Modeling landscape with high-resolution digital elevation model (DEM) in a geographic information system can provide essential morphological and structural information for modeling surface processes such as geomorphologic process and water systems. This paper introduces several DEM-based spatial analysis processes applied to characterize spatial distribution and their interactions of ground and surface water systems in the Greater Toronto Area (GTA), Canada. The stream networks and drainage basin systems were derived from the DEM with 30 m resolution and the regularities of the surface stream and drainage patterns were modeled from a statistical/multifractal point of view. Together with the elevation and slope of topography, other attributes defined from modeling the stream system, and drainage networks were used to associate geological, hydrological and topographical features to water flow in river systems and the spatial locations of artesian aquifers in the study area. Stream flow data derived from daily flow measurements recorded at river gauging stations for multi-year period were decomposed into "drainage-area dependent" and "drainage-area independent" flow components by two-step "frequency" and "spatial" analysis processes. The latter component was further demonstrated to relate most likely to the ground water discharge. An independent analysis was conducted to model the distribution of aquifers with information derived from the records of water wells. The focus was given on quantification of the likelihood of ground water discharge to river and ponds through flowing wells, springs and seepages. It has been shown that the Oak Ridges Moraine (ORM) is a unique glacial deposit that serves as a recharge layer and that the aquifers in the ORM underlain by Hilton Tills and later deposits exposed near the steep slope zone of the ridges of ORM provide significant discharge to the surface water systems (river flow and ponds) through flowing wells, springs and seepages. Various statistics (cross- and auto-correlation coefficients, fractal R/S exponent) were used in conjunction with GIS to demonstrate the influence of land types, topography and geometry of drainage basins on short- and long-term persistence of river flows as well as responding time to precipitation events. The current study has provided not only insight in understanding the interaction of water systems in the GTA, but also a base for further establishment of an on-line GIS system for predicting spatial-temporal changes of river flow and groundwater level in the GTA.