Predicting Bridge Pier Scour Depth Using LTS Regression.

Abstract

In this paper, we discuss seven equations that have been suggested by various authors to predict scour depth. The seven equations are (1) Inglis-Poona scour equation, (2) Laursen-Tach curve (3) Railway Practice after Lacey and Inglis (4) Mushtaq Ahmad's equation (5) Blench's equation and (6) Aminuddin and Nalluri 's equation. Almost all observed data must consist ofone or more extreme observations which are called outliers. In the presence ofoutliers ordinmy least squares (OLS) method used in multiple linear regression are not appropriate. Thus we use the least trimmed squares (LTS) method to obtain the modelfor predicting pier scour depth. The result obtainedfrom the analysis show that Aminuddin and Nalluri 's equation produced the best model.