Simulations play a fundamental role for the evaluation of vehicular network communication strategies and applications' effectiveness. Therefore, the vehicular networking community is continuously seeking more realistic channel and reception models to provide more reliable results, yet maintaining scalability in terms of computational effort. We investigate the effects of vehicle shadowing on IEEE 802.11p based communication. In particular, we perform a set of real world measurements on a freeway and study the impact of different obstructing vehicles on the received signal power distribution. Different vehicle types not only affect the average received power, but also its distribution, suggesting that the attenuation characteristics of the simulation model need to be tailored to the type of vehicle that obstructs the communication path. Based on these observations, we propose a novel way to compose shadowing and fading models to reproduce the observed effects.