This report describes a methodology for modeling and estimation of visibility in vegetated terrain, based on publicly available national geographic data. A model for visibility estimation has many military applications, e.g. assessment of observation opportunities and the risk of detection, or improved realism in simulations. Focus has been models that capture how vegetation in the form of tree trunks, branches, leaves etc. affects the visibility along the ground, and enables estimates from national geographic data. Other influencing factors, such as sensor charecteristics, weather (fog, rain, ...), lighting conditions, smoke or type of target, is not addressed within this report. The work with the model has drawn inspiration from established methods within the forestry industry for mapping and inventory of forests. Reference data of actual visibility have been collected with terrestrial laser scanning. Connection with national geographic data, laser data and national land cover data at the corresponding locations are then modeled using machine learning. The results show that visibility can be estimated from national data, albeit relatively roughly with the simple models, limited reference data and sparse national laser data used in this study. However, everything suggests that with future, more detailed laser data and collection of more extensive reference data, the quality of the estimates will increase.