Flood risk profile for Greater Monrovia Deliverable E - Final and summary report Flood risk profile for Greater Monrovia Deliverable E – Final and summary report Author(s) Bobby Russell Luisa Torres Duenas Maialen Irazoqui Apecechea Carter Draper Hélène Boisgontier Frederiek Sperna Weiland Ferdinand Diermanse Eelco Verschelling This report was prepared with support from the Global Facility for Disaster Reduction and Recovery (GFDRR) and the Japan–World Bank Program for Mainstreaming Disaster Risk Management in Developing Countries. 2 of 69 Flood risk profile for Greater Monrovia March 2021, Final © 2021 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org This work is a product of the staff of The World Bank and the Global Facility for Disaster Reduction and Recovery (GFDRR) with external contributions. The findings, analysis and conclusions expressed in this document do not necessarily reflect the views of any individual partner organization of The World Bank, its Board of Directors, or the governments they represent Although the World Bank and GFDRR make reasonable efforts to ensure all the information presented in this document is correct, its accuracy and integrity cannot be guaranteed. Use of any data or information from this document is at the user’s own risk and under no circumstances shall the World Bank, GFDRR or any of its partners be liable for any loss, damage, liability or expense incurred or suffered which is claimed to result from reliance on the data contained in this document. The boundaries, colors, denomination, and other information shown in any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions The material in this work is subject to copyright. Because The World Bank encourages dissemination of its knowledge, this work may be reproduced, in whole or in part, for noncommercial purposes as long as full attribution to this work is given. Any queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Summary This final summary report is the culmination of two stakeholder consultation workshops and extensive data analysis and flood modelling work. It supports the improved knowledge and understanding of flood risk in Greater Monrovia and contributes to the World Bank’s integrated urban development and resilience strategy in the region. The World Bank implemented the study with the support of Deltares, iLab Liberia and the stakeholders in Monrovia. The results show that direct rainfall flooding (referred to as ‘pluvial’ flooding) is the most significant flood mechanism, resulting in the highest risk. The rainy season lasts from May to October, and peak annual rainstorms can occur throughout the rainy season, and rainstorms outside the rainy season are not uncommon. The average annual rainfall in Monrovia is 5,250 mm, with June and September averaging in excess of 1000mm and July and August averaging in excess of 800mm. The highest recorded daily rainfall for Monrovia is 435mm, whilst rainstorms in excess of 100mm in a day are not uncommon. This makes Monrovia one of the wettest cities in the World. By comparison Accra, Ghana (a similar coastal city in West Africa) receives around 800mm a year with around 200mm in the rainiest month. Greater Monrovia also experiences regular flooding from both coastal flooding and flooding from the St Paul River to the north of the city. Though the most frequent coastal and river floods cover relatively smaller areas than the regular high-intensity rainstorm flooding. River flooding, referred to as ‘fluvial’ flooding, occurs during the middle of the monsoon in July and August, pushing water into the Stockton Creek from the St Paul River in the North, flooding low-lying areas along the creek and penetrating into the Mesurado estuary. The most severe fluvial floods affect larger areas of the city given the low-lying nature of the land around the Mesurado and Stockton Creek. Direct flooding from the Atlantic Ocean (overtopping) is very limited and flooding seems concentrated at New Kru Town and West Point, though erosion of the coastline seems a bigger risk to residents than flooding per se. Given climate change induced sea level rise and increased rainfall, all flood hazards will become worse in the future. Exposure, the assets and people at risk of flooding, mainly relate to the unplanned reclamation and settlement of low-lying wetlands (mangroves) which are prone to flooding. This is exacerbated by the lack of government capacity to prepare and enforce land use plans that would mitigate the exposure. Lack of basic services and unregulated informal buildings make residents especially vulnerable to flooding with lack of sanitation facilities being an important issue. Furthermore, inadequate stormwater drainage is an issue. The low gradient between most of the land and the open water, means flood water drains slowly. Most communities do not have an interconnected drainage network, so isolated areas can remain under water until it either infiltrates or evaporates. But even in the dry season patches of permanent water remain throughout the slum settlement areas. Isolated inundation of the road network also occurs throughout the city disrupting road transportation. This is primarily due to inadequate road drainage. Although the more extreme fluvial floods are of greatest depth, it is the pluvial flooding that creates the highest risk in terms of damage. This is due to the wider area that it covers and the frequency at which it occurs. Pluvial risk is more than 10 times higher than for the respective coastal and fluvial risk. Some 12% of the current population are affected by pluvial flooding depths of 10cm or more, and 5% by flooding of 1m or more. This rises for 2050 to between 13% and 16% of the population (>10cm) and 6 and 9% (>1m) depending on the population growth. Coastal risk does not increase significantly with more extreme events as there is no storm surge. Therefore, the relative damage does not increase much, but the number of residents affected becomes more significant as the inland tidal flooding is exacerbated by sea level rise in 2050. Some 1.5% of the 4 of 69 Flood risk profile for Greater Monrovia March 2021, Final current population is affected by fluvial flooding of 10cm and above or 0.3% by flooding of 1m or more. This rises for 2050 to about 3% of the population (>10cm) and 0.5% (>1m). While most of the frequent low magnitude fluvial flood events are confined to the Stockton creek area, around the 1-in-10 year flood event and above a topographical factor comes into play with the inland low- lying areas becoming inundated. This is particularly evident around the communities of Logan Town, Bong Mines Bridge, Duala Mombo Town East and New Georgia Road. The more severe (less frequent) fluvial floods result in significantly more residents being affected, with higher levels of damage. The effect of climate change also doubles the impact of fluvial flooding. Some 1.5% of the current population being affected by fluvial flooding of 10cm and above or 0.2% by flooding of 1m or more. This rises for 2050 to about 3% of the population (>10cm) and 0.5% (>1m). In monetary terms, for the current situation (2020) both coastal and fluvial risk are similar at USD1.1M/year and USD 1.4M/year in losses respectively, whereas pluvial risk estimates are USD 20.4M/year. This accounts for nearly 90% of the combined risk per year. As the value of assets grow with general economic growth, the risk also grows in monetary terms. The coastal and fluvial losses increase from around 0.03% in 2020 to potentially 0.09% of GDP for 2050 whilst pluvial losses increase from around 0.63% in 2020 to potentially 1.16% of GDP for 2050 (depending on the socio-economic and climate change scenario). Although the monetary estimates included in this assessment are based on the best available data, the values are uncertain due to the lack of primary data on actual damage costs, but that the order of magnitude is probably a reasonable approximation. Therefore, they can be used to compare areas within Greater Monrovia and thus ascertain which areas are high/low risk. This helps inform what investment levels might be realistic for cost-efficient solutions. The following key recommendations are made for further consideration by the Government of Liberia, assisted by International Financing Institutions. • Firstly, create a city-wide flood risk management committee, with representatives from relevant line ministries and city authorities, mandated by the President to coordinate and plan flood risk management measures. This would be the vehicle to discuss the potential measures. • Secondly, initiate a government, city and community wide awareness raising campaign as a basis for further consultations and to create the city-wide acceptance that action (both by the government and residents) is needed. • Thirdly, prioritise short term measures that address the highest risk mechanism (direct rainfall flooding) and those measures that can quickly reduce hazard, exposure and vulnerability for the most at-risk communities. • Fourthly, review and pre-select medium and long-term measures which may be necessary to address future hazard and risk, especially to address climate change induced increases in hazard. The planning of these measures should also take into account the future city with the short-term measures already implemented. Such medium and long-term measures may require extensive additional data collection and feasibility study which takes time to implement. This report describes pragmatic short, medium and longer term measures which may be applied to reduce risk in Greater Monrovia. They include long-term solutions, for the coastal and river flooding, but both would take significant amounts of time and financial resources to implement. So shorter term measures are also needed. A short to medium term solution would be comprehensive spatial planning based on the flood risk maps produced by this study. In the short term this would mean changing land use and zoning regulations, creating zoning maps and spatial plans for city development, new building codes for resilient construction (considering likelihood of flood depth now and in the future). This would be with the intent to implement urban consolidation, that is increasing the housing density in the safer areas of Greater Monrovia. The creation of the zoning itself would be a short-term measure whilst 5 of 69 Flood risk profile for Greater Monrovia March 2021, Final the implementation of the densification would be longer term. Requiring measures to capture low risk land value, incentives to build higher density housing developments, taxes on unused/underused land in safer zones and ‘No build zones’ in higher risk areas. Considering that flooding from direct rainfall creates the highest risk, short term measures that reduce flooding from the most frequent rainstorms is essential for Greater Monrovia. A possible solution to this is to create new, interconnected drainage systems that link traditional concrete drainage of roads with areas of existing wetlands and open water areas, creating sufficient temporary storage for stormwater. It should also be realised that even with such measures many residents will remain in medium and high-risk areas, predominantly poor slum communities with few means to reduce their vulnerability. Reducing their vulnerability could be achieved in the short term by improving the living standard in the slum communities, improving the resilience of housing to withstand flooding and reduce post-flood waterborne diseases through the provision of improved sanitation. Additionally, another short-term measure which would allow residents to make informed decisions and reduce vulnerability would be a flood warning system with community flood response plans Finally, the measures are strongly co-dependent with some measures only working effectively if key conditions are in place (such as city zoning). This reflects the interwoven nature of city living and the choices residents make for housing, such as relative distance from schools, services and work. Therefore, the measures need to be properly integrated within a supporting environment of coordinated ministries and agencies. Hence the inclusion of the multi-agency, city-wide flood risk management committee to coordinate and integrate the measures. Acknowledgements This report was developed by a team led by Robert Reid, comprising Linus Pott, Swati Sachdeva and Mathijs van Ledden of the World Bank with the financial support of the Japan –World Bank Program for Mainstreaming Disaster Risk Management in Developing Countries and the Global Facility for Disaster Reduction and Recovery (GFDRR). The study was carried out for the World Bank by a team from Deltares, led by Bobby Russell and comprised Luisa Torres Duenas, Maialen Irazoqui Apecechea, Carter Draper, F Sherman, Kayloe Frank, Hélène Boisgontier, Frederiek Sperna Weiland, Ferdinand Diermanse and Eelco Verschelling. This assessment was prepared in consultation with, and with support from, Bestman D. Toe and Chea B. Garley of the Slum Dwellers Alliance, Henry O Williams and Gemeh B Roberts from the National Disaster Management Agency, E Hodo Dossen and Alasca Y Cummings from the Ministry of Public Works, Arthur R M Becker and Fallah Nyumah of the Environmental Protection Agency, Dereck D Perkins, Daniel Sloh Sargbe and Alston Wolo from Monrovia City Corporation, Akoi Moiwai from Liberia Water and Sewer Corporation, Jannie Fahnbullah and Andy Tugbah from the Liberia Institute of Statistics and Geo-Information Services, William Dunn and Uriah Garsinii from Liberia Land Authority, Sumo Ballah from Liberia Hydrological Service, Amos J. Borbor and Albert M. Sherman of the Liberia Meteorological Services, Torwon T Yanting and Medecy Hessou from the Liberia Geographical Society, Paul Kennedy from Millennium Challenge Account – Liberia, Fred Abankwa and Christian Yeakula from Cities Alliance, Stephen B Lavalah from the Liberia Maritime Authority and Otis Kolie of the Liberia Association of International Geographers. We would like to thank the World Bank for the timely provision of extensive documentation which provided great context to the exposure profile of Monrovia, as well as the Liberia Institute for Statistics and Geographic Information Services and the Liberia Hydrological Service for providing data at short notice. 6 of 69 Flood risk profile for Greater Monrovia March 2021, Final Contents Summary 4 1 Introduction 8 2 Flood narrative 9 3 Methodology 16 3.1 Overview of modelling methodology 16 3.2 Flood Hazard Modelling 17 3.3 Flood Risk Modelling 20 4 Modelling results 23 4.1 Separate Fluvial, Pluvial and Coastal Hazard 23 4.2 Combined hazard 31 4.3 Risk 32 4.3.1 Risk results 32 4.3.2 Risk maps 35 5 Potential high-level risk reduction measures 40 5.1 Physical coastal defence structures 41 5.2 Physical riverine defence structures 43 5.3 City zoning and land use planning 45 5.4 Blue-green-grey drainage measures 48 5.5 Living with water 52 5.6 Flood Early Warning and Response 55 5.7 Greater Monrovia City flood risk management committee 57 6 Conclusions and recommendations 59 6.1 Conclusions 59 6.2 Recommendations 64 7 References 66 7 of 69 Flood risk profile for Greater Monrovia March 2021, Final 1 Introduction This report supports the improved knowledge and understanding of flood risk in Greater Monrovia, contributing to the World Bank’s integrated urban development and resilience strategy for the region. It describes the flood risk profile of the separate and combined risk from rainfall, river and coastal flooding for the city. Mapping the extent and magnitude of flooding from the three mechanisms, the population and assets at risk and identifies flood hotspots of high hazard and exposure. It is intended to assist in creating a resilient and economically vibrant city. The work has a direct relationship to the World Bank’s Liberia Country Partnership Framework (objective 8) on urban development and the Open Cities Africa initiative. It also recognises the importance of building urban resilience in post-conflict states, as safe and prosperous cities create the necessary stability to move on from conflict. This report is therefore an important stepping-stone, identifying the profile of flood risk and possible future measures, which can be further developed through site-specific analysis on specific vulnerability, risk and mitigation actions, as well as technical and economic feasibility studies. The risk assessment started in October 2019 with an Inception Report (Deliverable A) which detailed the initial stakeholder consultations, review of available information, data collection, modelling methodology and data gap filling strategy. This was followed by hazard and risk modelling work which included sourcing and procuring a new digital terrain model for the city. The result of this work was the Risk Maps and Report (Deliverable B/C) which described the hazard and risk assessment process, the current and projected hazard and exposure risks considering climate change scenarios and population growth projections. The results were compiled in a Geo- spatial database (Deliverable D). These results were presented and discussed with the stakeholders at the final workshop, where the potential high-level risk reduction measures were discussed. This Final and Summary Report (Deliverable E) describes the flood risk profile of Greater Monrovia with recommendations for the way forward. In this Final and Summary Report, Chapter 1 provides an introduction to the assignment, Chapter 2 is a detailed narrative of how flood risk occurs in Greater Monrovia, Chapter 3 summarises the modelling methodology (for more detail the previous flood risk assessment report should be consulted) Chapter 4 presents the hazard and risk results. Chapter 5 describes the high-level potential flood risk reduction measures discussed with the stakeholders, whilst Chapter 6 gives the conclusions and recommendations. 8 of 69 Flood risk profile for Greater Monrovia March 2021, Final 2 Flood narrative The flood narrative describes the process by which flood risk manifests in Greater Monrovia, accounting for the flood hazard mechanism itself and the exposure of assets and people as well as the vulnerability to the flood hazard. The narrative is derived partly from the stakeholder consultations, field work and the flood modelling study. The study area (Figure 2-1) is defined as ‘Greater Monrovia’, though the legal designation does not exist and the boundaries are not defined in law, it is understood to mean the two municipalities of Monrovia City Corporation and Paynesville City Corporation. The urban area historically centred on the high ground to the south (known as Central Monrovia) and the coastal strip to the east, as well as the west facing side of Bushrod Island (which is made up of New Kru Town, Logan Town, Clara Town and the freeport area and harbour), which is bounded by the St Paul and Mesurado river estuaries to the north and south and Stockton Creek to the east. The map below also shows the patches of low-lying wetlands, made up of mangroves in the Mesurado and patches of swampy ground. Stockton Creek Mesurado Figure 2-1. Map of Greater Monrovia. 9 of 69 Flood risk profile for Greater Monrovia March 2021, Final In more recent times the city has grown around the low-lying swampy eastern parts of Bushrod Island and the periphery of the Mesurado estuary and Stockton Creek areas (Figure 2-2), which are typically reclaimed mangrove swamp. Furthermore, the city has expanded to the east in Paynesville. Areas of permanent standing water are a feature of the slum settlements, which are used for waste dumping, creating a health hazard (Figure 2-3). Figure 2-2. The Stockton Creek photographed from Jamaica road, Logan Town. Figure 2-3. Isolated areas of open water, such as here at West Point, become waste collection areas, both solid waste, and human (note the latrine building on the left). The extreme nature of rainfall in Monrovia should be recognised, with a long rainy season lasting from May to October, and peak annual rainstorms occurring throughout the rainy season, and rainstorms outside the ‘rainy season’ are not uncommon (Figure 2-4). The average annual rainfall1 in Monrovia is 5,250 mm, with June and September averaging in excess of 1000mm and July and August averaging in excess of 800mm2. The highest recorded daily rainfall for Monrovia is —————————————— 1 Average and highest from 2007 to 2016 with 2011 and 2012 missing. 2 Data from Liberia Hydrological Service. 10 of 69 Flood risk profile for Greater Monrovia March 2021, Final 435mm, whilst rainstorms in excess of 100mm in a day are not uncommon. This makes Monrovia one of the wettest cities in the World. This direct rainfall is the largest contributor to the flood hazard, causing localised flooding throughout the city. Figure 2-4. Annual rainfall patterns in Monrovia (LHS station). The St Paul River is also a significant source of flooding during the monsoon season. The river flow ranges from around 25 cubic meters per second in the dry season to 2000 cubic meters per second in the monsoon. The high-water levels in the St Paul River cause flood water to flow south through Stockton Creek into the Mesurado Estuary (Figure 2-5), causing flooding in the communities of Caldwell and New Georgia Road and the low-lying eastern side of Bushrod Island. Figure 2-5. Topoe Village and the mangroves of the Mesurado estuary after the floods of August 2018. Coastal flooding mostly effects the west facing side of West Point, a sand spit at the mouth of the Mesurado River, and New Kru Town at the west facing north end of Bushrod Island. The coastline in between is protected by the Freeport’s breakwaters. Although direct coastal flooding is not significant in terms of the area flooded. Wave heights during storms are estimated to range from 2 to 2.7 meters, whilst increased sea levels due to storm surge are not significant. The tide ranges from around 0.6 meters and 1.8 meters, with the highest tides reaching into the Mesurado and Stockton Creek. The primary coastal hazard mechanism relates to the longshore erosion and transportation of beach material at West Point and New Kru Town due to the waves coming from the south which has been documented in a UNDP report (2019). 11 of 69 Flood risk profile for Greater Monrovia March 2021, Final As the city expanded from the 1980s onwards, the slum settlements have grown to accommodate two thirds of Greater Monrovia’s population3, often occupying marginal land on the eastern side of Bushrod Island, West Point and the fringe of the Mesurado wetland. Throughout the city, patches of almost permanently wet ground (remnants of wetlands) remain, often only a few 10s of square meters, between houses. The largest areas of wetlands are formed by the tributaries of the Mesurado river and Du-Junk river (to the east of the city boundary). Figure 2-6. Raised walkways in Doe Community, note the raised building footprints above the flood level of the two buildings on the right, photographed after the 2018 floods. Figure 2-7. Patches of semi-permanently wet ground between buildings in Topoe Village, photographed during the 2019 dry season. Although these areas evidently have a high groundwater table many residents have settled there. Adaptation is high, with residents raising the ground floor and foundations of the buildings more than 10cm above the surrounding ground surface (and hence the surrounding flood waters). For illustration see Figure 2-6 and Figure 2-7, showing firstly, the raised houses and semi-formal walkways, and secondly the patches of swampy ground and informal raised walkways. —————————————— 3 World Bank (2020b). 12 of 69 Flood risk profile for Greater Monrovia March 2021, Final Although the Mesurado estuary is an internationally protected area under the Ramsar convention, the stakeholders report that there is no effective enforcement. Currently both the Monrovia City Corporation, Paynesville City Corporation and the Ministry of Public Works have the right to issue building permits but in reality, there is little awareness amongst residents of the need to apply for permits and stakeholders report that there are effectively no consequences if developers build without permits. Additionally, land use planning is currently not done, with the Liberia Land Authority having only been established in 2016, subsuming the roles of several organisations and being mandated to create city zoning. So the city expanded in an unregulated way, with houses being built in protected areas and flood prone locations, thereby increasing the flood risk. The slum settlements have few services. None of the slums have formal water supply and sanitation and informal pit latrines raised above open water are the norm, with 80% of residents practicing open defecation. This has consequences for water-borne diseases during post-flood conditions. Solid waste collection and disposal is very basic with 95% of residents having no access to solid waste collection4, so garbage accumulates in open waterways and drains, blocking their ability to convey stormwater. The densely populated slum communities are also more vulnerable to epidemics, evidenced by the 2014-2015 Ebola Outbreak where many of the flood prone communities were affected. This is also evident from the 2020 COVID-19 pandemic5. Houses in the slum communities are typically constructed out of corrugated metal sheets or concrete blocks, and residents make extensive efforts to raise the internal floor level, either by compacting earth, garbage or anything else to hand (such as saw dust in Figure 2-8). Stakeholders report that the repair and strengthening of houses after a flood is an ongoing effort for residents. Although these buildings in the slums are relatively low-cost, the building and repair costs represent a significant investment for these poor communities. Also the poorly constructed single storey buildings are much more vulnerable to flood damage and residents have few means to store possessions above the flood water. Many residents in the slum settlements earn an income as petty traders or day labourers and flooding can prevent them from accessing work in the markets, Freeport or Downtown area. Moreover, residents report that they have to wade through polluted flood waters in their work or school clothes. This is exacerbated by the fact that typically residents must travel long distances to school or work, with a third reportedly walking 4. Considering that the Mesurado effectively divides the city in two, long travel times is a key factor for residents in choosing a location to live. Proximity to the downtown areas and markets being a major factor in their housing choice with ‘flood risk’ being a much smaller factor4. It should also be recognised that the impacts can be different for men and women, either in nature or in severity. Although a full gender disaggregated impact assessment was not performed we know that adult women in Monrovia are more likely than men to have no primary or secondary education4, women also have a reduced decision-making role in the community and are often restricted in their actions due to child care obligations6. This can reduce their capacity to respond to and recover from flooding of their homes. Women are also more severely impacted in relation to access to sanitation facilities, which during floods can become impossible. Women also have limited access to healthcare with high maternal mortality6. —————————————— 4 World Bank (2020b) 5 World Bank (2020a) 6 Nilsson et al, (2018). 13 of 69 Flood risk profile for Greater Monrovia March 2021, Final Figure 2-8. Resident of Topoe Village spreading and compacting saw dust to raise the floor of his building. Flooding not only affects the slum settlements in the former wetlands but also throughout the formal parts of the city. News articles report on torrential rain flooding streets and residential areas. Reports from 2018 describe three separate flood events in June, July and August, which caused residents to temporarily evacuate their homes with the household possessions they could carry7. This narrative was substantiated by the inception workshop and field visit where stakeholders and residents described leaving the flood affected areas to stay with relatives. Other residents described simply sitting out the flooding till it receded. Although, roads are often raised above the ground level, they are often hydrologically isolated, with rainwater becoming trapped on the road surface, unable to drain away. Stakeholders reported that drainage is simply insufficient, with major roads through the city being blocked by floodwater. Such floods can be very disruptive. Notably the junction between Somalia Drive and United Nations Drive regularly floods, cutting off the New Georgia and Gardensville areas from both the Freeport and Downtown areas, also blocking the main north – south artery. Figure 2-9. News article from August 2018. —————————————— 7 https://www.aljazeera.com/news/2018/08/flooding-leaves-parts-liberia-capital-water-180818085941804.html http://floodlist.com/africa/liberia-flooding-montserrado-margibi-monrovia-july-2018 https://themonroviatimes.com/2018/06/29/flooded-monrovia/ 14 of 69 Flood risk profile for Greater Monrovia March 2021, Final The Ministry of Public Works reported that there has been little construction of stormwater drains in the city, other than that built under a Japanese funded project. Most drains date from the 1980s and there is no comprehensive database of either the extent nor condition of drains. The National Disaster Management Agency reported that responsibilities for flood risk management sit with many organisations, and that implementation of risk management measures remains a major institutional challenge. Responsibility for flood knowledge and warning sits with both the Liberia Meteorological Service (for weather forecasts) and the Liberia Hydrological Service (for monitoring the rivers), and they sit under different line ministries. Currently the National Disaster Management Agency, other organisations and communities do not receive adequate flood warnings to be able to react effectively. This lack of preparedness means that flood impacts are worse and recovery slower than if flood response plans are in place and implemented. The National Disaster Management Agency reported that there is simply a lack of awareness amongst residents about flooding and the risk that is created when settlers build houses in the low-lying former wetlands. 15 of 69 Flood risk profile for Greater Monrovia March 2021, Final 3 Methodology 3.1 Overview of modelling methodology The methodology employed a chain of A - WFLOW Hydrological models to simulate the flood hazard from B - XBEACH Wave Model Model Coast separate and combined mechanisms, St Paul River namely rainfall, coast and river, resulting in Downstream Upstream Coastal flood hazard maps. This analysis was river in-flow conditions carried out for different return periods and climate change scenarios for the present situation (2020) and for a future time C - SFINCS Hydrodynamic Model horizon (2050). These hazard maps then Coast / River / Rain feed into the risk assessment model which combines exposure maps and vulnerability Flood hazard maps for relationships to produce risk maps for each each return period and hazard and for combined hazards. mechanism This follows a 4-step modelling approach to D - FIAT Risk Model derive flood maps for separate and Coast / River / Rain combined events (figure left) (A) a rainfall-runoff hydrological model using the wflow software to derive upstream Separate and combined risk river boundary conditions (St Paul River) for maps every return period. (B) a nearshore hydraulic model using the Figure 3-1. Model framework. Blue rectangles indicate XBeach software to derive the sea state model steps. through run-up analysis of waves from a global dataset. (C) a 2-Dimension hydrodynamic model using the SFINCS software, to derive separate and combined fluvial / pluvial / coastal flood hazard maps for each return period. Providing the simulated flood depths and inundated area. The input for the SFINCS model at the coastal boundaries are the water levels and wave data from the XBeach transects, while for the inland part, the input is precipitation (time-varying), St Paul River flow (constant) and soil infiltration rates. The topography was simulated using a newly purchased 0.5m resolution digital terrain model. (D) By combining the hazard (multiple return periods and future climate scenarios), exposure (current situation and future population scenarios) and vulnerability maps in Delft-FIAT model to generate the separate and combined flood risk maps. A summary of methodology is provided below, for further, more detailed information the Risk Report can be consulted8. —————————————— 8 Deltares (2020). 16 of 69 Flood risk profile for Greater Monrovia March 2021, Final 3.2 Flood Hazard Modelling The modelling requires a large amount of data, most of which is not locally available and needed to be sourced externally. Critical to such a modelling study is terrain data. Therefore, Deltares sourced a 0.5m resolution Digital Terrain Model (DTM) from AW3D, which was purchased on behalf of the World Bank from Restec (Figure 3-2). AW3D estimate that it is vertically and horizontally accurate to between 2 and 3 meters. Figure 3-2. 0.5m AW3D Enhanced DTM (with surface water masked out) and elevation above mean sea level (AMSL) referenced to EGM2008 geoid. The MERIT9 Digital Elevation Model (DEM) was used to create the wflow hydrological model of the St Paul River catchment. The lower spatial resolution (approx. 90m) is not a limitation when used for distributed hydrological modelling for such large basins and it was re-scaled to a 1km grid. For the coastal areas the General Bathymetric Chart of the Oceans (GEBCO) 2020 Grid was used to represent the bathymetry10. Given that there are no local tide or wave measurements, we derived the tidal data and offshore wave conditions using global datasets. For tides and surges, we use the Global Tide and Surge Model 11 which provide us with a coastal water-level dataset of 10-minute time-series at a 2.5km resolution, which captures storm processes at the city scale. For wave data, the ERA5 global dataset, which again covers 1979 to present and has a 0.5-degree spatial resolution on an hourly basis. It provides the maximum significant wave height, wave peak period and wave direction in a gridded format for both wind and swell generated waves combined. The daily river flow and rainfall data of Liberia Hydrological Service (LHS) for a limited number of locations in the St Paul River Basin was obtained. Only the station at Haindi has a reasonable length (2012 to present) but is not of sufficient length to derive extreme value analysis and thus return periods, but the combination of the wflow hydrological model and ERA5 climate data generated a 40 year record of river flow upstream of the city from which the return periods could be derived. This characterised the flooding frequency and magnitude from the St Paul River. —————————————— 9 Multi-Error-Removed Improved-Terrain (MERIT) : http://hydro.iis.u-tokyo.ac.jp/~yamadai/MERIT_DEM/ 10 https://www.gebco.net/data_and_products/gridded_bathymetry_data/gebco_2020/ 11 Muis et al. 2020 17 of 69 Flood risk profile for Greater Monrovia March 2021, Final The SFINCS flood model for the Greater Monrovia area is a 2-dimensional hydrodynamic model with a rectilinear grid and a horizontal resolution of 10 meters. The bed levels of the inland waterbodies (rivers, estuaries, creeks) had to be manually prescribed in the model, as the actual bed levels are not known. It is important to note that this is a limitation of the model and especially the modelled river flooding extent and depths will be sensitive to these bed levels. The boundary conditions applied that characterise the extreme pluvial, fluvial and coastal states were derived from extreme value analysis (Table 3-1) of the rainfall data, the modelled St Paul flow data and the modelled tidal/surge data and wave data. These were applied for the river mechanism (St Paul River, here we assumed a constant value for the discharge over 24 hours), rainfall mechanism (applied as mm/hour for a 24-hour high intensity rain event on the city) and the coastal mechanism (surge/wave/tide conditions over 24 hours). An 18% reduction was applied to the return period values for the St Paul River, given the overestimation of the hydrological model compared to the observed discharge values. Table 3-1. Return period values used in the flood hazard modelling. Return St Paul Rainfall Coastal Significant Tidal Tidal Range / period River flow (mm/d) surge (m) wave height conditions Amplitude (m3/s) (m) 1 1950 258 0.09 2.0 Minimal 0.64m / 0.32m 2 2435 304 0.09 2.1 Mean 1.16m / 0.58m 5 3069 363 0.1 2.3 Maximal 1.77m / 0.885m 10 3546 409 0.1 2.4 50 4638 514 0.11 2.7 The three hazards are run separately for the five return periods: • 1 in 1 year (or 100% chance of occurring in any given year) • 1 in 2 year (or 50% chance) • 1 in 5 year (or 20% chance) • 1 in 10 year (or 10% chance) • 1 in 50 year (or 2% chance) These are also referred to as T1, T2, T5, etc for simplicity. A worst-case-scenario is also run combining the 50-year return periods for the three hazards. Climate change The climate change impact assessment is based on climate datasets from the 5th Assessment report of the Intergovernmental Panel on Climate Change (IPCC-AR5; IPCC, 201312). For this study two RCPs were selected, with a base-year of 2020 and a future time-horizon of 2050: • RCP4.5 (the moderate change scenario) • RCP8.5 (the worst-case scenario) For the calculation of future changes in the flow of the St Paul River we used the precipitation and temperature datasets from General Circulation Models (GCMs) from the international inter- sectoral impact model inter-comparison project (ISI-MIP)13. Long-term average changes in monthly precipitation and potential evaporation were derived from the historical and future GCM simulations (see Table 3-2). The changes were applied to hydrological model of the St Paul River which showed that monsoon river discharges may start later in April / May / June than currently. —————————————— 12 IPCC. 2013. 13 Hempel et al (2013). 18 of 69 Flood risk profile for Greater Monrovia March 2021, Final River flows for August and September are likely to increase on average with climate change but do not differ significantly between climate scenarios. Table 3-2. Projected future changes (percentage increase) in monthly precipitation over the St Paul basin for the wet months. Scenario Jun Jul Aug Moderate climate change 1% 8% 2% Extreme climate change 1% 2% 7% For every degree increase in air temperature there is a 7 %/ºC more moisture capacity in the air14, therefore if air temperature increases due to climate change then rainfall increases. For the most severe events a scaling rate of 14 %/ºC has been observed15 in situations where enough moisture can be generated. The latter is expected to be the case because Monrovia is located at the coast. Based on this principle, the following precipitation change factors were derived for extreme rainfall over Monrovia (Table 3-3) and applied to the rainfall in the SFINCS flood hazard model. Table 3-3. Future climate change scenarios and related change in precipitation extremes over Monrovia. Scenario %/°C Change in Temp (°C) Precipitation change (%) Moderate climate change 7 1.3 9.1 Extreme climate change 14 2.3 32.2 The Jevrejeva global dataset of regional Sea Level Rise for each climate scenario was used for the 2050 time-horizon. Local SLR values are given in Table 3-4. These sea level rise values were applied to the coastal boundary conditions of the SFINCS model for all future scenarios. Table 3-4. Sea Level Rise for Monrovia as determined by Jevrejeva, 2017. Scenario Sea level rise by 2050 Moderate climate change 0.26 m Extreme climate change 0.30 m —————————————— 14 Allen, M., and Ingram, W. J. (2002). 15 Lenderink, G., and Van Meijgaard, E. (2008). 19 of 69 Flood risk profile for Greater Monrovia March 2021, Final 3.3 Flood Risk Modelling Flood risk modelling is based on combining hazard (flood depth and extent maps), exposure (maps of assets) and vulnerability (the relative damage for each asset based on depth-damage functions). It also accounts for non-monetarised impacts such as number of residents impacted, kilometers of roads impacted, number of schools and medical facilities impacted. Asset damage valuation was based on the Joint Research Centre (JRC) database16 and the R5 report17. The former provides construction cost estimates in 2010 Euros for every country. The latter gives the ratio of construction costs between ‘formal’ (i.e. 1 = permanent) and ‘informal’ (i.e. 2 = semi- permanent and 3 = temporary) which aided in estimating the ratios for Monrovia. Therefore a ratio of 1 is applied for the JRC values for ‘Concrete / stone / cement block’ properties in ‘Formal’ areas, whilst the following ratios were applied for ‘Informal’ areas: • Clay brick masonry / cement blocks = 0.4 • Zinc/iron/tin = 0.3 • Mud+sticks / mud bricks = 0.2 • Reed / bamboo = 0.1 Whether a particular residential property is ‘formal’ or ‘informal’ was taken from a gridded map provided by the World Bank. The building material information was verified using the 2008 Census of Liberia which showed a similar distribution. As the OSM dataset does not have material type, we then assumed that all residential properties in Monrovia have the percentage distribution given in the Census, which was then applied to the OSM dataset, assuming that in the ‘Formal’ areas only ‘Concrete / stone / cement block’ buildings are found, and in the ‘Informal’ areas the material distribution from the census applies (corrected for the formal areas). For Monrovia, the depth- damage fraction for “urban-houses” for Mozambique was used as a benchmark18. This damage fraction assumes 100% damages for a flooding depth of 3.0 m and it assumes the asset is located at ground level. Since many of the assets in the Greater Monrovia Region are elevated, it was decided to shift the damage fraction by 20cm to reflect that in average most of the buildings in Monrovia are slightly elevated. Future projections To be able to assess the risk for the future (2050), we need to identify how asset valuation and population (exposure) will change over time. For this the Shared Socioeconomic Pathways (SSPs) were used19. These are five scenarios projecting socioeconomic global changes including population and economic growth. They are used to derive greenhouse gas emissions scenarios for forcing General Circulation Models (Climate models) and have been applied to the IPCC sixth assessment report on climate change. The scenarios represent different world outlooks depending on different policies to geo-politics, economic development and clean energy. They as follows: SSP1: Sustainability (Taking the Green Road) SSP2: Middle of the Road SSP3: Regional Rivalry (A Rocky Road) SSP4: Inequality (A Road divided) SSP5: Fossil-fueled Development (Taking the Highway) SSP3 and SSP 5 have been selected as the representative socioeconomic pathways most closely matching the RCP4.5 and RCP8.5 being used for the climate change projections. —————————————— 16 Huizinga et al., 2017 17 ImageCat, 2017 18 Huizinga et al., 2017 19 UNECE, 2019. https://www.unece.org/fileadmin/DAM/energy/se/pdfs/CSE/PATHWAYS/2019/ws_Consult_14_15.May.2019/supp_doc/SSP2 _Overview.pdf 20 of 69 Flood risk profile for Greater Monrovia March 2021, Final SSP3 or the regional rivalry scenario represents a world with resurgent nationalism and increased focus on domestic priorities and internal energy security. Population growth is low in industrialized and high in developing countries and a low international prioritisation for environmental policies results in continued environmental degradation. SSP5 or the fossil-fuelled development scenario represents a world with increasing faith in competitive markets and technical innovation as the path to sustainable development, supported by the continued exploitation of abundant fossil fuel resources and the adoption of resource and energy intensive lifestyles. The SSP3 and SPP5 scenario database has been accessed20 and we extracted the GDP and population projections for Liberia the years between 2015 and 2050, deriving the rate of change. Then we applied the rate of change for those years to the population data (described above) to calculate the population for 2020 and 2050. Risk calculations The risk analysis itself was done with Delft-FIAT model (or Flood Impact Assessment Tool) which uses the flood maps, exposure datasets and vulnerability information as an input (see Figure 3-3)21. Producing tables and maps of flood risk for Greater Monrovia, for both the current situation and for a future time horizon considering climate change and future population/GDP increase. Figure 3-3. Overview of damage and risk calculations in Delft-FIAT. We also count the number of Residents, Health Services and Schools impacted by the flooding. This is computed as the yearly number of people affected by a flood of more than 30cm (flood nuisance), and the yearly number of people affected by a flood of more than 1m (severe impact). The mapping of the roads, health services and schools was provided by the World Bank. The gridded population data was created by combining a population analysis dataset produced by GMV for the World Bank for 2015, with the World Settlement Footprint (WSF)22, including building footprints, and the global population data from the Global Human Settlement (GHS) dataset 23, available at a coarser resolution (250 m). Providing a complete coverage of the city at a theoretical 10m resolution dataset of population. We employed the commonly used flood risk assessment approach24 to relate flood depth and damage for a specific asset category or type using a depth-damage-curve, rather than for individual properties (with all asset types assumed to be elevated 20 cm above the ground). The exposure elements are classified into several categories described Table 3-5. —————————————— 20 https://tntcat.iiasa.ac.at/SspDb/dsd?Action=htmlpage&page=10 21 Slager et al (2016). 22 Marconcini et al., 2019 23 Pesaresi et al., 2016; Freire et al., 2016 24 Merz et al., 2010; Ward et al., 2011 21 of 69 Flood risk profile for Greater Monrovia March 2021, Final The risk was computed taking into account the following return periods: 1, 2, 5, 10 and 50 years. The damages associated with the 0.3yr return period were subtracted from the risk calculation to prevent overestimation of the damages, under the assumption that such frequent events will not cause actual material damages as the residents would have adapted to the conditions. However, it is also a ‘calibration’ of sorts as it is not known for sure where this ‘zero damage’ point lies. Table 3-5. Exposure classifications for assets. # Building material Categories Formalitya Depreciated Repair Content Material Total Construction ratioc Ratioc Qualityd Damagee Costs JRCb (% from (% from (ratio to (USD/m2) (EURO)/m2 Survey) Survey) concrete) 1 Concrete / stone / Formal EUR 134.33 40% 25% 1 USD 134 cement blocks 2 Clay brick Informal EUR 53.73 40% 25% 0.4 USD 53 masonry / cement blocks Residential 3 Zinc/iron/tin Informal EUR 40.30 70% 25% 0.3 USD 59 4 Mud+sticks / mud Informal EUR 26.87 100% 25% 0.2 USD 51 bricks 5 Reed / bamboo Informal EUR 13.43 100% 25% 0.1 USD 26 6 Concrete / stone / Commercial - EUR 159.79 30% 50% 1 USD 196 cement blocks 7 Concrete / stone / Public - EUR 159.79 30% 50% 1 USD 196 cement blocks 8 Concrete / stone / Industrial - EUR 121.30 30% 75% 1 USD 195 cement blocks 9 Concrete / stone / Health - EUR 159.79 30% 50% 1 USD 196 cement blocks Services 10 Concrete / stone / Schools - EUR 159.79 30% 50% 1 USD 196 cement blocks a – Formal/Informal areas are defined in a GIS layer provided by the World Bank. b – A 60% depreciation is applied to the JRC figures. c – Both the content and repair ratio is taken from the field survey conducted by iLab Liberia. d – The reduction factor applied to the different building materials was estimated based on the R5 report which produced risk estimates for different countries in Africa. e – JRC values are in 2010 Euros which are converted to Liberian Dollars and inflation applied to reach 2019 Liberia Dollar values, before converting to US Dollars. 22 of 69 Flood risk profile for Greater Monrovia March 2021, Final 4 Modelling results 4.1 Separate Fluvial, Pluvial and Coastal Hazard Below we show the separate hazard associated with the separate mechanisms for the current climate, as well as a comparison of the T1 and T50 events for the current, future moderate and future extreme climate, to show the spatial variation between the most frequent and least frequent flood events and therefore least and most severe. One of the most striking conclusions from the hazard assessment (Figure 4-1) is the relatively broad city level impact of the pluvial hazard, with flooding throughout Greater Monrovia. Direct coastal flooding is limited to a narrow area in New Kru Town and West Point with tidal flooding around the low-lying areas around Mesurado estuary, Stockton creek and the St. Paul river mouth. Pluvial flooding shows a complex spatial pattern, with clearly identifiable accumulation areas. Fluvial flooding severely affects areas around Stockton Creek. With climate change these flood hazards all become worse with deeper flood depths and therefore greater inundated areas. Interestingly the difference between the moderate (RCP4.5) and extreme (RCP8.5) is quite small as the effect of climate change over the monthly rainfall over the large St Paul basin is quite similar (Figure 4-2 to Figure 4-7). Flood hazard and risk analysis considers both the probability (or chance) of a flood occurring and the size of the flood (depth and extent). This is defined as the floods ‘return period’. These are typically referred to as a flood with a 1-in-X year chance of occurring. With the floods with a less likely chance of occurring having greater depth and extent. For this assessment the return periods assessed are: • T1 (a 1 in 1 year flood or a flood with a 100% chance of occurring in any given year) • T2 (a 1 in 2 year flood or a flood with a 50% chance of occurring in any given year) • T5 (a 1 in 5 year flood or a flood with a 20% chance of occurring in any given year) • T10 (a 1 in 10 year flood or a flood with a 10% chance of occurring in any given year) • T50 (a 1 in 50 year flood or a flood with a 2% chance of occurring in any given year) 23 of 69 Flood risk profile for Greater Monrovia March 2021, Final Figure 4-1 Maximum inundation maps for coastal (top left), fluvial (top right), and pluvial (bottom) flood events with 50-year return period 24 of 69 Flood risk profile for Greater Monrovia March 2021, Final Year 2020 – Current Situation Year 2050 – Moderate Climate Scenario (RCP 4.5) Year 2050 – Extreme Climate Scenario (RCP 8.5) Figure 4-2 Pluvial flood hazard for T1 current climate, moderate and extreme climate change. 25 of 69 Flood risk profile for Greater Monrovia March 2021, Final Year 2020 – Current Situation Year 2050 – Moderate Climate Scenario (RCP 4.5) Year 2050 – Extreme Climate Scenario (RCP 8.5) Figure 4-3 Pluvial flood hazard for T50 current climate, moderate and extreme climate change. 26 of 69 Flood risk profile for Greater Monrovia March 2021, Final Year 2020 – Current Situation Year 2050 – Moderate Climate Scenario (RCP 4.5) Year 2050 – Extreme Climate Scenario (RCP 8.5) Figure 4-4 Fluvial flood hazard for T1 current climate, moderate and extreme climate change. 27 of 69 Flood risk profile for Greater Monrovia March 2021, Final Year 2020 – Current Situation Year 2050 – Moderate Climate Scenario (RCP 4.5) Year 2050 – Extreme Climate Scenario (RCP 8.5) Figure 4-5 Fluvial flood hazard for T50 current climate, moderate and extreme climate change. 28 of 69 Flood risk profile for Greater Monrovia March 2021, Final Year 2020 – Current Situation Year 2050 – Moderate Climate Scenario (RCP 4.5) Year 2050 – Extreme Climate Scenario (RCP 8.5) Figure 4-6 Coastal flood hazard for T1 current climate, moderate and extreme climate change. 29 of 69 Flood risk profile for Greater Monrovia March 2021, Final Year 2020 – Current Situation Year 2050 – Moderate Climate Scenario (RCP 4.5) Year 2050 – Extreme Climate Scenario (RCP 8.5) Figure 4-7 Coastal flood hazard for T50 current climate, moderate and extreme climate change. 30 of 69 Flood risk profile for Greater Monrovia March 2021, Final 4.2 Combined hazard Joint analysis of rainfall time series in Monrovia, sea water levels and St Paul river discharges was carried out. Looking at times floods are known to have occurred such as August 2018 and October 2018. Obviously, this is only possible for times where all three data sets overlap but we now have a better understanding of the joint occurrence probability of compound events occurring. The following conclusions are made concerning the combined hazard. a. Surge timing relative to tide Since the surge amplitude is small, the timing of the surge relative to the tide does not show to be a critical factor to consider. The inundation depth differences are below 10cm and the flooded surface areas are very similar. b. High volume rainfall event vs high intensity rainfall event The inundation map comparison between a 24-hour simulation with time-varying, high intensity rainfall events, and a less intense but constant rainfall rate over a period of 7 days shows that short, intense rainfall events lead to significantly higher inundation. c. Rainfall timing relative to peak surge/tide timing Some small differences are observed when peak timing in local rainfall is delayed relative to the peak coastal water-level timing. When peak timings coincide, the total water depth at that moment feels the maximum effects of both signals (peak in runoff due to rainfall and peak in backwater due to high tide). When the peaks in rainfall and high tide are more out of sync, the maximum water depth will obviously be lower. Also, as the combined event is more spread out in time, infiltration will be slightly more effective in reducing the maximum inundation depth in those areas that experience both coastal and pluvial flooding. In relative terms, however, the differences are very small, below 10 cm. d. Discharge timing relative to peak surge/tide timing Another phenomenon observed is that high river discharges lead to tidal damping at the St. Paul and Mesurado estuary mouths for compound flooding cases. However, these areas are mostly dominated by fluvial flooding, and the relative impact of the dampening is minimal. For high discharge scenarios, the differences due to relative timing of the local rainfall and high tide peaks reduce for those areas that experience combined pluvial, fluvial and coastal flooding. The reason behind this is that, while coastal and pluvial flooding take place on a relatively short time scale, the river flood wave takes place over a much longer time frame (the river discharge was consequently prescribed as a constant value in the simulations). High discharges therefore completely govern the infiltration rates during the simulations, thus reducing the impact of relative timing of the peaks in local rainfall and high tide in areas affected by all three components. 31 of 69 Flood risk profile for Greater Monrovia March 2021, Final 4.3 Risk The following section shows the flood risk results for: • Pluvial flooding • Fluvial flooding • Coastal flooding • All combined flooding (summation of 3 hazards). Scenarios To estimate flood risk induced by the 3 types of flooding (Coastal, Pluvial and Fluvial) for the year 2050, three scenarios were considered in order to estimate future risks in comparison to the current (2020) situation): Scenario 1: Without socio-economic growth (SSP) but considering changes due to climate change: Exposure and vulnerability datasets were kept constant while accounting for 2 RCP scenarios (RCP 4.5 and RCP 8.5) Scenario 2: With socio-economic growth but without climate change: Exposure and vulnerability datasets are changed according to the SSP scenario chosen (SSP3 or SSP5) but hazard remains constant. Scenario 3: With both considering socio-economic growth and climate change: Exposure and vulnerability datasets change according to the SSP scenario chosen (SSP3 or SSP5) and hazard information varies according to the climate change scenario (RCP 4.5 and RCP 8.5). For the combined case, impact was computed assuming all 3 hazards are independent of each other, given the assessment in Section 4.2 of the combined events. All results can be observed in Section 3. In the present report, the following indicators were used to define the impact and resilience potentials of the Greater Monrovia Region: • Expected Averaged Annual Losses (EAAL) [2020 USD ($)] • EAAL with respect to 2018 GDP25 [%] • Expected Averaged Annual People (EAAP) affected [number of residents] by floods higher than 10cm and 1m • Expected number of schools affected by floods higher than 10cm and 1m. • Expected number of health units affected by floods higher than 10cm and 1m. • Expected kilometers of road affected by floods higher than 10cm and 1m. • EAAP with respect to the total population of 200826 [%] Risk indicators are reported for each individual hazard and they are added up to show the total risk due to flooding in the greater Monrovia region. Risk maps (aggregating result at community level) are depicted for the 3 individual hazards and the summation only for scenario 1 (without considering socio-economic growth but considering changes due to climate change). The rest of the risk results (scenario 2 and 3) are summarized in the tables below. 4.3.1 Risk results The overall risk results are derived from the curve created by the risk calculated for each return period. Thus giving a USD losses value for a flood magnitude of known return period. The total risk being the area under the curve and the Expected Annual Average Losses being the total risk divided by the number years. Following the agreed methodology the T0.3 event, being an event that in reality probably results in no material damages, is excluded from the risk integral. —————————————— 25 As estimated from the World Bank: 3.071 Billion 26 As reported from the 2008 Census of Libera: 970,824 people 32 of 69 Flood risk profile for Greater Monrovia March 2021, Final Scenario 1 – Climate change and no socio-economic growth For this scenario, the hazard changed according to climate change (2020 and 2050 flood map for each return period and climate scenario (RCP 4.5 and RCP 8.5)) and the exposure value was kept constant (at 2020 values) The risk metrics in this scenario are shown in Table 4-1. The indicators used are ‘Expected Averaged Annual Losses’ (EAAL) with 2020 USDs (rounded off to the nearest thousand) and with respect to 2018 GDP [%], as well as Expected Averaged Annual People (EAAP) affected (number of residents). Table 4-1. Risk indicators for the separate hazards and total risk due to flooding according to Scenario 1 EAAP % GMR EAAP % GMR Yearly Yearly Time % GDP EAAL (USD) (WL> Pop. (WL> Pop. schools health units Horizon Loss* 10cm) affected* 1m) affected* affected affected COASTAL HAZARD 2020 1,085,000 0.03% 14,000 1.44% 3,000 0.31% 11 1 2050 2,473,000 0.08% 31,000 3.50% 5,000 0.52% 27 1 (RCP 4.5) 2050 2,781,000 0.09% 33,000 11.54% 5,000 5.05% 30 1 (RCP 8.5) PLUVIAL HAZARD 2020 20,354,000 0.63% 111,000 11.43% 49,000 5.05% 70 4 2050 24,876,000 0.76% 127,000 13.18% 59,000 6.18% 81 5 (RCP 4.5) 2050 35,795,000 1.10% 160,000 16.58% 84,000 8.65% 102 7 (RCP 8.5) FLUVIAL HAZARD 2020 1,405,000 0.04% 15,000 1.55% 2,000 0.21% 17 0 2050 2,932,000 0.09% 27,000 2.78% 5,000 0.62% 26 0 (RCP 4.5) 2050 2,712,000 0.08% 26,000 2.78% 5,000 0.52% 26 0 (RCP 8.5) COMBINED HAZARD 2020 22,844,000 0.70% 140,000 14.42% 54,000 5.57% 98 5 2050 30,281,000 0.93% 185,000 19.46% 69,000 7.32% 134 6 (RCP 4.5) 2050 41,288,000 1.26% 219,000 30.90% 94,000 14.22% 158 8 (RCP 8.5) * Compared to the latest population data (970,824) and latest GDP estimates (USD 3.264B). These results have been mapped and are shown in Section 4.3.2 showing both the spatial distribution of the Expected Annual Average Losses (EAAL) normalized for area (see Figure 4-8, Figure 4-9, Figure 4-10 and Figure 4-11), to provide an impression of where the largest impacts are concentrated. Given that pluvial is the largest mechanism and that it is distributed across the entire city, unsurprisingly the largest communities incur the largest losses, but on a USD/year/m2 basis the impact is surprisingly even across the city. The major exceptions for pluvial risk are the hotspot in Central Monrovia where higher value buildings are impacted by flooding between Benson Street and United Nations Drive and along the Monrovia Sewer Drain, New Georgia Road community next to Stockton Creek appears also as a hotspot, as does Clara Town, Logan Town and Duala Town-Mombo East which both have isolated low-lying swampy areas. As is the area on the boundary of Barnesville, Gardensville and Paynesville which borders a large swampy depression. For coastal risk, New Kru Town is a notable hotspot in the maps and Topoe Village and Matadi which experiences tidal flooding within the Mesurado estuary. Fluvial risk is concentrated around the Stockton Creek area, with the higher return periods impacting further into these communities. 33 of 69 Flood risk profile for Greater Monrovia March 2021, Final Scenario 2 – Socio-economic growth and no Climate Change For this scenario, the hazard was kept constant (2020 flood map for each return period) and the exposure value was changed according to a factor established by the SSP3 and SSP5 (in terms of increasing the maximum damage value of assets and the total number of people). The risk metrics in this scenario are shown in Table 4-2. • SSP3 factor to account for increase in value in assets due to increase of GDP: 4.13 • SSP5 factor to account for increase in value in assets due to increase of GDP: 13.31 • SSP3 factor to account for increase in population: 1.98 • SSP5 factor to account for increase in population: 1.87 Table 4-2 Risk indicators for the separate hazards and total risk due to flooding according to Scenario 2 % GMR % GMR Time % GDP EAAP EAAP EAAL (USD) Pop. Pop. Horizon Loss* (WL> 10cm) affected* (WL> 1m) affected* COASTAL HAZARD 2020 1,085,000 0.03% 14,000 1.47% 3,000 0.29% 2050 4,482,000 0.04% 61,000 3.17% 9,900 0.47% (SSP 3) 2050 14,444,000 0.04% 62,000 3.40% 9,00 0.51% (SSP 5) PLUVIAL HAZARD 2020 20,4354,000 0.63% 111,000 11.48% 49,000 5.00% 2050 84,064,000 0.66% 252,000 13.12% 118,000 6.12% (SSP 3) 2050 270,918,000 0.66% 300,000 16.46% 157,000 8.61% (SSP 5) FLUVIAL HAZARD 2020 1,405,000 0.04% 15,000 1.50% 2,000 0.26% 2050 5,801,000 0.05% 53,000 2.76% 10,000 0.54% (SSP 3) 2050 18,694,000 0.05% 49,000 2.69% 9,000 0.51% (SSP 5) COMBINED HAZARD 2020 22,844,000 0.70% 140,000 14.45% 54,000 5.55% 2050 94,347,000 0.75% 366,000 19.05% 137,000 7.13% (SSP 3) 2050 304,056,000 0.75% 411,000 22.54% 175,000 9.63% (SSP 5) * Compared to the 2050 population projection (SSP3 = 1,923,000 / SSP5 = 1,821,000) and latest GDP estimates (SSP3 = USD 12.7 B / SSP5 = USD 40.9 B). 34 of 69 Flood risk profile for Greater Monrovia March 2021, Final Scenario 3 – Socio-economic growth and Climate Change For this scenario (Table 4-3), the hazard increased with the impact of climate change (2020 and 2050 flood map for each return period and climate scenario (RCP 4.5 and RCP 8.5)) and the exposure value was changed according to a factor established by the SSP3 and SSP 5 (in terms of increasing the maximum damage value of assets and the total number of people). Table 4-3 Risk indicators for the separate hazards and total risk due to flooding according to Scenario 3 EAAP % GMR % GMR % GDP EAAP Time Horizon EAAL (USD) Loss* (WL> Pop. (WL> 1m) Pop. 10cm) affected* affected* COASTAL HAZARD 2020 1,085,000 0.03% 14,000 1.47% 3,000 0.29% 2050 10,212,000 0.08% 61,000 3.17% 9,000 0.47% (RCP 4.5 -SSP3) 2050 32,909,000 0.08% 58,000 3.16% 8,000 0.46% (RCP 4.5 -SSP5) 2050 11,486,000 0.09% 66,000 3.41% 10,000 0.51% (RCP 8.5 -SSP3) 2050 37,015,000 0.09% 62,000 3.40% 9,000 0.51% (RCP 8.5 -SSP5) PLUVIAL HAZARD 2020 20,354,000 0.62% 111,000 11.48% 49,000 5.00% 2050 102,738,000 0.81% 252,000 13.12% 118,000 6.12% (RCP 4.5 -SSP3) 2050 331,099,000 0.81% 238,000 13.09% 111,000 6.10% (RCP 4.5 -SSP5) 2050 147,833,000 1.16% 317,000 16.50% 166,000 8.64% (RCP 8.5 -SSP3) 2050 476,431,000 1.16% 300,000 16.46% 157,000 8.61% (RCP 8.5 -SSP5) FLUVIAL HAZARD 2020 1,405,000 0.04% 15,000 1.50% 2,000 0.26% 2050 12,110,000 0.10% 53,000 2.76% 10,000 0.54% (RCP 4.5 -SSP3) 2050 39,027,000 0.10% 50,000 2.75% 10,000 0.54% (RCP 4.5 -SSP5) 2050 11,202,000 0.09% 52,000 2.69% 10,000 0.51% (RCP 8.5 -SSP3) 2050 36,101,000 0.09% 49,000 2.69% 9,000 0.51% (RCP 8.5 -SSP5) COMBINED HAZARD 2020 22,844,000 0.70% 140,000 14.45% 54,000 5.55% 2050 125,060,000 0.98% 366,000 19.05% 137,000 7.13% (RCP 4.5 -SSP3) 2050 403,035,000 0.99% 346,000 19.00% 129,000 7.11% (RCP 4.5 -SSP5) 2050 170,521,000 1.34% 435,000 22.60% 186,000 9.65% (RCP 8.5 -SSP3) 2050 549,547,000 1.34% 411,000 22.54% 175,000 9.63% (RCP 8.5 -SSP5) * Compared to the 2050 population projection (SSP3 = 1,923,000 / SSP5 = 1,821,000) and latest GDP estimates (SSP3 = USD 12.7 B / SSP5 = USD 40.9 B). 4.3.2 Risk maps The maps below show the spatial pattern of risk as Expected Annual Average Losses normalised for area (USD/Yr/m2). They are also show the other impact indicators such as number of school and medical facilities affected, kms of road affected and population affected. 35 of 69 Flood risk profile for Greater Monrovia March 2021, Final Figure 4-8. Combined Risk with current climate, moderate and future climate change (normalised for area). 36 of 69 Flood risk profile for Greater Monrovia March 2021, Final Figure 4-9. Pluvial Flood Risk with current climate, moderate and future climate change (normalised for area). 37 of 69 Flood risk profile for Greater Monrovia March 2021, Final Figure 4-10. Fluvial Flood Risk with current climate, moderate and future climate change (normalised for area). 38 of 69 Flood risk profile for Greater Monrovia March 2021, Final Figure 4-11. Coastal Flood Risk with current climate, moderate and future climate change (normalised for area). 39 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5 Potential high-level risk reduction measures Several potential high-level flood risk reduction measures were discussed and elaborated with stakeholders. They are: A. Physical coastal defence structures B. Physical riverine defence structures. C. City zoning and land use planning D. Physical urban blue-green-grey drainage structures. E. Living with water (Urban design) F. Flood Early Warning System G. Establishment of a Greater Monrovia City flood risk management committee Whilst Measures A and B will reduce coastal and riverine flood hazard directly, they do not address the more significant pluvial hazard. Moreover, they are long-term solutions that require extensive technical, environmental, social and economic studies before financing can be sought. Measures C, D, E and F are shorter-term solutions which will more quickly reduce exposure and vulnerability for the most frequent events. Measures D, E and F are relevant for the pluvial hazard which is the most significant risk. The ‘Greater Monrovia City flood risk management committee’ is intended to better coordinate flood risk management and is in response to stakeholders highlighting the need for greater coordination and clearer demarcation of roles and responsibilities. The measures described in the following sections are based on several principles, accounting for the present conditions. They are: 1. Integration: Although presented separately, the measures are strongly co-dependent with some measures only working effectively if key conditions are in place (such as city zoning for flood risk). This reflects the complex, interwoven nature of city living and the choices residents make for housing, schooling, services and livelihoods. 2. No build zoning: That any measure must not create new risk in existing ‘no build zones’. These are especially important given the national obligation to protect the Mesurado wetlands, as a Ramsar site. Stakeholders have already stressed that people living there have done so illegally, though enforcement of regulations remains a challenge. 3. Community-led: That where possible, community-led risk reduction solutions should be explored given that central government has limited resources. These have the benefit of being steered by the communities affected, with seed money, training and community mobilisation being externally supplied. 4. Gender: It is also important to recognise the disparity in impacts that flooding causes on women and men. Therefore, the planning and implementation of measures must account for these differences, and fairly benefit everyone. The involvement of both men and women in consultations, planning and implementation of measures is a pre-condition. 5. Short term solutions: That measures that reduce risk in the short term, are prioritised given the high frequency of flooding (in some cases multiple times per year) and the potential challenges of longer term measures (physical river and coastal defences). 6. Mesurado ecosystem: That no measures are considered that would permanently or irreparably change the Mesurado ecosystem. This would mean maintaining the salinity gradient, by not imposing permanent restrictions on the inflow of sea water or freshwater. 7. Subsidence: That any measure must not initiate land subsidence, especially in the settled swamp area around Stockton Creek and the Mesurado. Any polderisation and pumped drainage of these areas must be avoided to prevent the organic soil from drying out and shrinking. Such a polderisation scheme, would also encourage further settlement, increasing exposure and increasing risk whilst operation and maintenance costs escalate. 40 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.1 Physical coastal defence structures Coastal flood defences constitute physical structures to reduce the chance of waves overtopping the beach and flooding the areas immediately behind the beach. The UNDP and Ministry of Environment commissioned a coastal risk assessment and feasibility study, which for Monrovia focused on the direct coastal flooding at West Point and New Kru Town, defining different options and costings27. The GoL has concluded on these options for implementation. The designed measures are stone revetments on the beach (Figure 5-1) which have the important effect of preventing further erosion of the coastline. West Point – Revetment New Kru Town – Revetment Figure 5-1. Artists impression of the New Kru Town and West Point coastal protection works under the UNDP/EPA project. —————————————— 27 UNDP and Liberia Environmental Protection Agency (2019). 41 of 69 Flood risk profile for Greater Monrovia March 2021, Final Recommendations for next steps: 1. The technical studies and assessments for the coastal defences have already been completed for Monrovia, therefore this measure is already in an advanced stage. 2. It is understood that the Government of Liberia has selected the most appropriate options for detailed engineering design and construction. Whilst also seeking confirmed financing for the works. 3. The only additional recommendation here is, as described in Section 6.2, to commission a coastal, riverine and estuarine bathymetric survey as part of those engineering works to better understand the coastal processes and optimise the designs as well as to commission both tidal and wave monitoring to improve understanding of the coastal flood hazard. The coastal measures for reducing risk in Greater Monrovia are already well worked out, but less so for the other measures. Therefore, cases from other countries and cities are presented as examples of how they may work. 42 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.2 Physical riverine defence structures Physical riverine defence structures may be necessary to reduce the fluvial flood hazard created by the monsoon flows of the St Paul River. The potential locations for the interventions are relatively easy to identify, being the Stockton Creek and area near the Bong Railway (which floods in the most extreme cases) in the North of Monrovia (Figure 5-2). Potential measures may include river embankments and an operational sluice to keep the monsoon high-flows in the St Paul channel. Additionally, there may be other options to reduce the fluvial flood hazard. These might involve temporary upstream storage in a series of reservoirs in the upper catchment, which might be more cost-effective and provide dry season flows for hydropower generation. It may also be possible to store monsoon river water in the flood plain above the Mt Coffee Hydropower reservoir. Given the modelling limitations in terms of the St Paul bed level and validation data, it is difficult to conclude on these riverine interventions and moreover would require several more years of feasibility studies before it might be implementable. It will also be an expensive option both to design and build with a substantial operational and maintenance burden for the Government of Liberia. Therefore the options will require considerable thought, not only in terms of the technical solutions but also in regards to roles and responsibilities of the different government institutions who have a vested interest in the management of the St Paul River. Figure 5-2. Flooding from the St Paul River (T1 – Left and T50 – Right) with the entrance to the Stockton Creek and the Bong Railway circled. The key government agencies in further elaborating this measure may be: • Ministry of Public Works • Ministry of Mines and Energy • Environment Protection Agency • National Disaster Management Agency • Liberia Hydrological Service 43 of 69 Flood risk profile for Greater Monrovia March 2021, Final Recommendations for next steps: 1. A bathymetric survey of the Mesurado estuary, Stockton Creek and St Paul River, which will probably take at least 1 year to commission and carry out (and is perhaps best done with a coastal bathymetric survey for efficiency). 2. Installation of water level gauges to monitor river levels on the: • Lower St Paul River (near upstream of the Stockton Creek confluence) • The Stockton Creek (Probably the Somalia Drive road bridge) • The Mesurado Estuary (Probably on the eastern side of Providence Island) Carry out river flow measurements at the gauge locations to understand the discharge distribution between the branches and create rating curves to generate discharge records. And a sea level gauge to monitor sea levels, on the: • Freeport breakwater/dock (to monitor tides/surge) • Offshore buoy (to monitor sea level) Two rainy seasons of data would be needed for the purposes described below. The design, installation and operation of the monitoring system will probably take two years before it provides data from two rainy seasons. 3. Adjust the existing flood hazard model to account for the actual bed levels of the Mesurado, Stockton Creek and St Paul River. Adjust the Digital Terrain Model for actual measured Mean Sea Level. Revalidate the model for the measured water levels and discharges to confirm the fluvial flood hazard extent. Commissioning and completion of this would probably take 6 months. 4. If confirmed, commission an options analysis and feasibility study for measures to reduce the river flood extent from the St Paul River. Commissioning and completion of this would probably take an additional 18 to 24 months. With some activities being in parallel, it would probably take 3.5 to 4 years to complete the studies on the river flood hazard reduction measures. Case study Whilst is it difficult at this stage to point to appropriate solutions for the fluvial flooding in Monrovia, it is informative to consider riverine flood control measures used elsewhere. These predominantly involve separating the river water from the exposed population, but there are many ways to do so. They are mainly in two forms, physical barriers preventing water from spreading out of the flood plain or entering urban river or canal systems; and upstream solutions that reduce the concentration of river water at a downstream point (usually the city). The physical separation of river water from the exposed population has been the approach of the Netherlands for a long-time using embankments, sluices and canals. This approach was also used in Jakarta, Indonesia, where a system of upstream reservoirs and urban canals with control gates reduce the rainy season peak water levels in the city. The most recent iteration of this approach is the ‘Room-for-the-River’ concept28 piloted in the Netherlands. This concept gives more space to the rivers reducing water levels at vulnerable locations. The measures can include removing road bridge embankments, moving flood embankments back, lowering the flood plain, raising the river banks, creating off-stream temporary storage or raising flood embankments (Figure 5-3). Figure 5-3. Some measures from the Room for the River approach used in the Netherlands. —————————————— 28 https://www.rijkswaterstaat.nl/english/water/water-safety/room-for-the-rivers/index.aspx 44 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.3 City zoning and land use planning During the stakeholder consultations, government officials repeatedly stated that residents were putting themselves at risk by settling in areas that frequently flood due to a lack of awareness. City land use zoning would, in part, address this awareness, but go further to manage the exposure of assets. Land use zoning can be used to identify, incentivize and encourage the densification of low to no-risk areas within the city. These low to no-risk areas can be where high-density living and critical infrastructure and services are focused. This ‘densification’ of the safer parts of the city, also called ‘urban consolidation’, will require range of measures to realise the new urban plans, these can include tools to capture low risk land value, incentive schemes to build higher density housing, taxes on unused/underused land in safer zones, building regulations requiring multi-storey housing and ‘No build zones’ in higher risk areas. City zoning was established by the 1958 Zoning Act, but there is no associated map showing where the zoning is and it is understood that enforcement is weak, with multiple organisations issuing building permits. Since, 2016 the Liberia Land Authority is mandated to create land use zoning maps and including cadastral services to register land and property29. So the legislative framework for the creation of city zoning for development is much clearer now, the stakeholders expressed the opinion that they would need considerable help in creating the zoning designations and maps. Also they stated that raising awareness was needed amongst the city authorities and residents about the zoning and the need for permits. And finally, an agency (possibly LLA or city authorities) would need the power to prosecute and enforce the zoning. An example of how zoning might work is given below. The flood hazard maps for a given return period are overlain on the city, and zones demarcated corresponding to ‘very low risk – high density’, ‘low risk – medium density’, ‘high-risk – low density/adapted living’ and ‘No build zones – temporary stormwater storage’ zones (Figure 5-4). The zoning would be based on an ‘acceptable level of risk’, and therefore mapped around a given probability of occurrence (i.e. a specific return period), in the example given it corresponds to a 1-in-5 year flood event (or a 20% chance of occurring in any given year). This is not a scientifically set protection level but often a combination of political and economic considerations. It would also take into account the current situation but also future climate change. The government would need to strike a balance between mitigating risks by enforcing no new build zones and also recognising the existing land tenure realities and citizens’ rights in higher risk areas, where residents have already settled. Zone 1: Zone 2: Zone 3: Zone 4: High density Medium Low density No build area living and density living living areas of used for work space in and work regular stormwater flood free space in areas periodic storage due to areas. of infrequent flooding, high high short duration levels of frequency of flooding. adapted flooding. living. 1 in 5 year flood Extreme climate change Moderate climate change Current Figure 5-4. Example of how flood risk zoning might look. —————————————— 29 Republic of Liberia (2016): Liberia Land Authority Act. 45 of 69 Flood risk profile for Greater Monrovia March 2021, Final The World Bank urban review of Greater Monrovia (2020) highlights that there is potential for densification of the lower risk areas and more efficient use of land. The report states that only 37% of Central Monrovia is built on (buildings or road/sidewalks) and 77% of that prime low risk real estate are single storey buildings. Part of the land and building stock is government owned and under-utilized. This presents an opportunity to develop unused or under-utilized land to create additional housing. Such city zoning works well considering the coastal and fluvial hazard, but the discontinuous nature of the pluvial hazard indicates that land use zoning can only work if the pluvial hazard is significantly reduced. Hence there is a strong dependency with the drainage measures. Recommendations for next steps: 1. A specific institutional assessment for the Liberia Land Authority and associated agencies to assess their needs to create and implement city zoning for Greater Monrovia. 2. Commission a project in support of the Liberia Land Authority, LISGIS and the city authorities to update the open-data of the city. This would effectively be a continuation of the work of the Open Cities Africa activities that mapped Clara Town and Doe Community in 2018. A good example of how this might be implemented is the Ramani Huria project30 in Dar es Salaam, where local NGOs and recruited and trained students to carry out the mapping. Massively enriching the mapping of city neighbourhoods, facilities, assets and population. Such a mapping activity would provide much greater context to the zoning at the community level than is currently possible. Additionally, an open spatial data platform could be incrementally developed that could start to be used as a decision support tool for resilient urban planning. 3. Commission a project to support the Liberia Land Authority and associated agencies to assess their needs to create, implement and enforce city zoning for Greater Monrovia. Parts 1 and 2 are probably achievable within 12 months, whilst the support to the creation, implementation and enforcement of the city zoning may take many years to fully implement. The institutional challenges to enforcement should not be under-estimated. Case study Flood zoning is a commonly used approach in developed countries to mitigate risk. It is often used to regulate development, inform insurance policies and reduce exposure and vulnerability, whilst accepting some level of risk is inevitable. Two examples are the zoning systems in the UK, France and the USA. In the UK, the Environment Agency defines three zones: Zone 1 has a less than 0.1% chance of flooding in any given year, where there are few restrictions on flood related land development; Zone 2 with between 0.1% – 1% chance of flooding from rivers or between 0.1% – 0.5% chance of flooding from the sea where for any land development a flood risk assessment (describing planned measures to reduce vulnerability) is needed before a building permit is granted; Zone 3 has a 1% or greater chance of flooding from rivers or 0.5% or greater chance from the sea, which has the strictest restrictions on development. The French zoning system31 is more related to the natural disaster insurance scheme, which is implemented through risk prevention plans led by local government. These risk prevention plans can define compulsory or recommended property‐level measures to lower flood risk and limit new development in zones of high flood. The zoning system in the USA31, implemented through the Federal Emergency Management Agency is used to determine insurance premiums, insurance purchase requirements, and building codes that are part of the National Flood Insurance Program. Again, three broad flood risk zones are defined corresponding to annual flood occurrence probability. The —————————————— 30 https://ramanihuria.org/en/ 31 Hudson and Botzen (2019) 46 of 69 Flood risk profile for Greater Monrovia March 2021, Final regulations, for example, limit development in floodplains and require that newly constructed or substantially renovated homes must be elevated above the expected 1-in-100 year flood depth. 47 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.4 Blue-green-grey drainage measures Stormwater from high intensity rainfall events is a major problem in Monrovia. The Ministry of Public Works, who are responsible for both construction and maintenance of drainage, confirmed that stormwater drainage is currently inadequate in both the formal and informal parts of the city. With little investment since the 1980s. Therefore, to address the pluvial flooding and convey stormwater away from exposed assets, investment in drainage is necessary. Blue-green-grey drainage measures primarily look at integrated solutions to address the pluvial flooding. This approach works by enhancing and building on natural infiltration whilst removing rainwater from key locations and retaining it in areas of low exposure. The combination of blue-green-grey drainage measures32, looks at: • Blue solutions - open water areas for temporary stormwater storage • Green solutions - wetland or green vegetated areas to maintain soil infiltration and temporarily store surface stormwater • Grey solutions - traditional constructed drainage of adequate size and design to drain roads and residential areas, diverting stormwater to either blue or green areas. These measures can be relatively cost effective but still requires careful engineering studies. They can be a combination of several measures (Figure 5-5), with traditional concrete drains conveying water towards rehabilitated/protected urban wetlands, permeable paving to enhance infiltration or green-swales (vegetated strips that capture and store stormwater from roads and residential areas). Noting that this measure needs to be combined with the city zoning and other infrastructure planning. Figure 5-5. Examples of Blue-Green-Grey drainage infrastructure. —————————————— 32 This is also often referred to as Sustainable Urban Drainage Solutions or Nature Based Solutions, but they are all in practice reasonably similar. 48 of 69 Flood risk profile for Greater Monrovia March 2021, Final One of the issues faced is the current localisation of drainage, with many patches of low-lying swampy ground that are hydrologically isolated from either the sea or the inland water, meaning that water cannot drain quickly. Also if the river/tide is high there may not be sufficient gradient to convey the water away from these areas even if drains are installed. If we take the Logan Town area as an example (Figure 5-6), the Bong Railway embankment, Logan Town Road and United Nations Drive are raised above the residential areas preventing drainage to the Stockton Creek. Within that area, informal housing has developed around a low- lying wetland which becomes flooded during a rainstorm. Many houses, both around the wetland area and along the Logan Town Road and United Nations Drive are also flooded. The image to the right shows how Green-Blue-Grey drainage infrastructure might be integrated to reduce the risk to the residents. The Blue solution might be a new shallow detention pond with increased storage, the Green solution is the conservation and protection from development of the existing wetland depression (i.e. these correspond to the ‘No build zone’ in the zoning measure) and the Grey solution (coloured red here) are improved road side drains with oversized openings, correctly levelled with an appropriate outlet structure at the Stockton Creek. Figure 5-6. Pluvial flooding of Logan Town (1 in 5 year flood event) on the left and possible Blue-green-grey drainage measures on the right to improve drainage. The Logan Town provides a good example of a possible drainage ‘unit’ but there is also the possibility to link several of them optimise temporary storage capacity and provide a connection to the permanent water bodies such as the sea and Stockton Creek. This connectivity, shown in (Figure 5-7), of the city drainage, low-lying swamps and water bodies is an important consideration because the topography is relatively flat and water cannot drain quickly. Therefore, if the drainage system is well connected, the overall storage is increased for localised rainstorms, and thus potentially dampen the flooding. This would involve a correctly levelled and graded channel connecting several of these existing green areas with the water bodies, including the associated concrete culverts (grey infrastructure). The outfalls of these drainage channels (blue arrows) would need to have flap gates to prevent water flowing into the drainage system during high sea or river states. 49 of 69 Flood risk profile for Greater Monrovia March 2021, Final New Kru Town Logan Town Figure 5-7. Possible linking of the existing green wetland areas. The Ministry of Public Works stated a preference during the consultations for traditional closed stormwater drains, with the stormwater being directly drained to the sea. They felt that they cannot protect ‘green’ areas from unregulated development. This is an understandable concern, but several factors need to be also considered. These are: 1. Can traditional ‘grey’ stormwater drainage systems be built of sufficient size to cope with rain intensities experienced in Monrovia? 2. Will the continued loss of the existing green areas (presumably due to unregulated development), reduce soil infiltration and stormwater storage and thus increase flood hazards? 3. Will that increase in stormwater, due to loss of infiltration/temporary storage capacity, be more than the new grey stormwater drains can convey? It is difficult to answer these questions right now but given the extreme rain intensities it is hard to imagine an entirely grey underground drainage system large enough to cope. This is especially given the relatively gentle slopes and the high costs associated with pumped drainage. The loss of the existing green space will undoubtedly increase stormwater further. Therefore, maintaining these green areas (and their storage/infiltration capacity) remains advisable. Fundamentally this blue-green-grey drainage infrastructure should be viewed as a combined suite of drainage measures as no one solution can resolve the problem. 50 of 69 Flood risk profile for Greater Monrovia March 2021, Final Recommendations for next steps: 1. Following a review of this measure, a decision is needed whether to proceed with a city- wide drainage master plan or to initiate a pilot scheme for small micro-catchments where the approach can be quickly tested, before spreading out across the city. 2. Mapping of the existing drainage network across Greater Monrovia would greatly assist in planning and designing a drainage system for the city. This does not necessarily need to be a traditional topographic survey but might be based on the mapping format developed for Open Street Map by Ramani Huria in Dar es Salaam, which was based on relatively cheap Precise Point Positioning (PPP) equipment and smart phone equipped surveyors. 3. Regardless of approach chosen, a technical feasibility study is needed. Given the discontinuous nature of the pluvial hazard and the obvious micro-topography of the low- lying vulnerable areas of the city, the technical study of the drainage options needs to look at delineating drainage areas and therefore the target community within it. The target communities would need to be extensively consulted about the potential drainage measures. This should include active involvement in planning, but also in designing the construction process in such a way as to create low-skilled employment for those communities. Finally, the planned drainage interventions would need to be designed to an appropriate design storm (e.g. 1 in 1 or 1 in 2 year event) which represents a politically acceptable level of protection. We would argue that focusing on the most frequent, and thus smaller storm events is both economically feasible and common sense, as these frequently occurring storms already cause extensive flooding. By creating a drainage system that already prevents these frequent floods a large part of the hazard has been removed (noting for example that the 1yr rainfall event is 65% of the 10yr rainfall event), thus reducing the severity of the more extreme rainstorms. 4. Due care and consideration are needed for the ongoing operation and maintenance of the drainage system. Recognising the limited government resources, reducing operation and maintenance costs, must be a factor in the design criteria. Even so simple drains, culverts and flap gates still require maintenance to function as designed. Therefore, the mechanism for funding and delivering maintenance must be carefully prepared with the government stakeholders, including identify who the responsible agency is. Case study Although, the blue-green-grey approach is innovative it has been implemented in both developed and developing countries. An example are the urban wetlands in Colombo, Sri Lanka which have been supported by GFDRR33. The Metro Colombo Urban Development Project invested in Colombo’s public spaces, wetlands, and infrastructure to improve the flood protection for 232,000 residents and improving the quality of life of over 6 million people. Figure 5-8. Revitalised urban wetlands in Colombo, Sri Lanka. Revitalized public parks, spaces and wetlands enhance livability of the city whilst aiding in flood control. To achieve this the gravity drainage capacity was increased to 185 cubic meters per second; two micro-drainage projects implemented to provide protection to localized areas (10-year return period); 45km of roads and drainage rehabilitated; and the primary canals connecting wetlands were improved. —————————————— 33 https://www.gfdrr.org/en/feature-story/reducing-colombos-flood-risk 51 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.5 Living with water The ‘Living with Water’ measure considers a suite of measures, which focuses on simple low-cost community level measures to reduce exposure and vulnerability. The measure could include: 1. Building codes for more resilient structures above flood levels (assuming a certain protection level). The building codes would be linked to the city zoning and the building permits to better regulate urban development and the risk exposure. 2. Simple standard construction designs for reinforced concrete piles and wooden board walkways, as well as for raised building foundations. They should be simple and cost efficient enough that local firms can implement them, with an emphasis on exploring low- cost/innovative materials for the pathways and foundations (such as recycled material paving/planks) thus reducing dependence on external financing and making the solutions more scalable across the areas. This approach would also generate local unskilled/semi- skilled jobs within the communities. Seed money and training can be provided so that residents can form cooperatives or small construction companies. 3. Solid waste clear-up before the rainy season to clear drains and reduce localised flood hazard. This might be a community-led activity with a centralized waste pickup point for the municipal waste collection. 4. Improved water, sanitation and health facilities to reduce the risk of post-flood waterborne diseases. This can include either connections to a central sewer network and/or deep sealed pit latrines (or similar solutions) with the latrines themselves above the flood level and accessible by raised walkway. Again, this would not only make a significant impact on post-flood recovery and reduce mortality rates in the slums but also improve the living standards generally. 5. The creation of simple, locally appropriate community level flood response plans which would detail: a. Warning dissemination process b. Evacuation routes to safe areas. c. Simple evacuation procedures d. Preparedness measures that residents can take to reduce vulnerability and improve post-flood recovery. 6. Finally, a community-level awareness raising campaign as the basis for all other activities. Potentially this could be the most innovative, and community led of the solutions, but there are challenges. The low-level of education amongst residents, especially in the slum settlements must be taken into account. Summarising flood risk information into locally appropriate language for radio broadcasts and community engagement is very important. Although a gender-sensitive approach is encouraged when planning all measures, whilst for the living with water concept it has special relevance, as it requires very close consultation with the residents. Therefore, the consultations for gathering information on the problem and for discussing solutions, must account for the different roles and responsibilities that men and women have in their communities. This concerns how men and women are impacted differently by floods, how men and women’s roles in community decision making differ, and also how men and women view potential solutions differently. When planning the measures the pre-existing patriarchal structures of the slum communities in Monrovia34 must also be taken into account, which may discourage full female involvement in the participatory planning process. So pro-active capacity building and female-empowerment should not be neglected, recognising that adult women in Monrovia are significantly less likely to have completed primary education 35. —————————————— 34 Nilsson et al (2019) and USAID (2019) 35 World Bank (2020b). 52 of 69 Flood risk profile for Greater Monrovia March 2021, Final Recommendations for next steps: 1. The first, initial step must be a government, city-wide and community-level awareness raising campaign to simply state the flood hazard and risks. Providing a simple message about actions taken that increase exposure and vulnerability and those actions that reduce risk. 2. The second step would be for the central government agencies and the Monrovia and Paynesville City Corporations to identify a few key ‘at risk’ communities in need of both urban upgrading and flood risk reduction. These might also be the same communities identified for drainage upgrading. 3. The third step would be to initiate a community level project with both international and local support to consult with communities and identify the final suite of measures to improve flood resilience. This shall build upon existing social capital and encourage local pre-existing community groups (e.g. savings group, etc.) to take ownership of both improving awareness of risk and developing solutions through a participatory approach. Case study There are few examples of an integrated ‘living with water’ approach being implemented in an African slum community and it presents possibly the most innovative of the solutions described here. The concept comes from developed countries where traditional flood defence approaches are no longer economically viable, and the increased risk posed by climate change must be accepted and adapted to. The Netherlands36 is the often-referenced example of this ‘living with water’ approach, though its implementation is quite different to how it is envisaged for Monrovia. The approach in the Netherlands is heavily integrated with the concept of green-blue drainage infrastructure to capture and store excess rainwater (such as the approach described in Section 5.4) but also highlights other key aspects such as: adaptation to flooding being a shared responsibility of municipal authorities and residents, municipalities have discretionary funds to subsidies adaption measures or fund local community projects. The City of Rotterdam also initiated projects to increase resilience of urban communities through a series of measures. These included ‘flood-proof’ buildings to reduce damage in the event of pluvial flooding, construction of flood-proof public areas (that would flood but recover quickly) and floating communities (houses built on concrete platforms anchored to the ground that float up with rising flood waters) (Figure 5-9). Although such adaptive living structures in the Netherlands are quite sophisticated, the concept of floating or raised houses in low-lying deltaic cities is not new nor confined to developed countries. Figure 5-9. An example of floating houses (Left) in the Netherlands that rise with flood levels whilst remaining accessible and an example of raised houses (right) in the Amazon basin designed to be above the monsoonal flood levels with wet season transportation provided by household/community boats. —————————————— 36 Dai et al (2018) 53 of 69 Flood risk profile for Greater Monrovia March 2021, Final There are also examples of the more general concepts described above. The Jipange project 37, implemented by the World Bank in Dar es Salaam which helps local communities to develop disaster plans. Allowing residents to better prepare for flood events reducing the negative impact of floods on their livelihoods. It has a focus on effective communication with disaster plans developed using appropriate language, utilising cartoons (Figure 5-10) and social media. Furthermore, slum upgrading projects are already well known with clear guidelines on good practices with well-developed solutions to improved (cost appropriate) housing, drainage, sanitation and water supply, waste management and other urban services. Figure 5-10. Disaster Planning posters for Dar es Salaam communities from the Jipange Project. —————————————— 37 www.floodtags.com/jipange-project/ 54 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.6 Flood Early Warning and Response A Flood Early Warning and Response measure would provide residents and government agencies with forewarning of potential flood events, allowing them to take measures to reduce exposure and vulnerability. This can include implementation of community level flood response plans (evacuation of people and possessions) as described above. Such a measure would require a flood forecasting system to simulate near-future water levels in the city, which is coupled to a Warning System. This may be as simple as the forecasting agency notifying the relevant agencies directly (e.g. police, local government offices, television news, print/e-media and radio stations). It would require close collaboration between: • Liberia Hydrological Service (responsible for river monitoring) • Liberia Meteorological Service (responsible for weather monitoring and forecasting) • National Disaster Management Agency (responsible for flood response planning and coordination) • Emergency services such as Police (responsible for flood response, such as road closures and traffic diversion) • Local Government (responsible to flood preparedness and response at local level). • Media Although flood forecasting and warning systems are relatively complex, requiring servers with continuous backup, connected to rain and river gauges, they are a game changer for flood risk reduction. Allowing government and citizens to make informed timely decisions. The reliability of internet connections and power supply are a major limiting factor for such systems, as is the availability of qualified system administrators. As with any country responsibilities are spread across several agencies. So, the other challenge is which agency should host, operate and maintain the flood early warning system. Recommendations for next steps: 1. Commission a design, build and install consulting assignment to prepare the Flood Early Warning System. 2. The consultants would identify the stakeholder’s roles and responsibilities in a forecasting and warning system and identify system requirements and training needs. Identifying the responsible agency(ies), i.e. the organisations that will operate the system. This assessment shall also identify key factors in ensuring sustainability of the system. 3. The consultants would work with the stakeholders to map out the warning information chain from the river, sea and rainfall monitoring, through the forecast modelling process, information compilation and review, issuing a forecast/warning, warning dissemination by agencies and media, before the final action by government actors (e.g. police and NDMA) and/or ordinary residents/community groups. 4. The consultants would then build and install the forecasting system, including identifying the most appropriate, cost-efficient and robust hosting solution and the responsible agencies (bearing in mind local capacities). Creating the ICT architecture (hardware and software) to host the forecasting system (either locally or virtually). 5. The consultants would train the responsible agencies in the interpretation of observation data and forecasts. 6. The consultants would work with the responsible agency(ies) to test the robustness of the forecasting and warning system over a system testing period (rainy season) against pre- agreed criteria and for a realistic level of certainty. 7. The consultants create a detailed operation and maintenance manual, with training for the key actors in the flood warning chain, with a role-playing exercise of a flood case study which all actors participate in. 55 of 69 Flood risk profile for Greater Monrovia March 2021, Final The commissioning, implementation and testing of such a system would probably take between 12 and 18 months to achieve. Assuming that a river and rain gauge network is in place before the work starts. Case study Although, these present a major challenge in the context of Liberia, such appropriate systems have been developed for other countries. A good example is the Community Water Watch pilot project38 for Dar es Salaam, which took advantage of an existing flood model and rain/river gauge network39 to provide short-term flood forecasts for the city (Figure 5-11). To reduce reliance on local monitoring and forecasting capacity there are also global forecast services which can be subscribed to, thus also reducing local hosting and ICT requirements. Figure 5-11. An example of a flood forecasting system for Dar es Salaam. In Dar es Salaam, the system is hosted by a private organisation in the Netherlands, thus avoiding issues of internet connectivity and institutional capacity. In the case of Monrovia, such a system could be hosted outside the country or by a private organisation in the city who would provide system hosting and administration services to maintain the servers and perform back-up. When combined with the community flood response plans developed under the Jipange Project described above, it is an effective tool. —————————————— 38 https://dashboard-dar.floodtags.com/ 39 Installed and operated by TAHMO on behalf of the World Bank / https://tahmo.org/ 56 of 69 Flood risk profile for Greater Monrovia March 2021, Final 5.7 Greater Monrovia City flood risk management committee One of the issues relayed by the stakeholders, is that responsibilities for flood risk management in Monrovia are spread across numerous organisations and that coordination and cooperation must improve. Therefore, a clear recommendation is to create a: “Greater Monrovia City flood risk committee” To enable this, it is important to make a full institutional assessment, not only looking at the organisations mandated tasks and responsibilities but looking deep into the organisational capacity, staffing levels, job descriptions and facilities. Identifying gaps and needs. Their mandate could be to: • Provide the President’s office and City Authority office with information and advice prior to, during and after flood events. • Share knowledge with each other on flood hazard and risk within Greater Monrovia. • Coordinate all flood risk reduction measures within Greater Monrovia to achieve an integrated flood risk reduction approach. • Ensure that city and community flood preparedness and response plans are created. • Issue flood warnings to government, media and citizens. Possible members of the Greater Monrovia City flood risk committee: 1. Ministry of State for Presidential Affairs (Chairperson) 2. National Disaster Management Agency 3. Ministry of Finance and Development Planning 4. Ministry of Public Works 5. Monrovia City Corporation 6. Paynesville City Corporation 7. Environmental Protection Agency 8. Liberia Land Authority 9. Liberia Meteorological Agency 10. Liberia Hydrological Service 11. Liberia Institute of Statistics and Geo-Information Services Placing the committee under the Ministry of State for Presidential Affairs would ensure that it has sufficient political support and is held accountable at the highest level. Recommendations for next steps, are to be taken as formative and require the input of government officials to work out the correct steps in accordance with the Liberian constitution: 1. A first step might be to carry out an institutional review of the different government agencies, identifying those with a key role in flood risk management. The participating organisations for this study are a good starting point (Deltares, 2019). 2. Lobby the Minister of State for Presidential Affairs regarding the flood risk profile of the city and the concept of creating a multi-agency committee to coordinate flood risk management. 3. Create the committee, with each agency or ministry nominating a focal point. 4. Initiate capacity building support to the committee members and contributing agencies and ministries. Case study Most countries struggle with the institutional challenges of scattered roles and responsibilities when it concerns flood risk management at the city level, with different solutions being applied. Rwanda is an interesting case in regard to Flood Risk Management for the City of Kigali, where a 57 of 69 Flood risk profile for Greater Monrovia March 2021, Final multi-agency ‘Task Force’ has been established to manage all aspects of flooding. It comprises the City of Kigali authority, Water Board, Ministry of Environment, Ministry of Public Works, Environmental Management Agency and Ministry in Charge of Emergency Management. They are tasked with coordinating flood management and mitigation measures in an integrated catchment approach, comprising wetlands restoration, green urban development, urban drainage (and road upgrading), urban planning and the relocation of exposed assets from flood prone locations. They are also tasked with improving the knowledge base on flood hazards and risk and coordinating donor funded urban development and flood risk reduction projects. It therefore presents a good example of a sub-Saharan country implementing a multi-agency approach to flood risk management, reflecting that it has been supported by capacity building. 58 of 69 Flood risk profile for Greater Monrovia March 2021, Final 6 Conclusions and recommendations 6.1 Conclusions Hazard From the most extreme separate and combined flood hazard model runs we can conclude that: 1. Pluvial flooding effects are spread across the whole city, with flood depths greater than 10cm found throughout the city even for very frequent rain storms. Deeper pluvial flooding is concentrated in the low-lying areas draining into the Mesurado river. We can assume this is due to the low gradient and thus slow drainage. 2. Fluvial and tidal flooding predominantly affects the area around Bushrod Island and particularly the Stockton Creek and Mesurado estuary areas. 3. Fluvial flooding causes inundation around Stockton Creek and Mesurado estuary areas, with the high monsoon driven St Paul River creating the high-water levels necessary to flood these areas to the south of the river itself. 4. We also note that in the most extreme return periods the St Paul River will flood the left bank area upstream of Stockton Creek, cutting the Bong Railway and main road to the hinterland. This flooding becomes worse with climate change. 5. The flood model shows high sensitivity to the river discharge values. Inundation depths dramatically increase between the low (average monsoon discharge) and high conditions (1:10 yr event) considered (1100 m3/s vs 3500 m3/s respectively). 6. Given climate change induced sea level rise and increased rainfall, all flood hazards become worse for the 2050 time horizon. Notably for the fluvial hazard, the moderate scenario results in worse flooding than the extreme scenario, although the difference is small. This is due to the relative affect of air temperature change applied as long-term monthly averages over the large St Paul Basin. For the occurrence of separate and combined flooding mechanisms we can further conclude: 1. Since the coastal surge amplitude is small, the timing of the surge relative to the tide does not seem to be a critical factor. 2. Based on 24-hrs and 7-day rain events, the model shows that short, intense rainfall events lead to significantly higher inundation, this is attributed to the role of infiltration in damping the surface water accumulation over longer periods 3. The timing of the peak local rainfall and peak coastal water levels results in small differences. When peak timings coincide, the total water depth at that moment feels the maximum effects of both signals (peak in runoff due to rainfall and peak in backwater due to high tide). When the peak timings are out of sync the affect is lower. 4. Important to recognise that the St Paul River experiences high water levels for long periods, whilst coastal and pluvial peak values are much shorter. Therefore, high river discharges are constantly filling up the inundated areas to their maximum, no matter the rainfall infiltration or tidal state. Risk For the risk modelling we can further conclude: 1. Although the more extreme fluvial and coastal return periods cause floods of greatest depth, it is the pluvial flooding that creates the highest risk in terms of damage. This is due to the wider area that it covers, as seen in the pluvial risk maps. The pluvial risk is more than 10 times higher than for the respective coastal and fluvial risk. 2. As the coastal risk does not increase significantly with more extreme events the relative damage does not increase much, but the number of residents affects becomes much more significant as the inland tidal flooding is exacerbated by sea level rise in 2050. 59 of 69 Flood risk profile for Greater Monrovia March 2021, Final Although the flood depths remain shallow. Some 1.5% of the current population being affected by fluvial flooding of 10cm and above or 0.3% by flooding of 1m or more. This rises for 2050 to about 3% of the population (>10cm) and 0.5% (>1m), with only small differences for climate change scenarios and population growth projection. 3. Fluvial risk does increase significantly as the return values increase. Most of the frequent low magnitude events are confined to the Stockton creek area. Around the T10 point and beyond a topographical factor comes into play with the inland low-lying areas becoming inundated. The more severe flood depths also result in significantly more residents being affected, with higher levels of damage. The effect of climate change also doubles the impact of fluvial flooding and interestingly the more moderate climate change scenario results in slightly lower risk than the extreme scenario. Some 1.5% of the current population being affected by fluvial flooding of 10cm and above or 0.2% by flooding of 1m or more. This rises for 2050 to about 3% of the population (>10cm) and 0.5% (>1m) with only small differences for climate change scenarios and population growth projection. 4. Pluvial risk is by far the most significant with the largest proportion of residents affected, impacting a larger area. With a more significant proportion of GDP and Greater Monrovia Population being affected. Some 12% of the current population being affected by pluvial flooding of 10cm and above or 5% by flooding of 1m or more. This rises for 2050 to between 13 and 16% of the population (>10cm) and 6 and 9% (>1m) depending on the population growth projection. 5. In monetary terms, both coastal and fluvial risk are similar at USD1.1M/year and USD 1.4M/year in losses respectively, whereas pluvial risk estimates are USD 20.4M/year. Pluvial risk is therefore also accounting for nearly 90% of the combined risk per year. With the increase in value of assets because of economic growth, the risk is expected to grow in monetary terms, with the coastal and fluvial losses increasing from around 0.03% to potentially 0.09% of GDP whilst pluvial losses will increase from around 0.63% to potentially 1.16% of GDP (depending on the socio-economic and climate scenario). 6. Although the monetary estimates included in this assessment are based on the best available data, the values are uncertain due to the lack of primary data on actual damage costs, but that the order of magnitude is probably a reasonable approximation. Therefore, they can be used to compare areas within Greater Monrovia and therefore ascertain which areas are high/low risk. Thus, helping inform what investment levels might be realistic for cost-efficient solutions. In terms of non-monetized impacts we should recognise that: • Approximately 2/3 of inhabitants of Monrovia live in ‘at risk’ slums in the lowlands and swamps along Stockton Creek and Mesurado River: Topoe Village, Battery Factory, Doe Community, Logan Town, Clara Town, Matadi and West Point. Many residents of these vulnerable communities’ work in the same (at-risk) area in trade, odd jobs (day wage workers), street vending (petty traders), and subsistence fishing, sand mining or employed in the port. In case of flooding, this means that not only dwellings are flooded, but also the majority of roads (which are often petty traders workplaces), effectively bringing a halt to economic activities for the duration of the floods; schools are closed as well. • The proximity of these (high-risk) settlements to Downtown and the Freeport area, is a major attraction for migrants from the hinterland. These location dependent livelihoods include petty traders, market traders, port workers and fisherman. Additionally, the World Bank’s Urban Review (2020) highlighted that the two thirds of households in the bottom 40% of wealth distribution are either tenants or living rent free. With areas closer to the central business district having a higher proportion of rental properties compared to further away, with the poor slum residents far more likely to live close to the central business district. This pattern highlights the importance of living close to the city centre for the poorest residents, making them more willing to locate in higher risk areas. Furthermore, it highlights an ill-functioning housing/land market. Stakeholders reported 60 of 69 Flood risk profile for Greater Monrovia March 2021, Final that there have been active measures to relocate people from vulnerable locations, although to date this has only focused on the erosion prone West Point. • The flood water is typically of low quality, which brings associated health risks. Some people go out during a flood to obtain food; the contact with the water brings risks. Also, everybody is susceptible to diseases such as malaria, diarrhoea and cholera, and risks of attracting these illnesses are aggravated by flooding. • Table 6-1 provides the number of basic services directly impacted by flooding, but impaired access to services due to flooding in the area is hard to quantify. The hazard maps, especially the pluvial hazard, show that the discontinuous nature of the flooding creates a fully disrupted city. • Health units are less affected compared to the schools. This may reflect the siting of health units in the more formal settlements which may be the older (low risk) parts of the city. Whereas schools, being sited near residential areas are scattered throughout the slum settlements in the city. • Pluvial has by far the largest impact, in terms of schools and health units affected, and the kilometers of road flooded. This fits with the stakeholder derived narrative as well as the annual news reports from Monrovia, which attributed flooding of facilities and roads to high intensity rainfall events, Table 6-1. Risk to basic services for Greater Monrovia. Time Horizon Affected schools per Affected health units Affected kms of road year per year per year COASTAL RISK 2020 11 1 2 2050 (RCP4.5) 27 1 28 2050 (RCP8.5) 30 1 34 FLUVIAL RISK 2020 17 0 30 2050 (RCP4.5) 26 0 55 2050 (RCP8.5) 26 0 52 PLUVIAL RISK 2020 70 4 121 2050 (RCP4.5) 81 5 157 2050 (RCP8.5) 102 7 242 The top three impacts are listed by residents during the field work were: • Inaccessible roads/transportation • Sickness outbreaks • Loss/damage to household property We should note that the first two are temporary impacts. The lack of reported flood impacts such as disruption of electrical, water or sewerage service probably reflects on the limited access to such services rather than the actual disruption. Noting that only 27% of residents report having access to electricity and 32% having access to solid waste collection. 61 of 69 Flood risk profile for Greater Monrovia March 2021, Final Identification of hotspots From the risk mapping we are able to identify some notably impacted locations (Figure 6-1). For pluvial risk, there is a hotspot in Central Monrovia where higher value buildings are impacted by flooding between Benson Street and United Nations Drive, New Georgia next to the Stockton Creek appears also as a hotspot, as does Clara Town and Duala Town-Mombo East which both have isolated low-lying swampy areas. The area on the boundary of Barnesville, Gardensville and Paynesville which borders a large swampy depression are also notably risk prone locations. For coastal risk, New Kru Town is notable hotspot as is Topoe Village and Matadi which incurr tidal flooding and pluvial flooding within the Mesurado estuary. Topoe Village, Matadi and potentially Chocolate City being vulnerable to coastal flooding after sea level rise in 2050. By contrast West Point does not present as a hotspot when only flooding related economic damages are considered, but it is identified as a hotspot due to the coastal erosion which is damaging houses in West Point. Fluvial risk is unsurprisingly concentrated around the Stockton Creek area, with the higher return periods impacting further into these communities. The hotspots were identified by considering the areas of higher risk that stand out in the risk maps (See Section 4.3.2) though no fixed threshold has been applied as the concept of what is deemed ‘significant’ is fluid and a matter for stakeholders and government representatives to confirm. The areas shown below are defined by the black boundaries that roughly follow main roads, waterways and community boundaries to define ‘hotspots’ and are named after the most prominent community name, but may include other communities within the defined area. Cardwell West New Kru Town / Community / Tweh Farm New Georgia Road Logan Town / Mombo East Battery Factory Jamaica Road Topoe Village Clara Town West Point Barnesville / Gardensille / Payneville Central boundary Monrovia area Flooding >10cm Fluvial Pluvial Matadi Coastal Permanent Surface water Figure 6-1. Identified flood hotspots – Note the boundaries do not delineate contiguous communities or drainage boundaries. 62 of 69 Flood risk profile for Greater Monrovia March 2021, Final Limitation of this study It must be recognised that the data and modelling chain, despite representing the best available, has some limitations. They can be summarised as follows: 1. The primary limitation is the lack of long reliable rainfall, river and coastal observations from which to derive the flood hazard input data. This is not uncommon in developing countries and modelled data is used to artificially create the necessary flood event return periods. Therefore, uncertainties exist to some extent in the model forcings and thus in the model results. 2. Although the Digital Terrain Model is a high spatial resolution (0.5m) the actual vertical accuracy is not known as there are no high-accuracy ground control points to compare it to. This leads to extra uncertainty in the model results. 3. A major consideration in the interpretation of the fluvial flooding is the role that the bed level (bathymetry) of the St Paul River has in driving flooding southwards along the Stockton Creek. We should caution that the river bathymetry for the St Paul and Stockton Creek are not known, and the modelled flooding may be sensitive to the bed level. If we are to make only one recommendation to update the flood hazard model, it would be to commission a full hydrographic survey of the St Paul, Stockton Creek and Mesurado before commissioning engineering feasibility studies. 4. Additionally, the lack of robust water level measurements for the St Paul River (below Mt Coffee dam), Stockton Creek and Mesurado Estuary, as well as the coast, effectively means the flood hazard model is unvalidated. This should always be taken into account when reviewing the hazard and risk results, but the relative hazard and impacts of the different mechanisms are probably reasonable. Thus providing sufficient information to inform the identification of potential high-level flood management measures. 5. The availability of comprehensive exposure and vulnerability data in the form of a national database of buildings, including classifications, valuations and materials, is lacking. Again, this is not uncommon in developing countries. 63 of 69 Flood risk profile for Greater Monrovia March 2021, Final 6.2 Recommendations Further model development The flood hazard model developed represents the best possible solution given the availability of data, there always remains the possibility to develop it further, especially if it is to support engineering and economic feasibility studies. Possible recommendations include: 1. Using measured water level data at multiple stations in St Paul, Stockton Creek and Mesurado estuary for at least two rainy seasons to quantitatively validate the hazard model. 2. Given the role that sea level plays in the flooding mechanism and the low-lying nature of the interior of Greater Monrovia, the vertical accuracy and reference to mean sea level of the Digital Terrain Model needs to be confirmed with the collection of Ground Control Points and confirmation of mean sea level. The collection of Ground Control Points could easily be carried out by a local survey team with survey quality GPS equipment. 3. A major consideration in the interpretation of the fluvial flooding is the role that the bed level (bathymetry) of the St Paul River has in driving flooding southwards along the Stockton Creek. It is therefore recommended to include a hydrographic survey of the St Paul, Stockton Creek and Mesurado before commissioning engineering feasibility studies for fluvial flood defences. Strategic questions regarding the potential high-level risk reduction measures Before concluding on which potential high-level risk reduction measures to proceed with, there are several strategic questions that the Government of Liberia may want to consider, in consultations with potential financiers. These questions are: 1. What protection level does the Government of Liberia want to achieve, and thus what level of risk is acceptable? Given that Monrovia faces quite regular rainfall induced flooding, we recommend prioritising short term measures that reduce the rain induced flood hazard and increase resilience in the most at risk slum communities. A pragmatic way to achieve this is by implementing the drainage measures described in this report. Even aiming at a drainage system designed to convey the T1 or T2 flood event (the most frequent events) would be a major factor in reducing damage and nuisance flooding for the city. Additionally, slum upgrading measures and other ‘living with water’ measures would reduce damages and improve post flood recovery in the short term. An initial short- term protection level of the 1-in-1 or 1-in-2 year flood, would achievable given the financial and institutional capacity in Monrovia. 2. Whilst short and medium term investments in innovative coping strategies are a must for the vulnerable population that resides in the low-lying flood prone slum settlements, there remains a question about the long-term vision for the urban population of Liberia. Given that Monrovia has had extensive under investment in infrastructure and urban services due to the civil war, and that large parts of the city are low-lying and flood prone, should longer term urban development plans include new satellite settlements away from the coast and Mesurado, possibly centred on the main arterial roads and airport? A pre- condition would be to create sufficient jobs, housing, schooling and facilities, in order to encourage residents to settle away from the higher-risk areas that currently provide work and services. 64 of 69 Flood risk profile for Greater Monrovia March 2021, Final The next steps The following key recommendations are made for further consideration by the Government of Liberia, assisted by International Financing Institutions. Potential high level risk reduction measures Without repeating the recommendations in Chapter 5, we would recommend the following: 1. The Government of Liberia, assisted by International Financing Institutions, review the flood hazard and risk assessment and discuss the potential measures identified here, taking into account the other ongoing and planned urban development work within Greater Monrovia. 2. Create a city-wide flood risk management committee, with representatives from relevant line ministries and city authorities, mandated by the President to coordinate, integrate and plan flood risk management measures. This would be the vehicle to discuss the potential measures. 3. Initiate a government, city and community wide awareness raising campaign as a basis for further consultations and to create the city-wide acceptance that action (both by the government and residents) is needed. 4. Prioritise short term measures that address the highest risk mechanism (direct rainfall flooding) and those measures that can quickly reduce hazard, exposure and vulnerability. 5. Review and pre-select medium and long-term measures which may be necessary to address future hazard and risk, especially to address climate change induced increases in hazard, but also takes into account a future city with the short-term measures already implemented. Also such medium and long-term measures may require extensive additional data collection and feasibility study which takes time to implement. It is also clear that the Government of Liberia would need significant support in further developing these the risk reduction measures. Further data collection As detailed elsewhere in this report, there are data, currently missing that would be necessary to further elaborate the risk reduction measures. Many of them are best collected or generated before commissioning detail feasibility studies. The data are: 1. A bathymetric survey of the Mesurado estuary, Stockton Creek and St Paul River (perhaps best done in conjunction with a coastal bathymetric survey for efficiency). 2. Installation of water level gauges to monitor river levels and sea levels 3. Mapping of the existing drainage network across Greater Monrovia would greatly assist in planning and designing a drainage system for the city. 4. Commission a project in support of the Liberia Land Authority, LISGIS and the city authorities to update the open data of the city. As mentioned in Section 5.3, this would effectively be a continuation of the data collection work of the Open Cities Africa activities that mapped Clara Town and Doe Community in 2018. The additional open data should at least include: drone imagery, building footprints, building classification, roads, drainage, utilities, services and sanitation facilities. 65 of 69 Flood risk profile for Greater Monrovia March 2021, Final 7 References Dai, L, Wörner R and van Rijswick, H.F. M. W. (2018): Rainproof cities in the Netherlands: approaches in Dutch water governance to climate-adaptive urban planning, International Journal of Water Resources Development, 34:4, 652-674, Deltares (2019): Flood Risk Profile for Greater Monrovia - Deliverable A: Inception Report. Deltares (2020): Flood Risk Profile for Greater Monrovia - Deliverable C: Risk Report. Freire, S., MacManus, K., Pesaresi, M., Doxsey-Whitfield, E. and Mills, J., 2016. Development of new open and free multi-temporal global population grids at 250 m resolution., in Proceedings of the 19th AGILE Conference on Geographic Information Science. GEBCO Compilation Group (2020) GEBCO 2020 Grid (doi:10.5285/a29c5465-b138-234d-e053- 6c86abc040b9) Hempel, S., Frieler, K., Warszawski, L., Schewe, J., and Piontek, F.: A trend-preserving bias correction – the ISI-MIP approach, Earth Syst. Dynam., 4, 219-236, https://doi.org/10.5194/esd-4- 219-2013, 2013. Hudson, P. and Botzen,. W.J.W (2019): Cost–benefit analysis of flood‐zoning policies: A review of current practice. WIRES Water. Volume 6, Issue 6 November/December 2019 Huizinga, J., Moel, H. de, Szewczyk, W. (2017). Global flood depth-damage functions. Methodology and the database with guidelines. EUR 28552 EN. doi: 10.2760/16510 ImageCat (2017): GFDRR Africa Disaster Risk Financing – Result Area 5. Exposure Development Replacement Cost Refinements To the Exposure data. IPCC, 2014: Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. IPCC, Geneva, Switzerland. Jevrejeva, S., Jackson, L. P., Riva, R. E. M., Grinsted, A. and Moore, J. C. (2017): Coastal sea level rise with warming above 2 °C, PNAS, 113(47), 13342–13347, doi:10.1073/pnas.1605312113, 2016. Lenderink, G., Mok, H. Y., Lee, T. C. and van Oldenborgh, G. J.: Scaling and trends of hourly precipitation extremes in two different climate zones – Hong Kong and the Netherlands, Hydrology and Earth System Sciences, 15(9), 3033–3041, doi:10.5194/hess-15-3033-2011, 2011. Liberia Institute of Statistics and Geo-Information Services (2009): Population and housing census 2008. Marconcini, M., Metz-Marconcini, A., Üreyen, S., Palacios-Lopez, D., Hanke, W., Bachofer, F., Zeidler, J., Esch, T., Gorelick, N., Kakarla, A., Strano, E., (2019): Outlining where humans live -- The World Settlement Footprint 2015. 66 of 69 Flood risk profile for Greater Monrovia March 2021, Final Muis, S., Apecechea, M. I., Dullaart, J., de Lima Rego, J., Madsen, K. S., Su, J., ... & Verlaan, M. (2020). A High-Resolution Global Dataset of Extreme Sea Levels, Tides, and Storm Surges, Including Future Projections. Frontiers in Marine Science, 7, 263. Nilsson, M., Howard, J., Konyon, F., Tucker, H. (2019): The challenge of unlearning: A study of gender norms and masculinities in Liberia Palacios-Lopez, D., Bachofer, F., Esch, T., Heldens, W., Hirner, A., Marconcini, M., Sorichetta, A., Zeidler, J., Kuenzer, C., Dech, S., Tatem, A. J. and Reinartz, P. (2019): New Perspectives for Mapping Global Population Distribution Using World Settlement Footprint Products, Sustainability, 11(21), 6056, doi:10.3390/su11216056, c. Pesaresi, M., Ehrlich, D., Ferri, S., Florczyk, A., Freire, S., Halkia, S., Julea, A. M., Kemper, T., Soille, P. and Syrris, V. (2016): Operating procedure for the production of the Global Human Settlement Layer from Landsat data of the epochs 1975, 1990, 2000, and 2014., 2016. Republic of Liberia (1958): Zoning Act. Republic of Liberia (2016): An Act to amend title 12, executive law of the Liberian codes of law revised and to add thereto a new chapter creating the Liberia Land Authority (Known as the ‘Liberia Land Authority Act’). Slager, K., A. Burzel, E. Bos, K. de Bruijn, D. Wagenaar, H. Winsemius, L. Bouwer, M. van der Doef. (2016). User Manual Delft-FIAT version 1, website: publicwiki.deltares.nl/display/DFIAT/Delft-FIAT+Home) UNDP and Liberia Environmental Protection Agency (2019): Provision of Consultancy Services for Engineering and Safeguards Assessment of Proposed Interventions in the coastal areas of the Monrovia Metropolitan Area (MMA) in Liberia – Preliminary Modelling Report. By CDR UNECE, (2019). Shared Socioeconomic Pathways (SSPs) https://www.unece.org/fileadmin/DAM/energy/se/pdfs/CSE/PATHWAYS/2019/ws_Consult_14_15. May.2019/supp_doc/SSP2_Overview.pdf USAID (2019): Sexual and gender-based violence and trafficking in persons in Liberia. World Bank (2020a): COVID-19 Country Social Impact Note – Liberia World Bank (2020b): Greater Monrovia Urban Review: A Spatial Analysis investigating Constraints and Opportunities. Websites ERA5: https://www.ecmwf.int/en/forecasts/datasets/reanalysis-datasets/era5 MERIT: http://hydro.iis.u-tokyo.ac.jp/~yamadai/MERIT_DEM/ 67 of 69 Flood risk profile for Greater Monrovia March 2021, Final 68 of 69 Flood risk profile for Greater Monrovia March 2021, Final 69 of 69 Flood risk profile for Greater Monrovia March 2021, Final