Mitigating Perennial Floods in Cities of the Global South: A Nature Based Approach

doi:10.21203/rs.3.rs-3587407/v1

Download PDF

Research Article

Mitigating Perennial Floods in Cities of the Global South: A Nature Based Approach

https://doi.org/10.21203/rs.3.rs-3587407/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The incidence of flooding in cities of the global south has become perennial with increasing urbanization, showing a lack of targeted approach to the challenge. The objective of this research was to develop a resilient and an integrated approach to stormwater management that could be used to inform policy. The Research for Design model was adopted using the Greater Accra Metropolitan Area (GAMA), Ghana as a case. Land cover types for three focal areas; Area1, Area2, and Area3, respectively 10.2 km², 10.2 km² and 10.5 km² were extracted and combined with daily rainfall values to estimate direct runoff, using the Curve Number method. Eight [8] landcover types were extracted based on which percentage imperviousness was determined for Area1 (54.1%), Area2 (43.66%) and Area3 (22%). Area3 had more large roofs (≥ 81m²) averaging 248.5 m² compared with Area2 (195.6 m ²) and Area1 (182.2 m²) but did not significantly impact runoff generation at 5%. Harvestable roof runoff for both small roofs (< 81m²) and large roofs (≥ 81m²) contributed 32.6% in Area1, 22.5% in Area2 and 14.7% in Area3, to runoff generation per rain event. The results suggests that retaining roof runoff at the plot level could significantly reduce stormwater runoff. To demonstrate this Living Labs concepts was combined with various design methods, aided by a schema to integratively design a selected site for roofwater harvesting for irrigation, storage for non-portable uses, infiltration to recharge groundwater and to create amenity values. The research contributes to knowledge by proposing a quantitative basis to target stormwater management at the plot level.

Curve number

Flooding

Harvestable roof runoff

Living Labs concept

runoff estimation

Synthesis-analysis approach

Research for design

Worldwide there is increasing trend towards urbanization and present projections show the trend will continue into the future where developing countries will see even greater increases in extent of urbanization. Literature agrees that increased urbanization has led to fragmentation of urban landscapes resulting in less infiltration but more stormwater generated runoff due to increased impervious surfaces, and reduced time to peak (lag time) where run off generated from impervious surfaces quickly accumulate leading to floods (Allen, 2013, p. 16). The role of urbanization in flooding in urban areas cannot be overemphasized and there is a troupe of research to support this. The primary effect of urbanization is the loss of natural places of attenuation and "evapotranspirational functions of soils and the vegetation". The result is that stormwater runoff which could have been stored in the natural places of attenuation and slowly infiltrated, or trapped by vegetation and transpired or evaporated, mostly "flows" quickly across the urban landscape arriving at discharge points such as streams, storm drains in large volumes with high energy (NRC, 2009, p. 5). The high discharge of stormwater is quickly conveyed by hydrologically efficient drainage systems which are not designed for that volume, causing the drainage system to overflow or even fail and thus discharge into nearby communities and settlements. There are instances where the drainage system which may have been filled with improperly disposed solid waste becomes silted up, and due to the reduced capacity, overflow or become completely blocked, resulting in inundation of surrounding communities and settlements. The argument that there is increased frequency in flood events because of climate change is supported by data which has been used to make projections into the future. This climate change projections show that in some parts of the world increased variability will lead to more serious floods which will increase in intensity, affecting new areas which previously were flood-free. At the same time in other places the increased variability will result in less rainfall and also less surface water flow. Increased variability could also result in more erratic rainfall which may produce flush floods in urban areas. It is estimated that by 2030, about 40% of global urban land will be located in high-frequency flood zones compared to 30% in 2000 (Güneralp, 2015) cited in (Itsukushima, 2018, p. 1) putting development, life and property in those areas at risk. Needless to say, the causes of floods is multi-dimensional(C. Y. Okyere, Yacouba, & Gilgenbach, 2013) and in an urban area in a developing country it is fittingly described as a wicked problem. But it is clear that urbanization in an urbanized age will continue to produce more runoff than the current drainage efficient systems can accommodate, resulting in floods in communities and settlements. It is also projected that the situation could become worse with climate change influenced variability producing more rainfall and thus more runoff than the existing systems can accommodate. Expressed another way, the current system of stormwater management to control floods in urban areas is not working; it's not adequate nor efficient (Ouikotan, van der Kwast, Mynett, & Afouda, 2017). This demands a smatter system which will be more sustainable to control the continued cycle of perennial floods in urban areas.

Gap analysis

In most developing countries runoff or stormwater (Odefey et al., 2012) generated from urbanized surfaces is managed using hydrologically efficient systems, the implementation of which has been adhoc (Bhattacharya-Mis, 2011; F. Frick-Trzebitzky, 2018; NRC, 2009; Ouikotan et al., 2017 a. Tengan & Aigbavboa, 2016); a piecemeal approach not based on science, and rather making worse the problem of floods. Indeed, the system in the face of climate change is making urban areas less resilient against floods, seen in increasing incidence (Jha, 2012; Ouikotan et al., 2017) as confirmed by several authors (CRED/UNISDR, 2016; Jha, 2012; Wallemacq, 2018; Yaghmaei, 2019). The absence of locally generated research content focusing interventions to address this problem has created a situation where often what is proposed in policy documents is not matched by what is implemented on the ground to address the challenge of stormwater management (F. a. B. Frick-Trzebitzky, Antje, 2017). This research was developed to address these gaps by proposing a resilience and an integrated approach to stormwater management. (Adger, 2006, pp. 268–270; Gupta, Singh, Wajih, Mani, & Singh, 2017, p. 35). An integrated approach to stormwater management to achieve flood control, according to (Adger, 2006; Gupta et al., 2017; ISWM, 2006) uses a design approach to combine drainage with runoff quantity and quality control.

This research is thus premised on the fact that policy documents guiding stormwater management within GAMA and other parts of Ghana has not been effective in addressing the challenge based on the gaps identified. This research adopts Research for Design approach which is a practice based research that rely on locally generated scientific data to inform design interventions to address the challenge of perennial floods(Cross, 1982; S. D. Lenzholzer, I and Jusuck Koh, 2013). Research for design is a research based approach which depends on knowledge generation (on stormwater) to feed and guide every activity in the design process to increase its reliability(Frayling, 1993; S. D. Lenzholzer, I and Jusuck Koh, 2013; S. D. Lenzholzer, I and van den Brink, Adri, 2017). The approach results in the creation of a designed artefact where every decision in the design process has to be explicit and supported by rational arguments on how the artefact was arrived at(Linden, 2001).

Study area as Case Study

The Greater Accra Metropolitan Area (GAMA) was used as the case study; a densely populated urban area in Accra-Ghana (Fig. 1). The choice of GAMA was guided by the fact that Ghana seem to be better organized in terms of data availability, relative to the rest of the sub region(F. Frick-Trzebitzky, 2018) and because of the researcher’s familiarity with the area. GAMA is made of 15 administrative areas (ARUP, 2016)but the research was conducted in 5 of these administrative areas. These five administrative areas lies within Longitude 5.804253 and 5.492637 West and Latitude 0.527292 and − 0.082525 North, covering a total land mass of about 891 sq km. The climate is Coastal Savannah with two rainy seasons of unequal intensity, averaging 730–800mm per annum. The soils in the area have developed on thoroughly weathered parent material with alluvial soils and eroded shallow soils (Oppong-Anane, 2006, p. 4).

The population density of the 5 administrative areas, Ga East, Ga West, Ga South, Ga Central and Accra Metropolitan Area is 1,238.8 persons/km², far higher than the national average of 103.4 (GSS, 2014). The landform is characterized by low lands with occasional hills averaging 20 m above sea level and a water table that varies between 4.8 m and 70 m below surface. Slopes are gentle, below 11 percent except in a few places where they are above 22 percent (Nyarko, 2000).

Design Methods and Strategy

Design methods were selected to chart well defined steps that guided the research(Yüncü, 2008) and to integrate the research findings into the design process. These included the application of the concept-test, optimizing essential functions, behavior patterns, analysis-synthesis (Alexander, 1964; Jones, 1992; Lynch, 1971; Rowe, 1982) and the Living Labs concept (Mats; Eriksson, Niitmano, Kulkki, & Hribernik, 2006; R. Bernhard; Katzy, S. Kulwant; Pawar, & Klaus-Dieter Thoben, 2012). The strategy was Research for design which used research as a vehicle to enrich and inform the design to ensure effectiveness (Cross, 1982; S. D. Lenzholzer, I and Jusuck Koh, 2013; S. D. Lenzholzer, I and van den Brink, Adri, 2017; Yüncü, 2008).

Data Types and Sources

The different data types and their sources are summarized in the figure below (Fig. 2)

Data Collection Process and Analysis

Data collected was both spatial and non-spatial and included land cover, rainfall, soil infiltration characteristics to model Runoff. The non-spatial data were obtained through site visits, interviews and information obtained through literature and included site descriptors on soil (eg presence of hard pan), historic information, some unique features on buildings in certain areas. Some of these non-spatial data were made spatial by associating them with various geographical elements using GIS techniques(Osman, Nyarko, & Mariwah, 2016). The spatial data included buildings, roads, streets (paved and unpaved), paved surfaces, bare areas, green vegetated areas.

Land cover extraction - Three focus areas were selected within the research area to capture variations in land cover types (Fig. 3). The focus areas and their respective sizes are; Area1 (10.2 km²), Area2 (10.2 km²), and Area3 (10.5 km²). Land cover classes within the focus areas were identified and extracted from 20 cm pixel 2014 Orthorectified aerial photographs of the research area, aided by a good knowledge of the research area obtained through field visits by the researcher (Justice, 2002; Stuebe, 1990, p. 612). The land cover types identified and extracted was based on an adaptation from the Climate Change and Urban Vulnerability in Africa (CLUVA) project (CLUVA, 2012).

The orthophoto images were obtained in tiles and were mosaicked together using the Erdas imagine software. Extraction of land cover was partly automated using the Feature Analyst software where the software was trained and used to extract the roofs of buildings and partly manual by digitizing the land cover types (Plates 1a, 1b,1c, 1d, 1e) using GIS techniques (Chabaeva, 2004, p. 2; Verbeeck, 2011). This methods of extraction actually allowed for more precise results to be obtained (Chabaeva, 2004). Un-extracted land cover types identified as ‘left over space’ (LOS) were visually assessed and classified as (LOS) bare, paved or vegetated based on the pre-dominate cover type, flood maps. The land cover types were used to estimate the proportion of each land cover as a percentage of the land area and imperviousness which is the percentage of the cover types made of impervious surfaces. The data collected was subjected to simple statistical analysis using central tendencies (mean and median), analysis of variance and the Linear Mix Model (REML) to determine the extent of the relationship between the land cover types in a Genstat environment.

Infiltration characteristics of soils: Soils at the study area (at the series level) were determined using soil maps which was used to guide a small scale field infiltration test as detailed in (Asiedu, 2020). Results of the field infiltration tests were used to categorize the soil series into Hydrologic Soil Groups (Table 1) using the USDA Hydrological Soil Group (HSG) classification method (Asiedu, 2020; United States Department of Agriculture, 2009).

Relating Land cover to HSG - The choice of CN factor values for each of the land cover types were based on HSG for the different soils of the three focus areas (Table 4) as described in (Asiedu, 2020). Area2 and Area3 were assigned different HSGs for the different soil series with corresponding CN factor values. Area1 had only one soil series and thus had a single HSG with a corresponding CN factor value. The composite approach where all hydrological soil groups in a focus area were considered yielded a composite CN value of 93 for Area1, 89 for Area2 and 83 for Area3 (Table 2).

Table 2

CN factor values for Landcover types based on HSG in focus area
Land cover type	Area1 HSG	Area2 HSG	Area3 HSG	Curve Number*	Sources
Rooftop Paved Street Pavement	C C C	B, A B, A B, A	A, B, C A, B, C A, B, C	98 98 98	(ISWM, 2010)
PC Building LOS (paved)	C C	B, A B, A	A, B, C A, B, C	98 98
Unpaved Street Baresoil LOS (bare) Playfield (bare)	C C C C	B, A B, A B, A B, A	A, B, C A, B, C A, B, C	Dirt road Fallow, Bare Fallow bare Fallow, bare	(ISWM, 2010) (USDA-NRCS, 2004a)
Backyard Farmland	C C	B, A	A, B, C	Cultivated land Cultivated land	(ISWM, 2010)
Trees UndeVeg	C C	B, A	A, B, C	wood/orchard wood/orchard	(USDA-NRCS, 2004a)
LOSVeg	C	B, A	A, B, C	Herbaceous, Poor & Cultivated poor	(ISWM, 2010; USDA, 1986)
Notes: PC Building – partially completed buildings; UndeVeg – Undeveloped vegetated land; LOSVeg, LOSpave, LOSbare – left over space covered with vegetation, pavement or bare surface, respectively. Curve Number* - apart from impervious areas like roofs and pavements whose CN value does not vary with HSG, the cited references for the other cover types have CN values which vary with HSG.

Rainfall Data

Daily rainfall data was collected from raingauge stations closest to the three selected sites; Accra Academy, Pokuase, and Mempehuasem, for Area1, Area2 and Area3 respectively as suggested by (Barbosa, 2012, p. 6791; Critchley, 1991, p. 3.4). These were used to estimate direct runoff at a given return period. A 44-year data period from five rain-gauge stations within the study area (Accra Academy, Pokuase, Mampehuasem, Weija and Korlebu stations) were also used to calculate the return period. Seven years daily rainfall data between 2006 and 2015 was obtained from Ghana Meteorological Agency in Accra and used for the runoff analysis.

Runoff Estimation

Runoff estimation was based on the Curve Number model and was used following guidelines in the National Engineering Handbook (United States Department of Agriculture, 2004b; USDA-NRCS, 2004a), supported by Integrated Storm water Management manual (ISWM, 2010) and Dile, et. al., (Y. T. K. Dile, Louise; Srinivasan, Raghavan, and Rockström, Johan, 2016). The Curve Number (CN) is a very popular model for direct runoff estimation under wide geographic and climatic conditions (T. Y. K. Dile, Louise; Srinivasan, Raghavan and Rockström, Johan, 2016; Ponce, 1996). The model is popular for its simplicity, low data requirement; especially important for a data-poor developing country like Ghana (Dingman, 2002, pp. 443–445). Its adaptability, extensive usage over many geographical areas, and high level of sophistication (Y. T. K. Dile, Louise; Srinivasan, Raghavan, and Rockström, Johan, 2016, pp. 1155,1156; Dingman, 2002, pp. 443–446; Grove, 1998, p. 1016; Ogden, 2017, pp. 49 − 41; Oliveira, 2016, p. 420; Ponce, 1996) has made it a preferred model especially in areas where information on antecedent conditions is not available (Dingman, 2002). Its main weakness though is its inability to make any distinction between actual retention / potential retention and actual runoff and potential runoff (T. Y. K. Dile, Louise; Srinivasan, Raghavan and Rockström, Johan, 2016). The model is also sensitive to spatial differences in rainfall within a catchment (Ponce, 1996). To reduce these weakness, data from closest rain gauge stations to the focus areas were used. Runoff analysis was based on daily runoff greater than 12.7 mm (0.5 inches) as stated in the United States Dept. of Agriculture, Urban hydrology document (USDA, 1986, pp. 2–11). Also rainfall ≥ 40mm was used to reduce error seen in very high direct runoff estimates as a result of small rainfall levels(Grove, 1998). Figure 4 summarizes the entire process. The data collected was subjected to simple statistical analysis using central tendencies (mean and median), analysis of variance and the Linear Mix Model (REML) to determine the extent of the differences between the three areas, Area1, Area2, and Area3 in a Genstat environment. The median was chosen as the central tendency for the rainfall- runoff analysis (Miller, 1994).

Development of Schema and Site Selection - A schema was developed for the entire study area using the overlay method patterned after (McHarg, 1992; Terrasa-Soler, 2012). The schema guided the selection of potential test sites for the location of various interventions by categorizing the areas into low risk, medium risk and high risk areas and was prepared using data from level of exposure to floods, soil type, extent of vegetation cover, geological cover, and extent of imperviousness, among others. Details on the preparation of the schema is beyond the scope of this research and is provided in another paper.

Selection of Test Sites and Design concept – Facilities with extensive roof area within residential areas in areas identified by the schema as low risk for each Focus area were selected and used to model the implementation of interventions. A test sites selected from Focal area1 (Living Labs Site, LLS) was modelled after the living Labs concept as explained by (Mat Eriksson, Niitamo, Kulkki, & Hribermik, 2006; R. B. Katzy, S. K. & Pawar, & Klaus-Dieter Thoben, 2012; Leminen, 2013) was used for the purposes of this paper. There were 5 separate sessions with the selected Living Labs site. The sessions were used to gather data on circulation, study activity patterns on the site as explained by Lynch (1971), conduct infiltration and soil profile tests, study the general nature of the site, alongside discussion sessions which lead to the development of outline proposals.

Site Description: The test site (LLS) belongs to a Parish and is located within a residential area. It sits on a crest (34 m elevation) and is composed of church buildings, conference and housing facilities, a cluster of basic schools for two communities and a French School for both Basic and Senior High School levels. The parish has about 2600 parishioners and 10 supporting staff. The schools have 1923 pupils and 82 teachers. The combined roof area of the buildings and structures on site is 6879.7 m² (1.7 acres) spread over 27283.9 m² (6.742 acres). The dominant soil type of the area is the Nyigbenya-Haacho complex which characteristically has hard pan formation in areas exposed to the elements (Brammer, 1967, p. 52; Obeng, 2000, p. 19). The site is almost bare of vegetation except a few scattered trees like Millettia thornningii (Miletia), Swietenia macrophylla (Mahogany) Albizia lebbeck, and Blaghia sapida (Ackee Apple), Eucalyptus sp. There are no ground covers. The bare nature of the site (Plate 2) exposes it to heavy dust generation during dry periods and sediment ladened runoff generation during rainy periods. The site relies on limited surface drains and an extensive pervious area to manage stormwater runoff. However, due to heavy compaction and an underlying hard pan in the area, a lot of sediment ladened runoff is generated from the extensive bare surfaces during rainfall which is emptied unto nearby streets. Infiltration tests and soil profile tests run for the site confirmed the presence of hard pan at 1.2m depth at areas with infiltration of between 14 mm/h to 190 mm/h. Areas with no hard pan at 1.2m depth had infiltration ≥ 300 mm per hour.

Data Processing

The figure below (Fig. 4) provides a brief on the process flow for the different types of data collected and derivations used and how they relate to each other.

Direct runoff – Direct runoff for each focus areas was estimated based on the relation:

$Qa=\frac{({P-0.2S)}^{2}}{P+0.8S}$ Eq. 1

Where Qa – direct runoff in depth over the drainage area (mm)

P - Actual rainfall (PQ) in depth over the drainage area (mm)

S - Potential maximum retention of rainwater by the soil after run-off begins (mm). S is related to the soil and cover conditions of the watershed through the Curve Number (CN), which has a range of 0-100 (Marsh, 1978, p. 59.52).

In this research S was related to CN (Chang, 2015; United States Department of Agriculture, 2004b) by the relationship:

$S=\left(\frac{1000}{{CN}_{w}}-10\right)x25.4$ Eq. 2

${CN}_{w}=\frac{{Q}_{T}}{ {A}_{T}}$ Eq. 3

Where CN_{w −} is Composite CN

Q_T - is sum of the products of the percentage of area of individual land cover types and their corresponding CN values; A_T is total area of focus site in Acres (ISWM, 2010; United States Department of Agriculture, 2004b, pp. 10–15).

Q_T is given as follows:

${Q}_{T}=\sum _{n}^{i}Ai x CNi+\dots An x CNn$ . Eq. 4

A_i - area of the ith land cover type, and CN_i is the CN value of the ith Land cover type for a given HSG (United States Department of Agriculture, 2004b, pp. 10–15).

Runoff Volume - Effective Runoff volume was estimated based on (Juliana, 2017) using the relation;

$ER=R x A x C$ Eq. 5

Where ER is Effective Runoff (m³); R is rainfall (mm); A is catchment area (m²) and C is runoff coefficient. Effective runoff is defined as the difference between rainfall and infiltration capacity (Ponce, 1996, p. 13).

Harvestable roof runoff - Harvestable roof runoff per rain event per household was calculated based on (Siabi 2015, Rahimi 2018):

$Q=R\_f x A\_r x C\_$ Eq. 6

Where Q - is quantity of water collected during rainstorm; R_f - is Rainfall in (mm);

A_r - is Surface area of roof; C -is the Runoff Coefficient.

Runoff Coefficient (C) was calculated based on (Marsh, 1978, p. 13) as:

$$\frac{Runoff\left(mm\right)}{Rainfall\left(mm\right)}$$

Relationship between roof runoff and Roof size – Harvestable roof runoff was estimated for roof sizes ≥ 81m² guided by (Juliana, 2017; Lani, 2018; Takagi, 2018). To account for the impact of the many small and illegal structures scattered throughout the study area, runoff estimates from roofs < 81m² was also included in the analysis. Percentage Harvestable Roof runoff was calculated from the ratio of daily harvestable roof runoff and daily direct runoff per rain event.

Development of a Schema - A schema was developed for the entire study area using the overlay method patterned as previously explained after (McHarg, 1992; Terrasa-Soler, 2012). The schema guided the selection of potential sites for the location of various interventions and was used to zone the focus areas into low risk, moderate risk and high-risk areas.

Development of Concept drawings - Combining a Living Labs approach with the concept-test model and analysis-synthesis approach as explained by (Mats; Eriksson et al., 2006; R. Bernhard; Katzy et al., 2012), research results were integrated into a design process for the design of the selected test site, leading to the preparation of concept drawings and a detailed design for the selected test site in Area1.

2.4 Application of Design Methods

This section shows how the data collected and the results of the analysis were integrated into the design process.

Outline proposal for Living Labs Site–To facilitate the design process, the selected test site was divided into four main zones based on activity patterns (Fig. 7).

Zone A is dominated by a bare-surface football park and the main church building. For the football park, roofwater from nearby buildings could be harvested and stored in trenches filled with grade gravel within the hard pan layer and overlaid with successive layers of sand and topsoil to allow planted grass to be sub-irrigated by capillary action. Roofwater from the main church building which is at a lower elevation could be managed using infiltration wells.

A secondary play area made of sand pit populated with various play structures for children was proposed for Zone B which may be fenced to define the space and modelled to receive runoff from the immediate surroundings.

Parking space was created in Zone C for Sunday church services. To reduce the volume of stormwater generated runoff from Zone C, roofwater may be harvested and stored in surface tanks which could be used to irrigate soft play areas and for flushing of toilets in the schools’ lavatories. In addition, infiltration wells and infiltration trenches could be introduced to infiltrate roof runoff to recharge groundwater in places where it was more appropriate.

Land cover types and Imperviousness - Ten land cover types were identified and extracted (Table 3). Analysis of variance performed on the ten cover types showed significant differences (p = 0.003) but no significant differences between the three focal areas (p = 0.395) (Fig. 7a). Percentage imperviousness increased from south, Area1 (54%) northwards Area3 (22%) (Table 2), but a linear mix model test produced no significant differences at 5% between the three focal areas (p = 0.475).

Table 3

**Summary of coverage of Land-cover types for the focus areas**
LandCT	Area1 (ac)	% Cov	Area2 (ac)	% Cov	Area3 (ac)	% Cov
Rooftops(i) P street(i) U Street(p)	884.78 162.31 141.62	34.99 6.42 5.60	572.58 30.95 30.95	22.64 1.22 9.34	371.95 17.61 147.86	14.8 0.7 5.6
Bare Soil(i)	44.52	1.76	15.24	0.60	17.68	0.7
Pavement(i)	24.44	0.97	33.27	1.32	133.73	5.3
U Build(i)	8.05	0.32	42.66	1.69	30.70	1.2
Playfield(p) Backyard(p) Trees(p) UV Land(p) Farmland(p) LOSbare (p) LOSVeg (p) LOSpave(i) Total (ac) % Imperv	7.75 0.94 104.96 1.10 - 861.21 - 287.07 2528.75	0.31 0.04 4.15 0.04 - 34.06 - 11.35 100 54.05	4.49 2.77 120.63 173.45 22.84 424.59 424.59 424.59 2528.75	0.18 0.11 4.77 6.86 0.90 16.79 16.79 16.79 100 43.66	4.49 - 132.12 15.69 505.85 394.56 794.95 - 2519.03	0.2 - 5.2 0.62 15.8 31.6 31.6 - 100 22.0
Notes: Land CT- Land cover type; % Cov - % coverage; P Street-paved street, U street-unpaved street; U Build-Uncompleted buildings; UV Land-Undeveloped vegetated land; LOSbare, LOSveg, LOSpave – left over space bare, vegetated and paved, respectively; % Imperv-%imperviousness; (i) and (p) – impervious and pervious areas, respectively; ac – acre.

Roof Size analysis - Percentage average roof area (≥ 81m²) increased from Area1–77.1percent > Area2–84.4 percent > Area3–88.6 percent, with decreasing imperviousness from South towards North (Table 3 & Table 4). There were far more small sized roofs of < 81m² in the south (Area1) which made up 23 percent of the roofs compared with 15.6 percent and 11.4 percent for Area2 and Area3 respectively (Fig. 7). A linear mix model test run for the three focal areas on roof size distribution (Table 4) showed significant differences at 5% (p = 0.005) between focal areas.

Table 4

Roof size distribution in the focal areas
Parameter	Area1	Area2	Area3	P-Value
MRA < 81m² MRA ≥ 81m² No.RA < 81m² No.RA ≥ 81m²	27.6 182.2 27,727 14,129	27.2 195.6 13,176 9,954	28.5 248.5 5,994 5,355	0.005
CRoofA < 81m²	765,256.2	358,387.2	170,829.0
CRoofA ≥ 81m²	2,574,304.0	1,947,002.0	1,330,718.0
%Roofs < 81m² %Roofs ≥ 81m²	22.9 77.1	15.5 84.5	11.4 88.6
Notes: MRA – mean roof area, No.RA < 81m² – number of roofs < 81m²; No.RA ≥ 81m² - number of roofs ≥ 81m²: CRoofA – Combined roof area

Rainfall and Runoff analysis - The mean (median) daily rainfall (> 40mm) among the three focal areas varied from 68.6mm, 62.8mm, to 69.2mm for Area1, Area2 and Area3 respectively (Table 5). A linear mixed model test showed no significant differences between the focal areas at 5% (P = 0.669). The highest daily rainfall of 313.8mm was recorded in Area1 in June, 2009; the highest for Area2 was 139.6mm in May, 2009, whilst that of Area3 was 148.3mm in June, 2015. Mean direct daily runoff (MDDRunoff) varied from 35.8mm for Area1; 21.1mm for Area2; and 22.1mm for Area3 and did not show any significant differences at 5% between the focal areas (p = 0.196), months and years (Table 5, Fig. 5 & Fig. 6, and Fig. 7).

Table 5

**Runoff levels and Runoff volumes from the three focus areas**
	Area1	Area2	Area3	P – Values
MDRainfall (mm)	68.6	62.8	69.2	0.669
MDD Runoff (mm) CDHRW(m³)	35.8 170,279	21.1 60,802	22.1 42,420	0.196 < 0.001
EMDRunoff (m³)	522,033	265,690	287,992	< 0.001
% DHRWater	32.6%	22.5%	14.7%	< 0.001
Runoff Coefficient	0.7463	0.5526	0.3739	< 0.001
Annual Rainfall (mm)	848.0	846.0	1057,5
Notes: MDRainfall; Mean Daily rainfall > 40mm, MDD Runoff – Median Daily Direct runoff (mm); CDHRWater – Combine Daily Harvestable roofwater per rain event, EMDRunoff – Effective Mean Daily Runoff (m³) per rain event, percent DHR water – percent Daily Harvestable roofwater which is the ratio of total harvestable roof water and effective daily runoff per rain event between 2006–2015. Source: PhD Research, Asiedu [2020]

For all three focal areas 28.6 percent of the daily mean rainfall produced direct runoff > 12.7 mm with Area2 producing the highest (32.5%) and Area3 the lowest (24.7%). Effective runoff which is the difference between rainfall and infiltration capacity (Ponce, 1996) ranged between 522,033 m³ and 265,690 m³(Table 5).

Runoff Contribution from roofs –The contribution of small roofs (< 81m²) to daily direct runoff in the focal areas was lowest for Area3 (11.4%), followed by Area2 (15.6%) and Area1 (22.9%). The combined contribution to daily runoff from small (< 81m²) and large roofs (≥ 81m²) varied among the focal areas as 32.6 percent for Area1, 22.5 percent for Area2 and 14.7 percent for Area3 (Table 3).

A similar analysis using only the dominant HSG in a focal area yielded far higher values from a combined contribution from roofs to daily direct runoff of 82.9 percent for Area1, 54.2 percent for Area2 and 34.26 percent for Area3.

Strategy for Stormwater management at plot level: Based on the finding that the roofs of buildings contributed between 14.7 percent – 32.6 percent to stormwater runoff, a strategy was developed targeting roofwater runoff at the plot level. The strategy was framed to ensure that stormwater generated runoff from roofs at the plot level is retained within plots through roofwater harvesting for non-portable use, irrigation and/or infiltration to recharge ground water. The implementation of this strategy is in the development the outline proposal which was translated into the conceptual plan.

Conceptual drawing or Design proposal for Living Labs Site: The final result from the outline proposals was used to develop the conceptual drawing for the site (Fig. 8).

Percentage imperviousness in Area3 (22%) puts the area under low density development while that of Area1 and Area2 with > 40 percent places them under high density according to Su (2014). This suggests that more than half of the study area with ≥ 40 percent imperviousness could be considered as high density (Su, 2014). The percentage of roofs is close to what was reported by Gogate (2012, p. 37) who gave a range of "25–35 percent of the total impervious area on a regional basis as roofs”. The figure of 70 percent is reported by (Garrison, 2011) as the contribution to imperviousness from roofs, parking areas and roads. The percentage of imperviousness seems to be low compared with (Garrison, 2011) and this could be because of the generally low level of development which according to (Parkinson, 2015) is characteristic of less developed countries. The observed level of imperviousness (22%-54.1%) is within range of the 16.1 percent and 66.1 percent for an urbanized watershed (Grove, 1998).

The observed trend in daily rainfall (> 40mm) and corresponding daily direct runoff (> 12.7mm) seem to suggest that rainfall levels alone might not have influenced direct runoff generation in magnitude (Fig. 7) in the research area with a relatively flat terrain (Darko, Barnes, & Sekpey, 1995; Nyarko, 2000; Programme, 1991) where 28.6 percent of all rainfall yielded runoff > 12.7mm. It also shows a relatively even distribution of runoff over the months with no particular month producing statistically significantly high concentrations (Fig. 6), a finding which does not conform with the generally held view that there are two major rainy seasons in a year in the study area with widely different levels of runoff generation (WRC-Ghana, 2007). The direct runoff generation trend especially between 2008 and 2013 suggest a downward trend, but this contrasts with the continuously high flooding situation in the study area over the same period (C. Y. Y. Okyere, Y.; Gilgenbach, D., 2013; C. a. A. Tengan, Clinton Ohis, 2016). The reason for the difference could be due to possible under-estimation by the runoff model (Bell, 1999).

Annual runoff was estimated at 4.1 x 10^− 3 km³/yr., 1.539 x 10^− 3 km³/yr. and 1.715 x 10^− 3 km³/yr. for Area1, Area2 and Area3, respectively which according to (Stuebe, 1990) should be 18 percent lower than manually determined runoff values. Area1 produced more annual runoff than the other focus areas, especially in 2015.

Roof size and Runoff correlation - There were very strong correlation (R² = 0.9999) between roof size and runoff generation, the differences showing significance at 5% (p = 0.00551) which suggests that the size of roofs could have influenced significantly runoff generation. This however was not reflected in the amount of daily harvestable roof runoff for both small (< 81m²) and large roofs (≥ 81m²) (Table 4 & Table 5) which was highest in Area1 (251,876 m³) and lowest in Area3 (112,567 m³). A similar trend is observed with percentage harvestable roof runoff which was highest in Area1 (32.6%) and lowest in Area3 (14.7%) (Table 6). It is possible that the number of large roofs (≥ 81m²) in Area3 might not have been enough to impact significantly on daily direct runoff, compared with the relatively small roofs of Area1 which was far higher in number.

Gogate (Gogate, 2012, p. 37) has observed that water from roofs and roads represents by far the most important contribution to urban surface runoff and the results confirm this trend where roof runoff alone contributed between 32.6 percent and 14.7 percent to surface runoff (Table 5). Pickett (Pickett, 2001) had earlier reported that 13 percent of rainfall in urban areas falls on buildings but a figure of 71.53 is reported by (Juliana, 2017). The observed average of 32.6 percent, 22.5 percent and 14.7 percent respectively from Area1, Area2 and Area3 shows that roofwater contribution to runoff in communities was significant although lower than what is reported by (Juliana, 2017). This may be explained by the general level of densification in the study area, as less dense areas produce less runoff compared with higher density areas (Asumadu-Sarkodie, 2015; Itsukushima, 2018).

There were far more small roofs (< 81m²) in Area1 which also contributed 22.9 percent to direct runoff compared with Area2 (15.6%) and Area3 (11.4%). The contribution of small roofs is a significant factor in runoff generation in the study area especially in the light of the numerous squatter camps (ILGS/IWMI, 2012; Jha, 2012) composed mainly of a collection of kiosks and make shift structures.

Developing a Strategy - based on the results that between 32.6 percent and 14.7 percent of stormwater runoff was contributed by the roofs of buildings, a strategy was developed for the study area to target roof runoff management at the plot level. The strategy was framed to ensure that 70–95 percent of stormwater generated from roofs of buildings at the plot level was retained within plots through roofwater harvesting for reuse, irrigation and infiltration to recharge ground water. This could provide the basis for the development of a policy on stormwater management at the plot level to help fill the gap identified by (F. Frick-Trzebitzky, 2018) and(Andoh, 2016). The resulting reduction in total stormwater runoff will keep runoff volumes low (effective runoff volume), making it easier for the rest of the stormwater to "infiltrate, evaporate" (Echols, 2007, p. 209) or easily channeled through the existing conveyance systems.

Living Labs Site - Roofwater from the main church building was infiltrated because the building was at a lower elevation and it will have required investing in machines to pump harvested roofwater up to the football park which was on a higher elevation. Harvesting and storing roofwater in trenches to irrigate the football park by subirrigation will ensure continuity between the underlying hard pan layer, imported topsoil and the grassed surface. This will prevent washing away of the imported material and the grass as has often occurred with previous interventions (Personal Communication, 2018. Parishner, Catholic Parish, Holy Family Church, Mataheko, Accra-Ghana). The second play space composed of a sand pit with various play equipment will provide extra play space in addition to the very limited existing play facilities provided for the 1923 pupils. It will also cover the bare space to reduce runoff generation. Provision of surface storage tanks to harvest roofwater and store for irrigation and flashing of toilets and infiltration will reduce the amount of stormwater runoff generated from the site. This will also reduce dependence on municipal water supply which unfortunately is not always reliable.

The concept of rainwater harvesting is not new and has received a lot of attention in recent times as a means to ease pressure on the existing public water supply system, decrease soil erosion and control flooding during rainy periods (Amisigo, undated; Asare, 2013; Jha, 2012; Shepherd, 2016, p. 6). Rainwater harvesting is a good means for communities to adapt to climate change (Boelee, 2013) and is an old method used almost throughout the world(Guebaili, 2011). Rainwater harvesting and infiltration are Low Impact techniques for stormwater management which could be developed both at the local and regional levels to manage urban floods (ActionAid, 2006, p. 6; Palanisamy, 2015). Rainwater harvesting is one of the improved sources of water which the urban population can easily access (Osumanu, 2010, p. 1) and is a critical intervention in "water conservation" efforts in areas with increasing climate variability (Armitage, 2013, p. 25). As an "age old practice", rainwater water harvesting is well known across Africa (Armitage, 2013; Boelee, 2013; Fuest, 2007; Guebaili, 2011). Its successful integration and development for domestic use to supplement pipe or Municipal water will not only impact the economy of Sub-Saharan Africa, where livelihood depends on reliable water supply, but most importantly, help address perennial floods (Boelee, 2013, pp. 509–510).

The Living Labs concept - The use of the Living Labs concept was a challenge as the actors were less forthcoming with ideas because perhaps they were not sufficiently motivated to challenge themselves to new ways of thinking and addressing issues in a way that was different from the prevailing believe that surface conveyance systems were the best way to address stormwater runoff to avoid floods. According to(Seybert, 2006) where people are not familiar with an innovative approach to address a problem, even with all its benefits, acceptance become difficult. It is also possible that actors did not feel morally obliged to intervene to address the challenge of stormwater management at the plot level as they could not appreciate the connection between plot level runoff generation and community runoff generation. (Beery, 2018) had earlier explained that actors feel more obliged to actively participate in a Living Labs experiment where they can see the “severe negative” impact of their inaction. Another reason could be from the perception the actors had that the exercise was futile and will most likely never be implemented. This made the work of the researcher, who acted more as catalyst very difficult and required that he often motivate by convincing the actors of the long term implications of the exercise(Capdevila, 2014). This also show that “curiosity” and a desire to contribute to find a solution to a problem alone could not be enough motivation for actors in a Living Labs experiment as suggested by (Schuurman, Ballon, Baccarne, De Marez, & Veeckman, 2016). The location and design of infiltration wells and infiltration trenches were guided by the infiltrating characteristics of the focus design sites to ensure that roofwater is infiltrated within a maximum of 72hours of a rain event to avoid problems with mosquitoes using a 25 year design storm(Blansett, 2016; PWUD, 2000, 2006). Other design standards were applied such as depth of infiltration wells(PWUD, 2006; RWRA, 2011; Strom, 2009), proximity to foundations and location away from various facilities(RWRA, 2011; Young, 2009), depth to water table (Young, 2009).

The research showed a clear trend in imperviousness which although not significant at 5 percent between focal areas increased from South to North; from dense urbanized Area1 (54%) decreasing towards the outer sparsely populated urbanized Area3 (22%). The research modeled runoff generation for a 7 year rainfall period based on the CN approach by using a composite CN to represent the different land cover types and soil hydrological groups of the focus areas. The research established a direct relationship between harvestable daily roof runoff and daily direct runoff using daily rainfall > 40mm from 2006 and 2015, where harvestable roof runoff contributed 32.6 percent (Area1), 22.5% (Area2) and 14.7 percent (Area3) to total direct runoff. This confirms the significant contribution of roof runoff to stormwater generation within communities. A similar analysis using only the dominant hydrological soil groups in the three focus areas gave a far higher percentage runoff of 82.9 percent, 54.2 percent and 34.3 percent for Area1, Area2 and Area3 respectively. The results suggest that less dense areas with large roof sizes did not impact significantly on direct runoff because there were fewer number of roofs. Based on the foregoing discussions, the authors proposes that a combination of roofwater harvesting, storage for reuse and infiltration at the plot level using infiltration wells or trenches could significantly reduce the amount of runoff generated within the research area.

This research affirms this position and proposes a "source disposal and treatment" control approach for urban areas in the study area which is holistic; an integrated approach which aims at reducing runoff generation by retaining runoff at the generation point(Urbonas, 2000, pp. 7 − 3) through roofwater harvesting and infiltration.

The major contribution of the research is to help bridge the gap and the missing link between what is proposed in policy documents on storm water management to control floods and what is actually implemented. The research closes this gap by using an evidence-based approach to support policy for stormwater management at the plot level in the study area through the significant contribution of roofs to runoff generation within the study area. The research proposed how through infiltration and rainwater harvesting at the plot level a strategy targeting roofwater could be implemented.

Future Direction

Scanning the literature, it is clear that little work has been done in quantifying runoff generation from the roof of buildings not only in the study area but also in the entire sub-Saharan Africa region and this goes to re-affirm observations made by several writers on the challenge of doing empirical research in developing countries (Eckart, 2012, p. 13; Ouikotan et al., 2017, p. 8; Tiepolo, 2014, p. 24). This has made it difficult to collaborate the findings of this research with similar works in the sub-region. The Author suggests further research to adapt CN values to increase efficiency of the model. Future research could also look at other methods for runoff estimation from ungagged watersheds being used in other parts of the world such as remote sensed approaches which combines vegetation, imperviousness and soil data in estimating runoff (Fan, Deng, Hu, & Weng, 2013). Another area of research should be to determine the critical minimum storage size and infiltration levels to significantly impact stormwater generation to control floods. The author is currently working to develop local standards to store roof water in the soil for sub-irrigation of lawns at the plot level.

Ethical Approval:

Ethical approval for this research was obtained from the University of Cape Coast Ethical Clearance Secretariat ID (UCCIRB/CHLS/2020/45). Ethical considerations included seeking the consent of participants during interviews and providing a declaration to the effect that participants will not be directly quoted in a document derived from the interview without the express permission of the correspondent. Also no part of responses provided will be used against them in any court of law.

Competing interest

We declare that this paper's authors have no competing or other interests that are perceived to influence the results or discussion presented in this paper.

Funding:

This is not applicable as the research was supported through personal funds of the authors.

Availability of Data material:

A third party provided some of the data used during the study. Direct requests for these materials may be made to the provider. Some of the data used to support the findings of this study are available from the corresponding author upon reasonable request.

Author Contribution

JBK - conceptualized the research, developed collected data processed the data analysed and conclusionBN - contributed to literature review, discussions and conclusions and the overall structure of the research

ActionAid. (2006). Climate change, Urban Flooding and the rights of Urban Poor in Africa. London, UK: Action Aid International.
Adger, W. N. (2006). Vulnerability. Global Environmental Change, 16, 268-281.
Alexander, C. (1964). Notes on the Synthesis of Form (7 ed.). Cambridge, MA. USA: Harvard University Press.
Allen, D. M. A., R. John; Geiger, F. Wolfgang; Johnston, D. Linton; Pezzaniti, David and Scott, Peter. (2013). Water Senstitive Urban Design: Basic Procedures for 'Source Control' of Stormwater (J. R. Argue Ed. Student Edition ed.). University of South Australia: Urban Water Resources Center, University of SOuth Australia.
Amisigo, B. a. K.-Y. (undated). Rainwater Harvesting for Resilience to the Impact Climate Change on Water Resources in Ghana. Center for Scientific and Industrial Research (CSIR) - Water Research Institute - Accra Ghana. Accra Ghana.
Andoh, R. Y. G. (2016). Addressing Accra's urban water challenges within a sustainable development context. Paper presented at the 39th Water Engineering and Development center (WEDC) International Conference, Kumasi, Ghana, Kumasi, Ghana.
Armitage, N. V., Michael; Winter, Kevin; Spiegel, Andrew and Dunstan, Jessica. (2013). Alternative Technology for Stormwater Management Alternative Technology for Stormwater Management. Cape Town, Republic of South Africa: Water Research Commission, South Africa.
ARUP. (2016). Future Proofing Cities - Ghana Metropolitan Cities (F. C. Africa, Trans.) Cities Alliance. London, United Kingdom: Cities Alliance- Future Cities Africa.
Asare, R. a. K., Eric. (2013). Report on Rain Water Harvestings (RWH) for Resilience to Climate Change Impact on Water Availability in Ghana Worskshop. Accra-Ghana: Council for Scientific and Industrial Research (CSIR)- Science and Technology Policy Research Institute (STEPRI).
Asiedu, J. B. (2020). Infiltrating to Control Floods - Suitability of Infiltration based Systems in Sub-Saharan Africa. American Journal of Earth and Environmental Sciences, 3(1), 1-12.
Asumadu-Sarkodie, S. O., A Phebe and Jayaweera, M. P. C. Herath. (2015). Flood risk management in Ghana: A case study in Accra Advances in Applied Science Research, 6(4), 196-201.
Barbosa, A. E. F., J. N.; and David, I. M. (2012). Key issues for sustainable urban stormwater management. Water Research, 46(20), 6787-6798.
Beery, T. (2018). Engaging the Private Homeowner: Linking Climate Change and Green Stormwater Infrastructure Sustainability, 10(12), 1-16.
Bell, S. (1999). Landscape: Pattern, Perception and Process. London, United Kingdom: E & FN Spon
Bhattacharya-Mis, N. a. L., J. (2011). A Review of the Flood risk situation in African growing economies. Paper presented at the Urban Flood Risk Management (UFRIM): Urban Flood Risk Management approaches to enhance Resilience of Communities, Graz, Austria.
Blansett, K. (2016). Design and Construction of Dry Wells in Pennsylvania. Pennsylvania State College of Engineering, Pennsylvania USA: Pennsylvania Housing Research Center.
Boelee, E. Y., Mekonnen; Poda, Jean-Noel; McCartney, Matthew; Cecchi, Philippe; Kibret, Solomon; Hagos, Fitsum and Laamrani, Hammou. (2013). Options for water storage and rainwater harvesting to improve health and resilience against climate change in Africa. Regional Environmental Chanage, 13, 509-519.
Brammer, H. (1967). Soils of the Accra Plains (Vol. Memoir No 3). Accra Ghana: Soil Research Institute- Ghana Academy of Arts and Sciences.
Capdevila, I. (2014). How can Living Labs enhance the participants motivation in different types of innovation activities. SSRN Eelctronic Journal. doi: 10.2139/ssrn.2502795
Chabaeva, A. A. C., L. Daniel and Prisloe, Sandy. (2004). Development of Population Density and Landuse based Regression Model to calculate the amount of Imperviousness. Paper presented at the ASPRS Annual Conference Proceedings, Denver, Colorado.
Chang, L.-F. H., Shu-Li. (2015). Assessing urban flooding vulnerability with an emergy approach. Landscape and Urban Planning, 143, 11-24. doi: 10.1016/j.landurbplan.2015.06.004
CLUVA. (2012). Assessing Vulnerability of Urban Systems, Populations and Goods in Relation to Natural and Man-made Disasters in Africa Climate Change and Urban Vulnerability in Africa (CLUVA): University of Manchester.
CRED/UNISDR. (2016). The Human Cost of Weather Related Disasters 1995-2015. Brussels, Belgium: Centre for Research on Epidemiology of Disasters (CRED) and United Nations Office for Disaster Risk Reduction (UNISDR).
Critchley, W. a. S., Klaus. (1991). Water Harvesting: A Manual for the Design and Construction of Water Harvesting Schemes for Plant Production. Rome: FAO.
Cross, N. (1982). Desingerly ways of Knowing. Design Studies, 3(4), 221-227.
Darko, P., Barnes, E., & Sekpey, K. N. (1995). Ground Water Assessment of the Accra Plains. Unpublished: Water Resources Research Institute, Center for Scientific and Industrial Research-Ghana
Dile, T. Y. K., Louise; Srinivasan, Raghavan and Rockström, Johan. (2016). Investigation of the Curve Number method for Surface Runoff Estimation in Tropical Regions. Journa of the American Water Resources Association, 52(5).
Dile, Y. T. K., Louise; Srinivasan, Raghavan, and Rockström, Johan. (2016). Investigation of the Curve Number Method for Surface Runoff Estimation in Tropical Regions. Journal of American Water Resources Association, 52(5), 1155-1169.
Dingman, S. L. (2002). Physical Hydrology (2 ed.). New Jersey, USA: Prentice Hall.
Echols, S. (2007). Artful Rainwater Design in the Urban Landscape. Journal of Green Building, 2(4), 101-122.
Eckart, J. G., Kebreab; Khatri, Krishna; Mutikamga, Harrison; Sempewo, Jotham; Tsegaye, Seneshaw and Vairavamoorthy, Kalanithy. (2012). The Future of Water in African Cities - Why Waste Water? : The World bank.
Eriksson, M., Niitamo, V.-P., Kulkki, S., & Hribermik, A. K. (2006). Living Labs as a Multi-Contextual R&D Methology. Paper presented at the 2006 IEEE International Technology Management Conference (ICE), Milan, Italy.
Eriksson, M., Niitmano, V.-P., Kulkki, S., & Hribernik, A. K. (2006). Living Labs as Multi-contextual R&D Methodology. Paper presented at the 2006 IEEE International Technology Management Conference, Milan, Italy.
Fan, F., Deng, Y., Hu, X., & Weng, Q. (2013). Estimating Composite Curve number using an Improved SCS-CN method with Remotely Sensed variables in Guangzhou. Remote Sensing, 5, 1425-1438.
Frayling, C. (1993). Research in Art and Design. Royal College of Art Research Papers, 1(1).
Frick-Trzebitzky, F. (2018). Riskscapes of Flooding - Social dynamics and Adaptation in a rapidly urbanizing wetland: The Densu delta case in Accra, Ghana. Humboldt-University, Berlin. Germany, Humboldt-Universität zu Berlin, Berlin.
Frick-Trzebitzky, F. a. B., Antje. (2017). Disparities in the Implementation gap: adaptation to flood risk in the Densu Delta, Accra, Ghana. Journal of Environmental Policy and Planning. doi: 10.1080/1523908X.2017.1343136
Fuest, V. a. H., A. Stefan. (2007). PPP - Policies, Practices and problems in Ghana's urban Water Supply. Water Policy, 9, 169-192.
Garrison, N. a. H., Karen. (2011). Rooftops to Rivers II: Green Strategies for Controlling Stormwater abd Combined Sewer Overflows. Retrieved from Rooftop to Rivers II website: http://www.nrdc.org/water/pollution/rooftopsii/files/rooftopstoriversII.pdf
Gogate, G. N. a. R., M. Pratap. (2012). Sustainable Stormwater Management in Developing and Developed Countries: A Review. Paper presented at the International Conference on Advances in Design and Construction of Structures (ADCS 2012), Bangalore, India.
Grove, M. H., Jon and Engel, Bernard. (1998). Composite VS. Distributed Curve Numbers: Effects on Estimates of Storm Runoff depths. Journal of The American Water Resources Association, 34(5), 1015-1023.
GSS, G. S. S. (2014). Population and Housing Census 2010 Vol. District Analytical Report. Retrieved from http://www.statsghana.gov.gh/docfiles/2010_District_Report/Greater%20Accra/GA%20WEST.pdf
Guebaili, A. D., Y.; Guedri, A. and Boukhari, S. (2011). Rainwater harvesting in North Africa: A Novel Method for Reservoir Sizing. Journal of Material Environmental Science, S1, 469-472.
Güneralp, B. G., Inci and Liu, Ying. (2015). Changing global pattern of Urban exposure to flood and drought hazards. Global Environmental Change, 31, 217-225.
Gupta, K. A., Singh, S., Wajih, A. S., Mani, N., & Singh, A. K. (2017). Urban Resilience and Sustainability - Through Peri-Urban Ecosystems (S. Katyal Ed.).
ILGS/IWMI. (2012). Community Adaptation to Flooding Risk and Vulnerability. Accra-Ghana: Institute of Local Government Stidues (ILGS) and International Water Management Institute (IWMI)
ISWM. (2006). Storm Water Management System Planning and Design Desdign Manual for Site Development. Arlington, Texas: Integrated Storm Water Management (iSWM).
ISWM. (2010). Hydrology. In E. a. D. North Central Council of Governments (Ed.), iSWM Technical Manual. Arlingron, Texas: Integrated Storm Water Management (iSWM).
Itsukushima, R. O., Yohei; Iwanaga, Yuki and Tatsuro, Sato. (2018). Investigating the influence of various Stormwater Runoff Control facilities on Runoff control Effciency in Small Catchment areas. Sustainability, 10(2). doi: doi:10.3390/su10020407
Jha, A. K. B., Robin and Lamond, Jessica (2012). Cities and Flooding: A guide to integrated urban flood risk management for the 21st century Retrieved from https://www.gfdrr.org/sites/gfdrr/files/urban-floods/urbanfloods.html doi:10.1596/978-0-8213-8866-2
Jones, J. C. (1992). Design Methods (2 ed.). New York: John Wiley & Sons, Inc.
Juliana, C. I. K., M. Syhril Badri; Cahyono, M; Kardhana, Hadi and Widjaja, M. (2017). Rainwater Harvesting System implementation for Domestic water use: The Environment and Financial benefits. Paper presented at the ICoSI 2014: Proceedings of the 2nd International Conferecne on Sustainable Innovation, Singapore.
Justice, D. a. R., Fay. (2002). Developing impoervious surface estimates for Coastal New Hampshire The New Hampshire Estruaries Project. New Hampshire, USA.
Katzy, R. B., Pawar, S. K., & Thoben, K.-D. (2012). Editorial: Living Lab Research Agenda. International Journal of Product Development, 17(1/2).
Katzy, R. B., Pawar, S. K., & Thoben, K.-D. (2012). Editorial: Living Labs Research Agenda. International Journal of Product Development, 17(1).
Landscape-Institute. (2009). ; Green Infrastructure Connected and Multifunctional landscapes. Position Statement. http://www.landscapeinstitute.org/PDF/Contribute/GreenInfrastructurepositionstatement13May09.pdf
Lani, N. H. M. Y., Zulkifli and Syafiuddin, Achmad. (2018). A Review of Rainwater Harvesting in Malaysia: Prospects and Challenges. Water, 10(4). doi: doi:10.3390/w10040506
Leminen, S. (2013). Coordination and Participation in Living Labs Network. Technology Innovation Management Review, 5-14.
Lenzholzer, S. D., I and Jusuck Koh. (2013). Research through Designing in Landscape Architecture. Landscape and Urban Planning, 113(2013), 120-127.
Lenzholzer, S. D., I and van den Brink, Adri. (2017). Research in Landscape Architecture and the special role of Designing. In D. B. Adri van den Brik, Hilder Tpbi and Siomn Bell (Ed.), Research in Landscape Architecture-Methods and Methodology. New York: Routledge Taylor & Francis group.
Linden, v. d. J. d. L., ANdre Pedroso; de Aguiar, Joao Pedro Ornaghi. (2001). The Evolution of Design. Paper presented at the 9th Internatinal Conference of the European Academy of Design, Porto-Portugal. https://www.researchgate.net/publication/273704768_The_evolution_of_design_methods
Lynch, K. (1971). Site Planning. London, England: The MIT Press.
Marsh, M. W. (1978). Environmental Analysis for Land use and Site Planning. New York: McGraw-Hill Book Company.
McHarg, I. L. (1992). Design with Nature. Singerpore: John Wiley and Sons, Inc.
Miller, S. (1994). Handbook of Agrohydrology. Chatham, UK: Natural Resources Institute, UK.
NRC. (2009). Urban Stormwater Management in the United States. Washington DC: National Research Council (NRC) of the National Academies.
Nyarko, B. K. (2000). Flood risk zoning of Ghana: Accra experience. International Archives of Photogrammetry and Remote Sensing, XXXIII(Part B7), 1039-1050.
Obeng, H. (2000). Soil Classifications of Ghana Selected Economic Issues. Accra Ghana: Center for Economic Policy Analysis (CEPA).
Odefey, J., Detwier, S., Rousseau, K., Trice, A., Blackwell, R., O'Hara, K., . . . Raviprakash, P. (2012). Banking on Green: A Look at how Green Infrastructure can save Municipalities Money and Provide Exonomic Benefits Community-Wide: A Joint Report by American Rivers, The Water Environment Federation, the American Society of Landscape Architects and EcoNorthwest. Retrieved 11/18/13, from www.americanrivers.org/assets/pdfs/reports-and-publications/banking-on-green-report.pdf
Ogden, L. F. a. W., S. Glenn. (2017). Watershed Runoff, Streamflow Generation, and Hydrologic Flow Regimes. In V. P. Singh (Ed.), Handbook of Applied Hydrology. New York: McGraw Hill Education.
Okyere, C. Y., Yacouba, Y., & Gilgenbach, D. (2013). The prohlem of annual occurences of floods in Accra: An integration of Hydrological, Economic and Political perspectives Theoritical and Emprircal Researches in Urban Management, 8(2).
Okyere, C. Y. Y., Y.; Gilgenbach, D. (2013). The prohlem of annual occurences of floods in Accra: An integration of Hydrological, Economic and Political perspectives Theoritical and Emperical Researches in Urban Management, 8(2).
Oliveira, P. S. T. N., M. A.; Hawkings, R. H.; Stone, J. J.; Rodrigues, D. B. B.; Panachuki, E. and Wendland, E. (2016). Curve Number Estimation from Brazilian Cerrado Rainfall and Runoff data. Journal of Sol and Water Conservation 17(5), 420-429.
Oppong-Anane, K. (2006). Country Pasture Forage Resource Profile - Ghana. Rome: Food and Agriculture Organization.
Osman, A., Nyarko, K. B., & Mariwah, S. (2016). Vulnerability and Risk Levels of Communities within Ankobra Estuary of Ghana. International Journal of Disaster Risk Reduction, 19(2016), 133-144.
Osumanu, I. K. A.-R., Lukman; Sangsore, Jacob; Braimah, R. Farouk and Mulenga, Martin. (2010). Urban Water and Sanitation in Ghana: How Local action is Making a Difference Human Settlements Working Paper Water and Sanitation. London, UK: International Institute for Environment and Development IIED.
Ouikotan, R. B., van der Kwast, J., Mynett, A. a., & Afouda, A. (2017, 29 May - 3 June, 2017). Gaps and Challenges of Flood risk Management in West African coastal Cities. Paper presented at the XVI World Water Congress. International Water Resources Association, Cancun, Quintana Roo, Mexico.
Palanisamy, B. a. C., Ting Fong May. (2015). Rehabilitation of concrete canals in urban catchments using low impact development techniques. Journal of Hydrology, 523, 309-319. doi: 10.1016/j.jhydrol.2015.01.034
Parkinson, J. a. M. O. (2015). Urban Storm Water Management in Developing Countries. London, UK: IWA Publishing.
Pickett, S. T. A. C., M.L; Grove, J.M.; Nilon, C. H.; Pouyat, R.V.; Zipperer, W.C. and Costanza, R. (2001). Urban Ecological Systems: Linking Terrestrial Ecological, Physical and Socioeconomic Components of Metropolitan Areas. Annual Review of Ecologcal Systems 32, 127-157.
Ponce, M. V. a. H., H. Richard. (1996). Runoff Curve Number: Has it Reached Maturity? Journal of Hydrologic Engineering, 1(1), 11-19.
Programme, A. a. D. (1991). Strategic plan for Greater Accra Metropolitan Area: Ministry of Local Government Department of Twon and Country Planning.
PWUD, L. (2000). Drainage Criteria Manual Lincoln USA: City of Lincoln Public Works and Utilities Department (PWUD).
PWUD, L. (2006). Alternative Stormwater Best Management Practices Guidelines. City of Lincoln, USA: Public Works and Utilities Department Watershed Management Division, City of Lincoln.
Rowe, G. P. (1982). A Priori Knwoledge and Heuristic Reasoning in Architectural Design. Journal of Architectural Engineering, 36(1), 18-23.
RWRA. (2011). Infiltration Basin (R. W. R. A. (RWRA), Trans.) Minimum Standard 14.01. Owensboro-Daviess County: Regional Water Resource Agency, Owensboro-Daviess County KY.
Schuurman, D., Ballon, P., Baccarne, B., De Marez, L., & Veeckman, C. (2016). Living Labs as Open innovation Systems for Knowledge Exchange: Solutions for sustainable innovation development. International Journal of Business Innovation and Research, 10(2-3), 322-340.
Seybert, A. T. (2006). Stormwater Management for Land Development: Methods and Calculations for Quantity control. New Jersey: John WIley & SOns Inc.
Shepherd, B. (2016). Ghana Sustainable Change Rainwater Harvesting. Ohio State University: Ghana Sustainable Change.
Strom, S. N., Kurt and Woland, Jake. (2009). Site Engineering for Landscape Architects (5 ed.). New Jersey: JohnWiley & Sons.
Stuebe, M. M. a. H., Douglas M. (1990). Runoff Volume Estimation using GIS Techniques. Water Resources Bulletin, American Water Resources Association, 26(4).
Su, W. Y., Gaobin; Yao, Shimou and Yang, Guishan. (2014). Urban Land pattern impacts on Floods in a New District of China. Sustainability, 6, 6488-6508. doi: 10.3390/su6106488
Takagi, K. O., Masahiro and Otaki, Yurina. (2018). Potential of Rainwater utilization in Households based on the distribution of catchment area and End-use water demand. Water, 10(12).
Tengan, C. a., & Aigbavboa, C. O. (2016). Addressing Flooding challenges in Ghana: A Case of the Accra Metropolis. Paper presented at the International Conference on Infrastructure Development in Africa (ICIDA-2016), South Africa. https://www.researchgate.net/publication/316093848_ADDRESSING_FLOOD_CHALLENGES_IN_GHANA_A_CASE_OF_THE_ACCRA_METROPOLIS
Tengan, C. a. A., Clinton Ohis. (2016). Addressing Flooding challenges in Ghana: A Case of the Accra Metropolis. Paper presented at the International Conference on Infrastructure Development in Africa (ICIDA-2016), South Africa. https://www.researchgate.net/publication/316093848_ADDRESSING_FLOOD_CHALLENGES_IN_GHANA_A_CASE_OF_THE_ACCRA_METROPOLIS
Terrasa-Soler, J. J. B., Mery and Laura Lugo-Caro, . (2012). A Green infrastructure Plan for La Parguera, Puerto Rico. Paper presented at the Designing Nature as Infrastructure, Technische Universität München, München-Germany.
Tiepolo, M. (2014). Flood risk reduction and Climate change in Large Cities South of the Sahara. In S. M. a. M. Tiepolo (Ed.), Climate Change Vulnerability in South African Cities. Switzerland: Springer International Publishing.
United States Department of Agriculture, N. R. C. S. (2004b). Estimation of Direct Runoff from Storm Rainfall National Engineering Handbook (Vol. Part 360).
United States Department of Agriculture, N. R. C. S. (2009). Hydrologic Soil Groups National Engineering Handbook - Part 630 Hydrology (Vol. Part 630): United States Department of Agriculture.
Urbonas, B. (2000). Assessment of Stormwater Best management Practice Effectiveness. In J. P. H. a. R. P. R. Field (Ed.), Innovative Urban Wet-Weather flow management Systems. Lancaster, Michigan Technomic Publishing Company.
USDA-NRCS. (2004a). Hydrologic Soil-Cover Complexes National Engineering Handbook (Vol. Part 630). United States: United States Department of Agriculture, Natural Resource Conservation Service.
USDA. (1986). Urban Hydrology for Small Watersheds TR-55. USA.
Verbeeck, K. O., Jos Van and Hermy, Martin. (2011). Measuring extent, location and change of imperviousness in urban domestic gardens in collective housing projects. Landscape and Urban Planning, 100(1-2), 57-66. doi: 10.1016/j.landurbplan.2010.09.007
Wallemacq, P. (2018). Natural Disasters in 2017: Lower mortality, Higher cost. CRED Crunch, 50.
WRC-Ghana. (2007). National Water Policy- Ghana. Ministry of Water Resources, Works and Housing: Water Resources Commission-Ghana.
Yaghmaei, N. a. B., Regina. (2019). Disasters in Africa: 20 Year Review (2000-2019). CRED Crunch, 56.
Young, D. K. Y., Tamim; Dymond, L. Randy; Kibler, F. David. (2009). Virginia's Stormwater Impact Evaluation: Developing an Optimization Tool for Improved Site Development, Selection and Placement of Stormwater Runoff BPMs. Blacksburg, Virginia: Virginia Polytechnic Institute and State University , Blacksburg, Virginia.
Yüncü, O. (2008). Research Design in Architectural Design Education. (PhD), Middle East Technical University, Ankara, Turkey.

Plates 1 and 2 are available in the Supplementary Files section

No competing interests reported.

Plate1.png
Plate 1- Orthophoto image (20cm pixel size) of sections of study area; a) Buildings, b) Digitized land cover types, c)Bare areas, d) Paved Streets, e)Unpaved streets, f)Vegetation cover.
Plate2.png
Plate 2: A section of Living Labs Site1 showing bare sediment generating Football Park

Download PDF

Version 1

posted

You are reading this latest preprint version

Mitigating Perennial Floods in Cities of the Global South: A Nature Based Approach

Status:

Version 1

Abstract

Figures

1.0 Introduction

2.0 Materials and Methods

2.4 Application of Design Methods

3.0 Results

4.0 Discussions

5.0 Conclusion

Declarations

References

Plates

Additional Declarations

Supplementary Files

Status:

Version 1