Distributed hydrological model becomes a focus of ecology hydrological research because it provides insights to solve the problems of multi-scale evaluation and upscaling of hydrological benefits of forest. However distributed hydrological model confront with some new challenges, hydrological process simulation in single scale is difficult to meet the needs of understanding multi-scale characteristic of hydrological process and link among processes at different scales. As basic unit of terrestrial hydrological cycle, watershed has clear boundary and become the foundation of hydrological process research and water resource management. Nested watershed is the expression of hierarchical structure of terrain and become a good experiment area for study on hydrological process and upscaling. In this paper, catchment hydrological theory, dynamic data driver application system (DDDAS) paradigms, and scale theory are applied to construct a multi-scale distributed hydrological model research framework, and some key techniques of multi-scale surface runoff model was solved to support the research on scale effect of hydrological process and upscaling in nested watershed. The main research result follows below.Based on digital terrain analysis and DEM, a research framework of multi-scale distributed hydrological model is put forward. It can support the research and test for identifying main hydrological process at different scales based on different basic model unit, for hydrological process interaction and nesting at different scales in nested watershed by nesting model output at small scale unit into model input at large scale, for capturing hydrological pattern at catchment scale and assisting decision of water resource management by combining multi-scale monitoring data with hydrological model at the proper scale.Based on research of automation segementation of watershed terrain, a multi-scale geomorphy unit family is formed. A new algorithm of auto-generating DEM from digitalized contour based on mathematical morphology is presented for Geomorphologically sound DEM. By comparison of the Methods for extracting digital channel network from DEM, a set of techniques is applied to discretization the watershed into small geomorphy element at different scales. Based on flow Path from DEM data, the method of automated extraction of shoulder line of valley is presented. All of above techniques is integrated to segment watershed into cell, hillslope, catchment and hierarchical nested watershed.A process system for automation of multi-scale geomorphy parameter is developed. Based onautomation segementation of watershed terrain, the algorithm of geomorphy parameters at different scale is described and their dependency relationships is analysed. Take the terrain data of Caijiachuan nested watershed in west Shanxi as an example, geomorphy parameters at different scales is produced and multi-scale analysis is conducted to discovery the rules of scales' changes on them.Under the guidance of DDDAS paradigms, the construction techniques of dynamic distributed hydrological model is studied and a distributed surface-runoff simulation system is developed with support for three basic model units. The simulation system gets the model unit parameters from above process system of geomorphy parameter and support coupling of watershed runoff processes and interaction between simulation processes at different scales. The system is designed to be open and flexsible, and provide virtual laboratory for research on hydrological process simulation and upscaling.Application research of surface runoff model in Caijiachuan nested watershed shows that the model developed in this paper can reflect the impact of the rain pattern on watershed hydrograph and route the simulation of small catchment into large catchment. Finally the impact of forest and human activities on hydrological process and nature environment can be evaluated.