ANALYSIS OF SPATIAL AND SEASONAL VARIATIONS OF WATER QUALITY USING STATISTICAL, CCME-WQI AND GIS METHODS IN NAKURU MUNICIPALITY, KENYA

Abstract

Nakuru Municipality lies in the larger Lake Nakuru basin where the natural chemistry of its watershed is characterized by high levels of inorganic salts posing a significant problem to water resources. River and ground waters are the primary sources of drinking water for the municipalities’ population but despite their significance, there are limitations to the availability of a robust database that can describe their quality status in a reliable way. Information that can explain spatial and seasonal variations of river and ground water quality is not sufficient. This study targeted to bridge this gap by adopting a quantitative research design. Stratified random sampling was applied in the study of pH, electrical conductivity, and levels of fluoride selenium, chloride and cadmium. Thirty eight sampling points and three sampling periods/seasons were selected for this study. The seasons were put in three groups: group 1: Transition Period (TP), group 2: Short Rain Period (SRP) and group 3: Dry Period (DP). The data were analyzed, processed and interpreted using SPSS software Version 22.0, Microsoft window Excel tool, statistical, Geographical Information Systems (GIS) and Canadian Council of Ministers of the Environment Water Quality Index (CCME-WQI) methods. Water quality data were presented as mean values and evaluated based on guideline values to assess its suitability for drinking. Spatial and seasonal variations of examined parameters were descriptively analyzed. The results showed significant spatial and seasonal variations in some of the examined water variables (pH, chloride, fluoride, selenium, cadmium and electrical conductivity). Kruskal - Wallis H test was applied to test for seasonal variation in mean concentration values of water parameters where results were reported at 0.05 level of significance. Seasonally, the results indicated a statistically significant difference between the means of pH (P = 0.001) and cadmium (P = 0.000) but no significant difference for electrical conductivity (P = 0.130), selenium (P = 0.981), fluoride (P = 0.293) and chloride (P = 0.228) in ground water. Post hoc results showed significant existence of mean difference in pH between the TP and DP (P = 0.000) and TP and SRP (P= 0.006) while for cadmium there existed significant mean difference existed between TP and SRP (P = 0.001) and TP and DP (P = 0.000). For natural river water samples, Kruskal - Wallis H test results showed existence of significant differences between the means of pH (P = 0.050) and electrical conductivity (P = 0.008). However chloride (P = 0.629), cadmium (P = 0.122), fluoride (P = 0.105) and selenium (P = 0.472) did not show any significant difference. Post hoc results showed existence of significant mean difference in electrical conductivity between TP and DP (P = 0.004) and SRP and DP (P= 0.016) while for pH significant mean difference existed between TP and DP (P = 0.034) and SRP and DP (P = 0.34). Spatial distribution maps gave a pictorial representation of the spread of water quality parameters at different sampling points. Based on the CCME-WQI, index values for drinking water quality of river and ground water samples were calculated in a range of 20.94-39.69 and ranked as poor. Selenium, cadmium and fluoride were taken as important parameters impacting water quality as their concentrations were elevated. Aquifer geological formations and water quality characteristics showed spatial variation among the boreholes. Considering all examined variables, correlation results showed that correlation coefficient (r) values ranged from -0.85 to 0.70 in river water and -0.87 to 0.56 in ground water. The study concluded that the quality of river and ground water of the study area was mainly influenced by geological hydrogeological settings of the area, not potable and need proper treatment before consumption.