Abstract: Based on DEM, hydrology principle, this paper synthesizes the basic principles of GIS technology and urban intrinsic, impervious surface and considering infiltration and drainage, simulates water flow along the grid unit from high to bottom, distributes flow according to slope ratio, and forms the urban interior scenario simulation in different rainfall intensity, which extracting information such as the risk range and distribution characteristics of typical water points in Xi'an. The spatial analysis and data validation results are consistent with the actual results. Superimposed surface coverage of classification results and statistical yearbook data include population, economy, traffic, buildings, it get different risk of inundation, then introduce grey correlation theory to quantify the degree of correlation between rainfall and loss factors, and to explore comprehensive influencing factors of the intrinsic loss in the city. It shows that the correlation between rainfall and economic loss factor is the largest, reaching 0.823, and the correlation with the population loss factor is the smallest, reaching 0.725. The order from large to small is:economy>building >Inundation area>Road>Population. So, fully considering the artificially controllable urban intrinsic impact factor, it proposes to reasonably control the urban land expansion speed and improve land use efficiency, aiming at providing reference for the management, prediction and decision-making of urban inland.

Key words: urban guilt, scenario simulation, GIS, grey correlation degree, guilt loss