Fluctuations in river flows are important as they affect irrigation and therefore the livelihood of people. The Vaal river in South Africa is and important water source in the region and was studied by Alexander et al (Journal of the South African Institute of Civil Engineering Vol 49 No 2, June 2007, Pages 32–44, Paper 659) who give this record of its flow which demonstrates the huge fluctuations in water supply over time:
The spectrum of the Vaal river flow shows a number of prospective cycles, several of which are significant (in bold):
Alexander claims a 22 year cycle based on sunspot cycles with alternate cycles reversed. However the spectrum shows a cycle of 18.3 years which is different enough from 22 years to be confident that it is not the correct period. I think that Alexander gets this value because he uses only the lows of the cycles and not the highs:
It seems more likely that the detected period of 18.3 years is related to a lunar cycle than a solar one. The lunar cycle of 18.6 years strongly affects patterns in tides in the sea and there are reports of it affecting rainfall through affecting atmospheric tides. Sometimes it also appears as a 9.3 year cycle.
The 10.35 year cycle might possibly be related to solar activity, although the measured period is rather shorter than the expected 11.1 years. Along with the 18.3 and 10.35 year cycles, other cycles of 6.49, 4.46 and 2.37 years were compared to the fluctuations:
When these five cycles are combined the fluctuations can be compared to the actual variations in river flow and a projection can be made. Because the latest data in the study are some 14 years ago, the projections could be tested against real data if it is available.
Whether this projection is useful or not is not known at this time. However cycle projections might be a useful tool in planning in the many regions where water flows sometimes lead to disasters.
Cycles Research Institute
Leslie Carol Botha
K. David (Rudraksha) Katzmire
Ray Tomes
tallbloke
Cycles Research Institute's Blog 
Excellent piece of work! Well done Ray.
My father was a water resources civil engineer and he also noticed a ~10.5 year cycle in drought and flood in my home area of Yorkshire UK.
Regarding the solar connection, it’s worth noting that the short high amplitude cycles which dominated the second half of the period of this record will have brought the 11.07 year long term average solar cycle down a bit.
To me, it makes sense that both Sun and Moon would afect river flow through precipitation changes.
Another study which may interest you is this one from South America.
Solar Forcing of the Stream Flow of a Continental Scale South American River
Pablo J. D. Mauas*
Instituto de Astronomı´a y Fı´sica del Espacio (CONICET-UBA), C.C. 67 Sucursal 28, 1428, Buenos Aires, Argentina
Eduardo Flamenco
Instituto Nacional de Tecnologı´a Agropecuaria, Rivadavia 1439, 1033, Buenos Aires, Argentina
Andrea P. Buccino
Instituto de Astronomı´a y Fı´sica del Espacio, (CONICET-UBA), C.C. 67 Sucursal 28, 1428, Buenos Aires, Argentina
(Received 18 April 2008; published 17 October 2008)
Click to access reprint_parana.pdf
Thanks Rog.
Yes, the average sunspot cycle length has been lower in the last century, but 10.35 years seems too low, so correlation analysis is the way to go to be sure.
In that paper, the graph in figure 2 I fancy that I see cycles of about 35 years and 18-20 years also.
The long term English temperature series shows strong cycles of 24.3 and 22.7 years, the latter being near the magnetic sunspot cycle period.