Nyarko, Benjamin Kofi 
ng The area is characterized by lowlands and occasional hills with an average altitude of 20 meters 
above sea level. The slopes are generally gentle with most slopes below 11 percent, except few 
places such as MaCarhty hills, the television transmitting station near Abokobi and Kwabenya 
| hills, where slopes are above 22 percent. The water table varies between 4.80 meters to 70 
d | meters below the surface at places like Ofankor, Kantamanso and Accra Brewery Limited 
bottling house in Accra. 
The area is drained through by natural streams (Figure 1.2) and valley network and artificial 
  
ly drains. Most of the streams like Odaw, Sakumo, Mahahuma, Lador, and Dzorwulu, take their 
source from the Akwapim range. The artificial drainage is mostly built-up structures that enable 
df quick discharge of waste and storm water. 
er The area falls within the anomalous dry equatorial climate region and experiences double 
nt maxima rainfall and a prolonged dry season with occasional dry harmattan condition being 
vs experienced. The hottest months are February and March, just before the rainy season, with 
te mean monthly temperatures of 27°C, whilst the coolest months are June-August. During this 
)0 period temperatures are around 21?C. Rainfall in this area has two peak periods in May to 
d August and October to November, with an annual amount between 780mm and 1200mm. 
y, There are two main vegetation types within the area; namely the coastal scrub and grasslands, 
es and mangrove forests. The coastal scrub and grasslands are in patches at certain places with 
id occasional trees such as Nim and Baobab. The mangrove forests are found in the coastal lagoon 
1 areas where the soil is waterlogged and salty. 
2 DATA SOURCE AND IDENTIFICATION 
nt Data to be used for this study were sought from two main sources namely, official and primary 
al sources. Data from official sources were obtained from published information on flood 
ja generation factors such as, rainfall, discharge of the various rivers and human activities within 
Be the study area compiled and used by specialized organizations such as Water Research Institute, 
ly Meteorological Services, Hydro-Division of Ministry of Works and Housing. The data collected 
d from primary source included vegetation characteristics, landuse pattern, channel characteristics 
er (cross section area) and soil characteristics. The methods used to collect these information 
included, field observations, field measurements and satellite/photographic and topographic map 
interpretations. 
n? To easily incorporate the data collected into the proposed geographic information system and 
it hydrological models for analysis, the Terrain Mapping Units (TMU) technique as described in 
ls. the works of Meijerink (1988), Moore et al (1991) Meijerink et al (1994) and Mitchell (1973) 
de was followed. Using 1997 aerial photographs with a scale of 1:10,000, Thematic Mapper 
satellite image of 1991 and the 1972 topographic map of scale of 1:50,000, the terrain of the 
study area was sieved using various cover classes into thematic maps (topography, elevation, 
slope, land use, and vegetation, drainage channels). Out of these data collected, total discharge 
was calculated using the modified rational model and a digital elevation model generated from 
1972 contour map of Accra of a scale of 1: 50,000 using ILWIS2.1 software. 
3 MODEL FOR THE STUDY 
The modified modeling flow based on the relational rule, was employed for the study as it 
concerns spatial extent of flood occurrence. A spatial model was adopted because it has the 
capability of using point data to represent an area in which spatial variability of specific 
parameters of an area can be integrated to help provide an understanding of interdependence in 
hydrology (Molenaar, 1998; Este, 1992; Doe III et al., 1996; and Baumgartner & Apfl, 1996). 
The application of a Geographic Information System Model (GISM) to study hydrological event 
in its spatial form is therefore appropriate, the reason being that it has the capabilities of 
incorporating physical and stochastic models for spatial analysis of a phenomenon. 
  
International Archives of Photogrammetry and Remote Sensing. Vol. XXXIII, Part B7. Amsterdam 2000. 1041