combinations of I-, Q-, and V-designs to lay out LID devices upstream, drainage systems through the watershed, and detention systems downstream. The alternative studies aim to minimize the costs and maximize public safety.

Taiwan faces an extremely high risk of flood disasters due to abundant rainfall. Moreover, with a dense urban population, numerous flood control projects have been constructed. Building large-scale flood control projects has become challenging given the current land use patterns. Therefore, utilizing LID in conjunction with micro-detention effectively reduces peak discharge and decreases the area and depth of flooding. This study aims to provide an adaptation strategy based on the distribution of flow loading and control of flow releases approach of watershed-integrated stormwater management. A disaster hotspot, the Dapu area, located in Chiayi County, Taiwan, is adopted as the research area. Firstly, hydrological practical and land use development processes are considered. Then, the StormWater Management Model (SWMM) is embedded to simulate the runoff flow in the Dapu area. Under the above framework, three sets of plan methods are separately produced for (1) river-based MDP, (2) city-based MDP, and (3) site-based MDP. Finally, a reliable result can be suggested as an improved decision-making approach. MDP will provide a quantifiable basis to develop future management plans on land uses and infrastructures for stormwater and transportation systems under different precipitation scenarios (slight, medium, and climate change). In conclusion, the highlight of this study is to integrate a framework for urban design and hydraulic engineering and provide a practical approach to MDP.

Similarly, an MDP must be developed for each city. The city's significant drainage method should be treated like a river. At all design points on the drainage route, the design flows should have been defined by the city-based MDP. In case any site has been overdeveloped, the outfall point of the site needs a mitigation plan to reduce the flow release to stay within the allowable release. A site should have a small-scale, detailed MDP that defines the flow path, location of low-impact devices, etc. The MDP network is integrated with multiple levels of details, including at least three levels: (1) river-based MDP, (2) city-based MDP, and (3) site-based MDP. An overall effectiveness of MDP is the enforcement of MDP from all individual sites, cities, and the entire river basin. All levels of MDPs define the design flows at all design points. These design flows set the level of safety as to whether we need a detention basin for runoff flow reduction and porous low-impact designs for runoff volume reduction. As a rule of thumb, a river-based MPD can be numerically modeled using the kinematic wave approach without consideration of backwater effects (Guo et al. 2012). A city-based MPD shall consider the backwater effect of the river. Therefore, a dynamic approach is preferred. For a site-based MPD, the construction plans shall provide detailed flow paths and the location of low-impact facilities such as porous pavers, gardens, infiltration beds, etc.

After developments, the natural waterway is replaced with the significant sewer trunk line along a collector street, while the incoming creeks are replaced with the branch sewer lines. Urban overland flows are generated from adjacent paved areas along a street. Most of the natural depressed regions and lakes are filled or leveled. Of course, most of the previous surface areas are converted into impervious areas such as roofs, parking lots, streets, etc. As aforementioned, the mitigation plan for increased stormwater includes LID devices as a replacement for natural depression and a regional detention basin as a replacement for raw storage such as pools and lakes.

Urban development is always staged and random in time and space. As illustrated in Figure 3, the continuous development from Stages A to D requires early land use planning to accommodate the spatial and temporal needs in the MDP. For instance, which streets should be selected to lay out the major sewer trunk lines, and where should land dedications for future LID or detention facilities be? All these details must be planned ahead of random urban development. For an urban renewal project, the old town may need more flexibility to accommodate the flooding problems. An MDP has to be developed with various alternatives to mitigate the flooding problems to an acceptable risk level.

The urban drainage system carries cascading flows from micro, minor, and major systems. An MDP is a spatial analysis used to optimize land use. For instance, how to select the best strategic location for a city park so that it remains open for recreation during dry days and is reserved for flow flood detention during wet days, how to use the available space along waterways as greenbelts, and how to convert floodplains into bike paths and waterfront parks.

The stormwater systems, composed of major, minor, and micro facilities, are critical, as shown in Table 1. A microsystem is designed to intercept the 3- to 6-month frequent events for water quality enhancement. Examples of micro facilities are LID devices, including rain gardens, infiltration beds, porous pavers, etc. When the microsystem is overtopped, the excess water flows into the downstream minor system, which includes street inlets, underground sewers, roadside ditches, etc. A little system is designed to pass the 5- to 10-year events. After the minor system becomes full, water will overflow into the significant systems, such as street gutters, flood channels, natural waterways, etc. All primary drainage routes must be capable of passing the 50- to 100-year events. The application method will be presented in the following chapter.

The hydrologic impact of urbanization is an integration of watershed hydrologic cycles and human activities (Sheng & Wilson 2009). Most studies on urban watersheds have addressed various methodologies to simulate urban stormwater generation and movements (Galster et al. 2006). Still, the fundamental concern is how to conduct RDMP (regional drainage master planning) to quantify the hydrologic changes in watershed regimes. In practice, an RDMP is performed using computer numerical simulation. The most popular computer models are the US EPA SWMM5 and US COE HMS (Huber et al. 2005; Chu & Steinman 2009). The input parameters include design rainfall information, watershed hydrologic parameters, street and sewer network, and hydraulic features, including culverts, LID devices, detention basins, etc. In practice, the existing watershed model should be developed to reflect the current condition, while the future watershed model should be designed according to the future fully expanded condition. To evaluate the impact of a specific development project, the pre-and post-development models must be developed to identify the changes with and without the proposed project. A mitigation plan for the changes should be developed using alternative models to minimize the cost and maximize the level of safety. Various alternatives can be created with different combinations of drainage features.

Taiwan is a small island in the South China Sea, spanning 36,000 km². Despite its size, the island has high mountains that reach peaks of 2,000–4,000 m, running through its center. Due to the island's geological and climate conditions, it experiences frequent and significant natural disasters, such as earthquakes, landslides, and floods caused by typhoons. This study focused on the Dapu township, situated upstream of Tsengwen Reservoir, and examined its MDP levels. The Tsengwen River basin is a crucial reservoir water source, the sole water provider for the southwest region of Taiwan. This area has a population of approximately 2.67 million people. The terrain in the Dapuurban planning area slopes generally from east to west, with ground elevations ranging from 230 to 410 m. The average slope is 3.47% west and 37% east. The minimum temperature in January is about 16.16 °C, while the maximum temperature in July and August is 35.42 °C.

This study adopted the Matou Shan rainfall station situated at an elevation of 1,020 m in the Tsengwen Reservoir catchment area. According to the statistics from 1969 to 2022, the mean annual rainfall at Matou Shan is 2,817 mm; the maximum annual rainfall is 5,046 mm. The catchment area experiences a dry season between October and April each year. It has been observed that the precipitation in winter has a slight impact on the catchment area. Heavy rainfall, with more than 100 mm of daily rainfall to the Tsengwen Reservoir catchment, is often observed in the summer. The average rainfall varies monthly, with the highest in August, July, and June. The maximum daily reached 695 mm, and the 3-day maximum rainfall was 1,352 mm. The Dapu Township Urban Plan aims to achieve its target year of 2015. The plan includes a land use zoning plan that projects a population of 10,000 and a residential density of 220 people per hectare (ha). The project area covers a total of 203.67 ha.

The tributary area to the Tsengwen River has been developed into several small mountain towns that have caused tremendous increases in runoff flows and vast amounts of landslides and bank erosion upstream of the river. The deterioration of the Tsengwen River directly impacts the Tsengwen Reservoir. Tsengwen River must be preserved with its design flows because all bank protections and drop structures along the river were built according to the design flows stated in its MDP.

Proper land use planning is imperative during Stage D. The paper highlights the connection between land use and MDP in watershed regions. It also proposes adopting the LID approach at the upstream street corner to minimize runoff volume and implementing a detention basin at the downstream outfall point to decrease peak flow. This holistic strategy can aid urban planners in effectively mitigating floods. Moreover, the paper offers valuable perspectives on MDP practices.

It has been determined that the H1M250 station's proximity to the local climate and rainfall pattern is significant, so a chosen set of rainfall data from several representative areas in Taiwan's City of Dapu was analyzed. The data was collected from the Water Resources Agency and spanned an average of 25 years. A line chart displayed the cumulative distribution of rainfall events from small events to 100-year events over 48 h. The distribution was found to be expected. The statistics and parameters for high peak flow were also analyzed, including 100 years = 141 m³/s, 10 years = 76.4 m³/s, 2 years = 61.7 m³/s, and small event = 12.3 m³/s.

The Tsengwen River is an important waterway that flows through Alishan Mountain and eventually empties into the Tsengwen Reservoir. The framework demonstrates that V-design uses retention and detention facilities to reduce the release of flows into the downstream receiving water bodies. To check the stagnant flood volume and flood capacity of the base drainage road at the development site, the outer water level will be used as the downstream water level boundary condition. The extreme water level record line can be calculated from the external drainage discharged into the regional drainage or the river at each recurrence of the flood level calculation or the outer drainage road known water level of the cross-section. According to the 2000 Taiwan River Database, the 100-year design flow rate at Dapu in the Tsengwen River is 15.8 m³/s. This means that the maximum amount of water that can be released from the Dapu area is 15.8 m³/s. As a result, this study aims to establish a Maximum Permissible Discharge (MPD) for Dapu, which ensures that the 100-year peak flow does not exceed 7.62 m³/s. To calculate the peak flow of each recurrence interval at the outflow of the development base, a one-dimensional hydrological calculation of the joint outflow into the regional drainage or the familiar outflow road at the river will be conducted. The calculation method of the peak flow of each recurrence interval of the familiar outflow road, the intercepting waterway, and the crossing waterway will be explained. This calculation will be used as the basis for verifying the flooding capacity of the watercourse and calculating the outer water level.

The paper proposes an MDP that has three main parts. Firstly, it calculates the peak flow at each recurrence interval of the joint outfall, the interceptor, and the crossing. Secondly, it calculates the outer water level history. Finally, it calculates the discharge flow of the development base, ensuring that it does not affect the flooding capacity of the downstream joint outfall. When designing a microsystem to handle a 5- to 10-year return flood, it is essential to consider the pumping station's capacity, the downstream culvert section, and the availability of proper outlets for the drainage trunks simultaneously (Hsu et al. 2000; Cheng et al. 2015). However, due to the lack of measured flow data for storm sewer systems in Taiwan, only the depth of inundation is available, which makes it impossible to directly verify the accuracy of the design model incorporated into the design rainfall pattern, given that the SWMM model has been validated for the depth of inundation in many areas in Taiwan, its credibility is acceptable (Huang et al. 2018; Lin et al. 2021).

Taiwan's Department of Water Resources (DWR) offers a simplified method to calculate the impervious surface ratio (IMP) for different land use zones. The study area's underlying surface was classified into greenspace, building, and road. Greenspace is considered a pervious surface, while buildings and roads are impervious. The imperviousness of each sub-catchment is calculated as the percentage of building and road areas: ⁠.

The Q-design uses streets, sewers, and ditches to collect and transfer runoff effectively to watershed outlets and at the outfall point. The effective collection and conveyance of runoff flows through streets, sewers, ditches, and canals converge to small flood storage ponds. These ponds are usually covered with concrete to store water, but they do not reduce the volume of water. Instead, they only change the drainage distribution over time, which means they only reduce the flow rate. When designing the roadway in urban planning land use detail plans, designers need to refer to the ‘Highway Drainage Design Guidelines’ to determine the scope of the study for the design of drainage facilities. Designers can then make appropriate decisions on the size of the drainage facilities, taking into account the risks and economics involved.

The average KW plane slope (S_w) for the study area is 4.3%, and the average width of the KW plane (_LW) is 115.2 m. The infiltration rate for the sand mix used in this tributary area was reported to be 12.5–75 mm/h. The detailed hydrologic parameters are summarized, including pre-post watershed imperviousness, length, S_w, LW, and soil infiltration rates. Using the impervious percentage of 15%, the 100-year peak flow must be reduced from the post-development flow release of 7.62 to 3.78 m³/s. In this case, setting the control of flow releases to be the peak outflow under the pre-development condition, the after-detention hydrograph is approximated by a linear rising hydrograph (Guo 2006), as shown in Figure 8.

City-based MDPs were integrated to explain the peak flow reduction by detention ponds. The model determines detailed investigations by applying long-term runoff simulations to assess the needed stormwater quantity capture volume for a site, stormwater drainage system, municipality, or region. The control of flow releases analysis results reveal that the land use layout is linked with the drainage master plan. This method was improved using SWMM and GIS to determine the future development and road structures most closely associated with existing city-based detention ponds. This analysis primarily considers the economics of each detention facility since hydraulic facilities will cost more to infrastructure and occupy large tracts of land in the future.

According to the stage area storage for the proposed detention basin described in Table 2, the stage–storage curve is developed using triangular cross-sections. The 10-year and 100-year water depths in this basin under design are identified to be 1 and 3 m. The required stormwater detention volumes were 8,014 m³ for the 100-year event.

The primary purpose of the I-design is to enhance the quality of the water environment. It achieves this by improving the ability of runoff to infiltrate the ground surface in combination with an LID facility. This helps to prolong the time it takes for the runoff to collect and reduces the amount of pollutants that can contaminate the water supply through stormwater runoff.

The flow loading distribution represents the site's value to be developed with the micro facilities, including the LID devices, infiltration beds, and porous pavers. This paper modifies integrated land use and drainage planning; the two flow paths can be identified, as shown in Figure 11.

This case introduces a pervious outlet to drain the overland flow from the impervious area to the pervious area. In this case, setting the peak outflow to the peak outflow under the pre-development condition, the after-LID hydrograph is approximated by a straight line, which means that the LID devices are set to reduce the flow to enhance the water quality further.

The primary function of Type I-design is to improve the quality of the water environment, together with LID facilities to increase the capacity of runoff to infiltrate the surface, and to lengthen its collection time, in order to minimize pollutants from stormwater runoff to contaminate the water source, as shown in Figure 6. In terms of the base, the ordinances that specify the amount of rainwater storage or runoff from the building site, for example, the standard of the amount of runoff from the base of the development of the city of Taipei City that discharges into the sewers. The New Taipei City Urban Plan stipulates that rainwater storage and culvert water reuse related facilities should be set up to apply for the code of practice or the Central Statute – Architectural Technical Rules, Article 4-3, which are mainly for the purpose of disaster mitigation and flood prevention, and stipulates that the building base should be designed with the base amount to achieve the effect of disaster mitigation and flood prevention, and the Ministry of the Interior, Department of Construction, can refer to the ‘Low-impact Development Facilities in the Water Environment Operational Manual and Case Evaluation Plan’ as the reference for the future land use control. The Ministry of the Interior's Department of Construction's ‘Water Environment Low-Impact Development Facility Operation Manual and Case Evaluation Program’ can be a reference for future land use control.

Planning for land use and master drainage can be a highly effective way to adapt to flood control. Detention areas, like schools and parks, have been found to mitigate flood risks due to MDP. Stormwater management has become a significant concern in many cities, leading to an increased focus on the distribution of flow loading and control of flow releases. To determine the best placement of LID devices, this study has developed hydrological analytical techniques using SWMM-based analysis to create a land use layout. This paper aims for MDP to be developed for each city. At all design points on the drainage route, the design flows should have been defined by the City MDP. The network of MDP is integrated with multiple levels of details, including at least three levels: (1) river-based MDP, (2) city-based MDP, and (3) site-based MDP. This model described in this study reveals the differences in the hydrologic parameters in each tributary area. The results of the SWMM model simulate the effects of pre-development conditions on progressive stages in watershed development.

Previously, the MDP study only focused on flood flow calculations before and after site development without considering the relationship between land use and flood flow. This was achieved by estimating the flood flow based on the SCS uncalculated unit calendars. However, a new approach to MDP has been proposed, introducing a refined river, city, and site based on MDP. This new approach gives urban planners and water engineers a more comprehensive understanding of the relationship between land use and flood flow. It enables them to engage in a productive dialogue, facilitating better decision-making and more effective flood management strategies.

The river based on MDP considerations, uses wetlands, retention, and detention facilities to reduce the release of flows into downstream water bodies. An external water level record line is maintained from joint shallow drainage discharges into the regional drainage or river at intervals to check stagnant flood volume and the drainage road flood capacity of the downstream water level boundary conditions.

A city based on MDP considerations, the Q-design utilizes streets, sewers, and ditches to collect and convey runoff flows to the watershed outlet effectively. To ensure the flood capacity of the waterway and downstream water level boundary are within acceptable limits, it is necessary to calculate the flood peak flow for each recurring distance of the joint external drainage road, the intercepting waterway, and the traversing waterway. Q-design aims to improve the quality of the waterway.

The site based on MDP considerations and design aims to enhance the quality of the water environment. This is achieved through LID facilities, which help increase the capacity of runoff to infiltrate the surface. This, in turn, helps improve the water environment's overall quality. My design approach focuses on minimizing development's environmental impact and promoting sustainable practices.

Strategizing for land use and master drainage is a powerful means of adapting to flood control. Integrating detention areas such as schools and parks has proven effective in mitigating flood risks through MDP. Stormwater management has become a pressing issue in numerous cities, highlighting the need for increased focus on the distribution of flow loading and control of flow releases. To optimize the placement of LID devices, this research has devised hydrological analytical techniques utilizing SWMM-based analysis to generate a land use layout. This research employs a land use model and explores a runoff volume-based method that converts the effective imperviousness ratio into volume-weighted effective imperviousness and perviousness for two sequential plans. This approach is valuable in simulating impervious and pervious outlets for LID designs using an on-site MDP basis.

Financial support for this study was partially provided by a grant from the Ministry of Science and Technology (project number: 111-2625-M-035-007-MY3; 112-2625-M-033-001).

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.