Using Multiple Regression Models to Investigate Effectiveness of Watersheds Properties on Suspended Sediment Yield (SSY)

Hamid Jalilvand¹, Mahmooh Habibnejad², Rajabali Fatahi

¹ Department of Forestry, Mazandaran University, Iran
Presenting Author: Dr. Hamid Jalilvand
phone: +98-152-4222981-5, fax: +98-152-4222982
e-mail: hj_458_hj@yahoo.com

² Department of Pasture & Watershed, Mazandaran University, Iran

Abstract

In this research as a result of reaching suitable method for purpose of mean suspended sediment yield prediction under ten measuring and hydrometric stations were selected. These stations must have statistical data and appropriate information scattering in region of Neka Rood and Gorgan Golf to measure effective different variations on suspended sediment yield. Different effective variables on suspended sediment yield were measured. Also twenty years of mean suspended sediment yield stations (1360-1380) were selected to estimate curve rating method using multiple regression models.  Multiple regression models with step by step method when suspended sediment yield was as a depended and watershed characteristics were as an independent variables were used. At the first step, totally, 104 variables were identified that could inserted in the simple regression models. The percentage of area was sensitive to erosion before quaternary formations (R2=0.996) the percentage area relatively resistant to erosion of quaternary formations (R2=0.996), gravelius coefficient (R2=0.997), the percentage area of dry farming and range mixed (R2=0/997), the percentage area of high density and medium density forest mixed (R2=0.993), mean annual of discharge (R2=0.61), had effectively identified on SSY.

Keywords: Suspended Sediment Yield (SSY), Multiple Regression, Area, Nekarood and Golf of Gorgan