Vol.28 No.3(6) [2006]

Multiple Regression Analysis for the Discharge of Changjiang in China

Tomoharu SENJYU and Hirofumi ENOMOTO


An estimation method for the Changjiang discharge from precipitation anomalies is proposed on the basis of multiple regression analysis. The estimated discharge anomaly shows a significant correlation with the observed one (with a multiple correlation coefficient R of +0.45). The root mean square between the estimated and observed discharge anomalies is 5968.6 m3s-1. The partial regression coefficients show a relatively larger contribution to the discharge anomaly from precipitations in the middle region of Changjiang.