Ngaina JN and Opere AO
This study sought to investigate hydrological characteristics of lower Nzoia sub-basin in Kenya. Trend (long and medium) utilized graphical and statistical approach. Extreme value (EV) analysis based on frequency, annual maximum flows, exceedance probability, low flows and probable maximum precipitation was used. Quality control showed consistency in rainfall, temperature and discharge datasets. Maximum-minimum monthly and annual flows and rainfall showed maximum flows centered in March-to-May (peak) with increasing temperature. Trends, seasonality and cycles were identified and maximum values in rainfall and discharge closely followed the pattern for peak rainfall seasons. Based on the flow magnitudes and 100-year return period, the upstream station (1EE01) had lower values compared to downstream station (1EF01) for different assumed distributions and thus 1EF01 assumed to be more reliable. Exponential and Pareto distributions indicated a normal tail and thus appropriateness of EV1/Gumbel distribution in calibrating AM series. Best conventional calibration results based on assumed distributions using EV1/Gumbel superimposed with the extreme value distribution fitted along with the exponential/Pareto Q-Q plots for comparison. Estimated maximum withdrawal in monthly terms for 1EE01 and 1EF01 was 262.5 and 368.4 cumecs respectively. Analyses of low stream flow indicate probable availability of water in streams at different return periods.
Comparte este artículo