Developing Sediment Yield Prediction Equations for Small Catchments in Tanzania - Advances in Data, Methods, Models and Their Applications in Geoscience - page 255

10100

9100

8100

7100

6100

5100

4100

3100

2100

1100

100

0 50 100 150 200 250 300

Catchment area (km2)

Observed data Predicted fill-yield Upper 95%

Predicted fill-yield Lower 95% Predicted fill-yield Average fit

Fig. 3.2. Relationship between sediment fill-yield and catchment area for both dry and
moderate climatic zones of Tanzania

were developed in the form of a power function as proposed in section 3.1 above (Table
3.2.2) and as illustrated in Equation 3.2. Besides, the estimated parameter uncertainty
bounds are presented in the table. Specific equations for administrative regions are also
included.
One would note from Table 3.2.2 below that the performances, as measured by coefficient of
determination, R 2 , of the developed prediction equations are higher in moderate climate
zone than in dry climatic zones. The strength of correlations between sediment yields and
catchment area sizes could be categorized as high, moderate and low for Arusha and
Tabora, Dodoma, and Dodoma and Shinyanga and Singida regions, respectively. These
results suggest that large variance in sediment yields remains unexplained by the developed
regressions in dry climatic zone. This may be attributed to high uncertainty in
representations of long term sediment yields in catchment with high temporal variability of
rainfall intensity, runoff and sediment (Mulengera, 2008). However, independent analysis
indicates that sediment fill-yield data used for the dry climatic zones are good in space
representation with coefficient of variation, CV, between 23 and 26 in percent. The
corresponding values for the moderate and all climatic zones ( i.e. , dry and moderate) range
from 64 to 128 in percent. The high sediment yields for Arusha region were expected as the
soils in the region are mostly recent volcanic ash and/or highly erodible. The overall