Comparative study of original Ackers & White relation along with modified coefficient for total sediment load prediction.


Sahita Waikhom, Dr. Sanjay M Yadav

Wednesday 1 july 2015

9:15 - 9:30h at Mississippi (level 1)

Themes: (T) Sediment management and morphodynamics, (ST) Sediment transport mechanisms and modelling

Parallel session: 8A. Sediment - Transport


Reliable and quantitative estimates of sediment transport are very important in river engineering and water management projects. Numerous comparison studies, by researchers like Nakato T, Wu et al. and others, have been conducted on sediment transport using various total load formulas for wide range of flume and field data. Most of these studies conclude that Ackers & White relation is quite reliable and a good predictor. In light of this outcome, an analysis was done using this relation for the flume data set of Samaga et al, Soni J P and river data of Tofalleti F. B, with an aim to verify the validity of result for Indian terrain envisaging further test and applicability of the Ackers & White relation. Comparison was also done with Ackers & White relation incorporating modified coefficient. The performance of the selected sediment formula was tested on the basis of accuracy in predicting the observed sediment within a certain discrepancy ratio. Statistical analysis and graphical comparison are utilized to demonstrate the performance and variations for different data sets. Overall, original Acker & White function under predicts, with percentage error ranging from -20% to -90 %, for all the selected data sets. The evaluation revealed that, for all four mixtures (of Samaga et al data set), prediction of sediment load is superior in comparison to other data sets. It was also observed that both Acker & White (1973) and Acker & White (1990) function gives better value for d35 as representative size of the bed material instead of d50 in Samaga et.al data set. It was also noted that the result predicted by both selected transport function has little difference for d50 ranging from 0.24 mm to 0.35 mm in almost all chosen data sets. Analysis of the result also conclude that Ackers and White (1973) function changes its trend from under prediction to over prediction as the diameter decreases but interestingly, predictions with the modified parameters (Ackers and White, 1990) deviate significantly from that of the original formula, except for a very few cases for sediment diameter less than 0.18 mm as unpredictable trend is observed in Ackers and White (1990) approach. When the diameter decreases it over predicts but again in some cases it under predicts to high degree of deviation so diameter of sediment alone cannot judge the trend of prediction which was observed in the prediction of transport rate using Tofalleti. F. B. data set.