Flow event deposits in turbidite lobes are modeled with stochastic surface-based simulation. This method honors the geometries and compensational stacking of flow event deposits. Flow event deposit geometries are based on a flexible lobe parameterization. Compensational stacking is the tendency of flow event deposits to fill topographic lows and to smoothing of topographic relief. The surface-based model may be conditioned to well data. Models of reservoir proper-ties such as porosity and permeability are constrained by the resulting geometric models. This approach is applied in a geostatistical workflow to better integrate available geologic information. The resulting models may improve the accuracy of model reservoir response and account for the uncertainty in the heterogeneity of turbidite lobes.