Every year,floods damage farms,agriculture lands,roads,dams,bridges and ect.in some events dead many people and sticks, inresults damage social structure.Frequency and numbers of floods in Iran increasedamages,inrecent many watershed management studies do in floods control majour,so with investigating,influence factors on floods and developing quantitative models can estimate peak floods discharges. This research do in Can basin , it is one of the nourth catchments in Tehran city ,it has five main subbasins (Emamzadeh Davood,Taloon,Rendan,Keshar and Soleghan).developing quantitative models infact are linear & noun linear regression models that best model has maximum (Rē) and minimum (Se) ,these models obtain with make regression relationships between estimated peak floods discharges and geometricparameters(Area,Median slope weighting,form coefficient,drainage density, main chaels longth,median elevation,no infiltration lands in percentage, geology factors in PSIAC model,plant cover percentage weighting ,curve number (CN) and precipitation intensity) that most of them estimate in GIS softhwares, finally area, main channels longh and median slope weighting were important factors influence on peak dischargein linear &noun linear methods.quantitative models for estimated peak discharge,explained when we have 3 geometric factors (area,main channel longth,median slope weigthing) we can substitude in models in every return period and give (2,5,10,25,50,100)peak discharges for every basins in Can catchment.

Key words:

Investigation,floods, flood products,geographic information systems(GIS) and geometric parameters