Identification of basin characteristics influencing spatial variation of river flows

The selection of basin characteristics that explain spatial variation of river flows is important for hydrological regionalization as this enables estimation of flow statistics of ungauged basins. A direct gradient analysis method, redundancy analysis, is used to identify basin characteristics, which explain the variation of river flows among 52 selected basins in Zimbabwe. Flow statistics considered are mean annual runoff, coefficient of variation of annual runoff, average number of days per year without flow and selected percentile flows. Basin characteristics investigated are those likely to reflect climatological, topographical and hydrogeological influences including that of land cover on river flows. The first ordination axis of flow statistics is strongly correlated with mean annual precipitation, mean annual potential evaporation and median slope. This ordination axis explains 64% of the variation of selected flow statistics among the selected drainage basins. The proportions of a basin under cultivation, and that with grasslands are correlated with the second ordination axis, which explains 6% of the variation of selected flow statistics. Mean annual precipitation is the most important basin characteristic, and this alone explains 50% of the variation of flow statistics. Median slope is the second most important basin characteristic. Proportions of a basin underlain by different lithological types had no effect on flow characteristics of selected basins. The paper has demonstrated the ability of redundancy analysis to identify basin characteristics that explain the variation of river flows among basins, including estimating the relative importance of these basin characteristics.