Civil Engineering March 2021 | Vol 29 No 2

Civil Engineering March 2021 55 time series for some stations with very few field measurements,” says Nandiga. Now that the data scarcity constraint is being addressed in the Wami-Ruvu and Rufiji basins, inadequate knowledge of the quality and quantity of the available water resources should no longer delay their development for the benefit of the people in these basins. Hopefully, the use of a similar system in the other river basins in Tanzania will follow, as the Ministry of Water is determined to see this happen. Kizigo River A single precision measurement on the Kizigo River provides an example of how modern measurement and analysis techniques can help enormously with the determination of river flow over an extended period. On the Kizigo River, in a four-year period in the 1970s, more than 250 streamflow measurements were collected, followed by a few more measurements in the early 1980s. After that, no more reli- able measurements were conducted due to the difficult terrain, and hydrologists were unable to determine the rating curve and by association the stream flow record. This measurement of the flow of the Kizigo River is vital as it is the largest contributor to the Mtera reservoir, which provides water for the Mtera hydroelectric facility. Streamflow data for the Kizigo River is needed to plan the operation of the power plant’s turbines. Yet this information has not been available with any degree of confidence for the last three decades. River flow is one of the more difficult variables to measure on a continuous basis, so it is commonly derived by converting daily level recordings into discharge by using a rating curve. Establishing the shape of the rating curve historically requires lots of streamflow measurements. However, in this case the difficulty of accessing the remote station, and insufficient funding for a field measurement programme, meant that data was scarce. To solve this problem, WARIDI used Aquarius Time Series software to rede- velop rating curves using a robust concep- tual model that requires far less field data. This ‘hydraulic’ approach is based on the premise that the parameters of the rating curve function can be inferred from field observations of river channel properties. Discharge measurements at the station’s control section then serve to calibrate and fine tune the conceptual model. WARIDI has supported an extensive field measurement programme to collect up-to-date station data needed to develop the conceptual rating curve models. 16 stations were visited in the Wami-Ruvu basin and 26 in the Rufiji basin, including hydro station Chinugulu at Kizigo. WARIDI trained 10 hydrologists from the Wami-Ruvu and Rufiji Basin Water Board Offices on river flow measurements using state of the art Acoustic Doppler Profilers and on Aquarius software to improve rating curves and sustain application of both practices by Basin Water Office staff. The visit to Kizigo station took place during the rainy season in March when water levels were high, and the river was in full flood. A high-quality discharge measurement was conducted – the first for a flood event for Kizigo in over 40 years. This single data point proved invaluable in re-establishing the rating curve parameters and calculating the dis- charge for the entire historic data record. Kizigo, which has an entrenched channel with a narrow floodplain on the right bank, had essentially maintained its channel shape since 1977, apart from a gradually rising riverbed caused by sedi- ment load deposits, probably because of upstream land degradation. Nevertheless, the old field observations in combination with the latest measurement – and using the hydraulic approach facilitated by the Aquarius software – proved that the basic shape of the 1973 to 1977 rating could be maintained. It only required adjustments for the rising riverbed. “A single measurement provided the verification. Aquarius provided the tools. The persistence and commitment of the observer provided the basic informa- tion. Combined, it resulted in a 40-year record of highly valuable water data,” concludes Nandiga. eNDnOTE This work was made possible by the support of the American people through the United States Agency for International Development (USAID.) The contents of this article are the responsibility of Aquatic Informatics and do not necessarily reflect the views of USAID or the United States Government.  South Africa