Outage probability and Bit Error Rate (BER) are both important link quality parameters for wireless communica-tion systems. In this paper, the error in dB between the Multi-Resolution Frequency-Domain ParFlow (MR-FDPF) model and the channel measurement is first modeled to follow the normal distribution which is verified by performing the Lilliefors good-ness of fit test. Then a realistic BER map as well as its confidence interval can be predicted. In addition, the normally distributed error in dB can be regarded as a kind of lognormal shadowing introduced by the MR-FDPF model due to, for instance, the inaccurate scenario modeling. Then a realistic outage probability can be computed by taking into account both the large scale lognormal shadowing and the small scale Nakagami-m fading.