Development of low flow prediction models for southern Africa
| dc.contributor.author | Mngodo, Raymond Julius | |
| dc.date.accessioned | 2020-06-05T06:50:14Z | |
| dc.date.available | 2020-06-05T06:50:14Z | |
| dc.date.issued | 2002 | |
| dc.description | Available in print form, East Africana Collection, Dr. Wilbert Chagula Library, Class Mark ( THS EAF GB1201.A3M6) | en_US |
| dc.description.abstract | Hydrological Network does not cover all catchments in the Southern Africa region. There are many ungauged catchments that could have the potential for water resources development. A database was established for 638 river flow stations from 11 Southern Africa countries, covering an area of 6,929,826 km2. Spatial database consisting of river basins, gauged catchments, national boundaries; river; rainfall; potential evaporation; wetlands; and geology was established using ARC-INFO. The objective of this study was to develop low flow prediction models. The low flow index Q70 from flow duration curve was used to describe flow regime and map spatial variability of mean annual runoff, temporal variability of annual runoff, temporal variability of annual runoff and base flow contributions to river flow. About 25% of the rivers are ephemeral, 30% are intermittent and the remaining 45% are perennial rivers. Linear regression models were developed to predict Q70 for ungauged catchments in the nine primary basins of Southern Africa using catchment area, MAR, AAR, BFI and their GIS coverage’s. The BFI has a strong influence in estimation of Q70. For Tanzania, a non-linear regression equation was obtained by including geology indices. Eight homogeneous regions of 10-day annual minimum flows for Tanzania were delineated using a simple test based on the variability of at-site values of Cv: The L-moment ratio diagrams and the goodness of fit test of Hosking and Wallis are used to assess the suitability of selected distributions as regional parent distributions. The LLG distribution provides a good fit to low flows in one region while the LN distribution fits well in seven regions. The GEV-4 and non-parametric kernel estimation models were explored as other possible methods of low flow analysis. The GEV-4 models and the NKE when compared with Weibull models for its predictive and descriptive ability tests showed better results. | en_US |
| dc.identifier.citation | Mngodo, R. J (2002) Development of low flow prediction models for southern Africa, Doctoral dissertation, University of Dar es Salaam | en_US |
| dc.identifier.uri | http://41.86.178.5:8080/xmlui/handle/123456789/12056 | |
| dc.language.iso | en | en_US |
| dc.publisher | University of Dar es Salaam | en_US |
| dc.subject | Rivers | en_US |
| dc.subject | Stream flow | en_US |
| dc.subject | Stream flow measurements | en_US |
| dc.subject | River channels | en_US |
| dc.subject | Africa Southern | en_US |
| dc.subject | Regulations | en_US |
| dc.title | Development of low flow prediction models for southern Africa | en_US |
| dc.type | Thesis | en_US |