Global NEST Journal, Vol 14, No 3, pp 362-370, 2012 Copyright© 2012 Global NEST Printed in Greece. All rights reserved REGRESSION ANALYSIS BETWEEN SEDIMENT TRANSPORT RATES AND STREAM DISCHARGE FOR THE NESTOS RIVER, GREECE I. ANGELIS Department of Civil Engineering A. METALLINOS Democritus University of Thrace V. HRISSANTHOU* 67100 Xanthi, Greece Received: 20/3/12 *to whom all correspondence should be addressed: Accepted: 27/7/12 e-mail: vhrissan@civil.duth.gr ABSTRACT Systematic measurements of sediment transport rates and water discharge were conducted in the Nestos River (Greece), at a place located between the outlet of Nestos River basin and the river delta. This basin area is about 838 km 2 and lies downstream of the Platanovrysi Dam. Separate measurements of bed load transport and suspended load transport were performed at certain cross sections of the Nestos River. In this study, relationships between sediment transport rates and stream discharge for the Nestos River are presented. A nonlinear regression curve (4 th degree polynomial curve; 2 r equals 0.62) between bed load transport rates and stream discharge, on the basis of 63 measurements, was developed. In addition, a nonlinear regression curve (5 th degree polynomial curve; 2 r equals 0.95) between suspended load transport rates and stream discharge, on the basis of 65 measurements, was developed. The relatively high 2 r values indicate that both bed load transport rates and, especially, suspended load transport rates can be predicted as a function of the stream discharge in the Nestos River. However, the reliability of the regression equations would have been higher if more measured data were available. KEYWORDS: Sediment transport, Bed load, Suspended load, Stream discharge, Measurements, Regression analysis, Nestos River. 1. INTRODUCTION Stream sediment transport is classified into bed load transport and suspended load transport on the basis of the two different motion patterns. Bed load transport depends mainly on the hydraulic characteristics of the streams (e.g. discharge; Maniak, 1988), consists mainly of coarse material, and originates from stream bed erosion. Suspended load originates from soil erosion products inflowing into the streams from the surrounding basins and stream bed erosion, and depends on both hydraulic and rainfall characteristics (e.g. rainfall depth, intensity). This can be explained as follows: rainfall and runoff cause soil detachment and hence erosion products can reach the streams and are then transported in the streams as suspended materials. However, the measurements of suspended load transport are not usually conducted promptly after rainfall events, and hence suspended load transport mainly originates from river bed erosion. On this basis, both bed load transport and suspended load transport are correlated with stream discharge. A common method to represent the dependence (linear or nonlinear) between two measured variables (in this case between bed / suspended load transport and stream discharge) lies on the regression analysis. The dependence between two variables could be depicted graphically using a curve, whereas the degree of the linear dependence is usually expressed by the correlation coefficient (Pearson product-moment correlation coefficient). Numerous regression relationships