**5. Study methodology**

The present study employed both field data collection and secondary data sources to advance the researcher's opinion that due to the increased human activities and water quality degradation, the basin is likely going to experience severe water shortages and scarcity (hence affecting people's livelihoods).

Field and laboratory analysis and observations, standard empirical formulae for the determination of suspended sediment loads and discharge measurements were used to analyze the collected data. The used quantitative methods included principal component analysis, time series and chemical methods in determination of the water quality status.

Data on water quality and river discharge was necessary as well as the information on land use changes over time and space. This is relevant to the study since it is a factor associated with changes in river flow characteristics and quality degradation. Monthly data on river flow characteristics and water quality parameters were used, some of which were measured in the field during the period of investigation. Water sampling points indicated in **Figure 3** were used both for field samples and river gauging stations for secondary data.

The collected water samples were analyzed in the laboratory to quantify the sediment loads and the water quality deterioration changes along the river profiles. The total suspended sediments concentrations (TSSC) were determined on two replicate samples by gravimetric method according to Mcgrave [12] and Woodroffe [13]. The filtration was carried out by suction in which two vacuum flasks were run off one pump with open filter holders being mounted on each flask. After the determination of the water volume, sediment-water mixture was filtered through pre-weighed Whatman GF/C filters (4.7 cm wide, pore size 0.4 μm, thickness 260 μm) and kept in individually numbered aluminum packs. These were handled at the edges using flat-bladed tweezers. Following the filtration, the filtrates were washed with filtered distilled water and dried in oven at a temperature of 105°C for 24 hours. After removal from the oven, the filters were left to cool to room temperature for about 2 hours before they were reweighed using a sensitive electronic balance to the nearest 0.0001 g. Whatman GF/C filters of 47 mm diameter were used because they have a high flow rate and take a high sediment load [12]. Heavy metal analysis for mercury, chromium, cadmium, zinc, nickel, copper, iron,

**Figure 3.** *Water sampling points in the Nairobi River sub-basin (upper Athi catchment basin).*

manganese and fluoride was done to establish the concentration variations along the stream course distance-wise before and after the rivers pass through the city of Nairobi (urban/commercial areas) or part of its environs using the same method.

Water samples for water quality determination were collected in a depth integrated manner at the middle of the river. Each sample was taken according to the standard laboratory analysis and the concentration of each parameter determined. The basic method applied in the determination of each metallic cations was the Atomic absorption spectrometry (AAS). Measurement of other water quality parameters was done using standard laboratory methods and included BOD5, COD, TSS, pH, Total Alkalinity, Total dissolved solids, Conductivity, Calcium, Magnesium, Sodium, Potassium, Chloride, Fluoride, Total hardness and Turbidity.

BOD5 was determined over 5 days for biochemical oxidation of organic substances at 20°C. The detection limit was set at about 5 mg l<sup>1</sup> and the same was done for the chemical oxygen demand (COD). The total dissolved solids (TDS) concentration was done through evaporating the water sample on a previously weighed dish and the residue dried at 180°C and then weighed again to allow for the concentration to be determined while suspended solids (SS) were determined by filtration of a well mixed sample on a standard glass-fiber filter disk (0.45 mm filter; detection limit SS ≤5 mg l<sup>1</sup> ). Fluoride was determined by titration of the water sample with a standard Thorium nitrate in a solution buffered at about 2.9–3.3, using sodium alizarin sulphonate as indicator. Volumetric analysis or otherwise filtration method was used in the analysis of most of the other water quality parameters since it is usually more rapid than the gravimetric analysis if the filtrate reagent is specific for the water quality constituent. The sensitivity and/or precision may also exceed that of a gravimetric analysis for some determinants. It was widely used in the present study for the analysis of water quality parameters as pH, total hardness, and alkalinity according to Mancy [14]. Their results were used to group the parameters into three main categories; chemical, biological and physical related water pollution parameters. In addition, their magnitude values (concentrations) were used to infer on whether there was water quality deterioration or not.
