The proposed framework produces wind resource maps and estimates the energy production in complex terrain by combining computational fluid dynamics (CFD) with observations from an arbitrary number of masts. It provides a fast and comprehensive solution to the research gap that is to obtain accurate CFD inflow conditions and to mitigate the modeling error by incorporating observations to produce a wind resource map and estimate energy yield. Wind data are processed by direction bins both to initialize simulations and to combine with the latter. The wind resource map is then obtained by superposing the power density maps from each sector. Ultimately, a database containing different turbine models and hub heights is used to filter the best performing cases. Validation at two complex terrain sites reveals that the wind flow model is sufficiently accurate using the proper parameters, and that reasonable inflow conditions and assimilated wind speed fields are achieved. The framework was tested at a complex site in Brazil and was more sensitive to the number of simulated wind directions than to grid refinement. This research provides a comprehensive contribution toward wind resource mapping in complex terrain because it is computationally fast and flexible in the number of masts. (C) 2020 Elsevier Ltd. All rights reserved.