Flood disaster response to rainfall dynamics and land cover change scenarios in the urbanized watershed of Accra, Ghana
Introduction: Globally, flood hazards are common in highly developed low-lying or floodplains. In India, for example, as many as 78 cities are subject to perennial flooding. Rapid urbanization has been identified as a major cause of regional hydrological change, as confirmed by Dan-Jumbo et al. in Nigeria.
Ouma and Tateishi, along with Areu-Rangel and others, have identified the impact of LULC changes and urban growth in upstream watersheds on flood hazards. Changes in LULC in forest watersheds that may increase or decrease peak runoff.
According to Solín and Panahi, among others, changes in land cover or a decrease in vegetation cover in river basins can increase the intensity and frequency of flood hazards.
Brownster et al. also showed that the combination of higher rainfall intensity and river catchment characteristics helped increase flow. Recent developments in the adoption of green infrastructure (GI) to mitigate urbanization and climate change-driven flood intensification have spurred the use of LULC and rainfall input variables to model and simulate floods for effective decision-making.
The Odo River Basin belongs to the five metropolises, municipalities and districts (MMDAs) of the Greater Accra region, which is the most urbanized, developed, and densely populated region in Ghana. A larger proportion of the catchment area has been built and some streams have been diverted.
The total population of the basin is about 4 million, and another 2 million migrants travel to the area every day to engage in different socio-economic activities. These MMDAs are centers of major activities such as industry, commerce, finance, culture, and transportation.
The area has many slums and flood-prone communities, made up of poor immigrants from other parts of the country and natives of the area.
The basin has the characteristics of both high and low, Gadong District has a high altitude, and AMA is a coastal low-lying area.
The vegetation of the watershed is dominated by woodland upstream and coastal shrubs and grasslands in low-lying areas. The basin has savanna climatic conditions (based on Corpen's climate classification) with annual rainfall of about 730 mm, short duration of rainfall, and strong storms leading to localized flooding.
The LULC in the Odau River Basin changed significantly from 2000 to 2020, with the largest change in coverage for built-up areas in 2000-2011 and 2011-2020.
Grass increased from 2000 to 2011, but decreased from 2011 to 2020. Bare land decreased the most from 2011 to 2020, while woodland gradually decreased throughout the study period. Bare land yields the highest gains, while woodland loses the most.
All changes between 2011 and 2020 indicate that all LULC types have been converted to built-up types, with bare ground decreasing the most, followed by grass.
Built-up areas were the only type of land cover that increased during this period. From 2000 to 2020, bare land cover lost the most.
The relationship between rainfall intensity, LULC and peak runoff within a watershed is clear. The highest runoff occurred in built-up areas, while woodland recorded the lowest runoff for all varying rainfall intensities and land cover.
Estimated peak flow per unit area for land cover types are BU: 0.0742; baseline: 0.045; GL: 0.0337; white balance: 0; white line: 0.0202.
Thus, the peak flow per unit area of BU was 3.67 times that of woodland, 2.20 times that of bare land, and 1.6 times that of grassland, indicating that replacing vegetation cover increased peak flow several times.
Over the years, the survey has found that the type of LULC in the Odo River Basin has undergone hidden changes, with an increase in the construction area and a decrease in the area of bare land, grassland, water bodies and woodland.
The largest expansion of built-up areas occurred in the last decade, between 2011 and 2020, and in their recent study of LULCC in Accra, Akubia and Bruns observed a similar increase in built-up coverage in Accra and attributed the increase to population growth and related housing demand for residential and non-residential uses.
Recently, cities in other countries have seen similar increases in built-up areas.
The study modeled the response of peak runoff and flow velocity to different scenarios of rainfall intensity of different durations and recurrence periods, as well as LULC dynamics in the urbanized Odaw River Basin in Accra.
Surveys in recent years have found that built-up areas have expanded, while vegetation areas, particularly woodland areas, have decreased significantly, with peak flow of LULC types increasing in all rainfall intensity and return periods during the period under consideration.
For heavy rainfall events with long recurrence periods, runoff peaks occur over a relatively short period of time; However, less intense rainfall with shorter return periods takes longer to reach peak runoff.
Conclusion: The results of the methodology used indicate that LULC conditions, rainfall intensity, and duration of different recurrence periods have clear potential effects on individual and synergistic effects of peak runoff and flow. Therefore, rapid warnings should be provided when long-term reproduction of extremely heavy rainfall is expected, and future studies must also consider the vulnerability and risk of flood hazards, which are important inputs to overall flood DRR planning.
