The main visibility observation indicators for civil aviation ground meteorology include Runway Visual Range (RVR),Dominant Visibility (VIS),and Meteorological Optical Range (MOR). RVR is the visibility standard recognized by the International Civil Aviation Organization (ICAO) for determining the takeoff and landing of aircraft. The RVR estimation models recognized by ICAO are monopolized by foreign aviation instrument manufacturers,which hinders the development of RVR early warning and forecasting technology. Compared to MOR,RVR is also affected by the vertical distribution of visibility and is closely related to background brightness and runway light intensity. The complexity of the boundary layer structure and the unpredictability of non-meteorological factors both pose challenges for the early warning and forecasting of RVR. This article introduces the relationship between RVR,MOR and VIS,analyzes the impact of meteorological factors,background brightness,and runway light intensity on RVR,reviews the current technological progress in visibility forecast in the aviation industry,and explores possible ways to improve the aviation visibility forecast skill. RVR, MOR, and VIS all represent visibility,with different physical significance,detection standards,and influencing factors,leading to significant different values. But,they have a high correlation and can be substitutable under some certain conditions. The influence mechanism of meteorological factors on MOR has been widely studied,but there is a lack of research on the impact characteristics of background brightness and runway light intensity on RVR. There is no brightness factors forecast at present,which affects the the accuracy of RVR forecast. Compared to the current ground observation of MOR,using lidar and satellite measurements of slant visibility can provide a more a more RVR-physical-significance-based fundamental visibility. Currentaviation visibility forecasting technologies based on deep learning models struggle to forecast the inflection points in visibility evolution trends and are unable to provide long-term effective forecasts. Looking forward to the development trends of artificial intelligence and numerical forecasting technologies,combining these two methods and integrating effective meteorological factor forecasting to establish a physically contrained intelligent visibility forecast model may be able to provide high-precision visibility forecasting products.