An extraction of model parameters is the key for remote sensing application in the calculation and forecast of flood

which mainly depends on relationships between characteristic informations of a basin obtained by remote sensing data and characteristics of model parameters. There are close relations between the flood stochastic model-rgeomorphoclimatic instantaneous unit hydrograph （Rodriguez-Iturbe

et al.

1979.） and the characteristics of a basin. The model is deduced through Marcov process of raindrops uniformly distributed over the river ordering system of a basin

which is advanced by R. E. Horton and is defined by A. N. Strahler. At this basis

we take Jieting basin of Puyong river in Zhejiang Province for the experimental area of the model. According to the map of river ordering system of a basin

which is interpresed from LANDSAT MSS image we extracted the geomorphic parameters of the model and calculated the processes of five floods in the basin. And the results indicated that the calculative precisions of the model can meet the requiements of flood calculation.