Unmanned aerial vehicles, i.e., drones, have recently caught attention for providing on-demand capacity to wireless networks as drone-base-stations (drone-BSs). Many studies assume simplified channel models based on average characteristics of the environment to estimate the placement of drone-BSs. However, especially in urban areas, positioning of drone-BSs with respect to intersections and roof-top heights of buildings can severely change the path loss characteristics. To address this issue, we adopt an ITU channel model utilizing more information about the environment, such as the shapes of the buildings. We optimize parameters of the selected ITU model, so that it can be used for altitudes both strictly lower and higher than building roof-tops. Using ray-tracing simulations as a benchmark, we compare the proposed model with a widely used simpler model. Results show that the proposed model can reduce the root-mean-squared error from 35 to 10 dB, which may have critical implications for drone-BS operations, such as planning for the required number of drone-BSs to cover outdoor urban users, as demonstrated with simulations.