Non-contact road surface meteorological detection technologies have emerged as a significant area of development due to their non-destructive impact on the road foundation and the simplicity of installation and maintenance. Typically, these non-contact road surface meteorological detection technologies utilize optical detection methods, and factors such as the roughness of the road surface and the optical angle of incidence significantly influence the system's performance and the accuracy of the meteorological measurements.According to the optical geometric ray method, an improved microfacet model is proposed, which introduces multiple random parameters generated by the reflection of light from rough road surfaces, and establishes a hemispherical equivalent simulation model. This model microscopically elucidates the reflective properties of photons when interacting with rough road surfaces, and it allows for the convenient and precise simulation and analysis of the distribution of photons after reflecting off rough surfaces. Building on this, a rough road surface link transmission model based on wireless laser transmission theory has been developed to study and simulate the optical power characteristics received by the detection system under different road roughness levels and angles of incidence.The random distribution function of the normals of road microfacets under varying degrees of roughness is obtained by using refusal sampling technique, which determines the changes in photon reflection direction, and the distribution state of photons after reflection from the rough surface is statistically analyzed by using the Monte Carlo method, which derived the variations in reflected optical power under different angles of incidence and road roughness conditions. Subsequently, the validity of the model is confirmed. For the experimental design, a non-contact laser-based road surface meteorological condition detection system operating at a wavelength of 850 nm is constructed, which mainly consists of the light source drive circuit with emitting the light power of 50 mW, the laser receiving unit, and the optical system (including an optical antenna, the optical filters, and an optical collimator, etc.). The system is positioned at a vertical height of 2 m from the road surface to be measured, which is capable of not only monitoring road conditions in real time but also validating the photon distribution and optical power variation predicted by the simulation model. The simulation results and experimental data both reveal a trend where the received optical power gradually decreases as the incident angle between the incident light and the road surface normal increases. Notably, at an incidence angle less than 15°, the greater the road surface roughness, the lower the received optical power. Conversely, at angles greater than 15°, the trend reverses—the greater the road surface roughness, the higher the optical power, and this relationship tends to become linear at certain roughness levels. When the incidence angle reaches 60°, the received optical power stabilizes and undergoes minimal further change. Additionally, the experimental results indicate that the signal-to-noise ratio of the received optical signal does not change with the variation of road roughness, but closely correlates with the incident angle.This study presents and validates an equivalent simulation model for the reflection of light from rough road surfaces, and confirms the model's accuracy and feasibility in practical applications through experiments with an actual non-contact road surface meteorological detection system. The findings not only enhance our understanding of road surface reflective properties but also offer practical insights for the optimization of road detection techniques and meteorological condition monitoring. Thus, the research provides a theoretical and technical support for further improving road detection technology and monitoring meteorological conditions, ultimately contributing to the advancement of road safety measures.