Rooftop photovoltaic (PV) systems, as a common form of photovoltaic applications on buildings, can harness solar energy while reducing air-conditioning energy consumption. However, they also alter the rooftop airflow field and surface convective heat transfer, thereby influencing building thermal loads and the photoelectric conversion efficiency of PV systems. This study investigates the forced convective heat transfer characteristics of rooftop and PV surfaces under varying wind directions. First, a forced convective heat transfer model based on the shear stress transport (SST) k-ω turbulence model is developed and validated. Subsequently, the effects of wind direction θ, PV tilt angle φ, PV spacing D, installation distance W, and setback distance S on the paired convective heat transfer coefficients (CHTCs) are presented and discussed. The results show that when θ = 30°, the convective heat transfer on both the roof and PV surfaces reaches its maximum, whereas other wind directions lead to an approximately 30% reduction. Across different wind directions, W modifies the convective heat transfer by up to 45%, while φ and D induce variations of about 20%, and the effect of S is smaller. Finally, a new correlation for the paired CHTCs is proposed as a function of θ, φ, D, W and the reference wind speed U₁₀. This work provides a basis for predicting building thermal loads and PV performance in buildings equipped with rooftop PV systems.