The comparison of visibility measurement between image-based visual range, human eye-based visual range, and meteorological optical range

Abstract

Before the 21st century, visibility monitoring techniques are mainly categorized by human eye-based and optical-based measurements. In order to automate and objectify the former measurement, the latter measurement method has been developed and meteorological optical range (MOR) is used as a concept similar to the visual range. However, in the 21st century, MOR has been found to have some differences with visual range that is recognized by human beings. As image analysis technology is improved, a variety of image-based on-site visibility measurement techniques are being developed using a camera. In this study, image-based visibility measurement that is based on the measured distance to visible objects was performed at about 3 km northwest of the Seoul Meteorological Observatory where the human eye-based and the optical-based visibility monitoring are observed simultaneously. As a result of correlation analysis between three measurement data, the image-based visual range (IVR) was found to be more comparable to the human eye-based visual range (HVR) than to the optical-based MOR. HVR tended to be lower than IVR as visibility improved, but MOR was predominantly higher than IVR as visibility was reduced. Perceived visibility measurement of IVR and HVR were revealed to be more accurate in severe poor visibility conditions. In this study, the system uncertainty of IVR estimation is expected to be within 10% up to about 33 km and approach 20% under above a hundred km of visibility. The Visual Range Meter (VRM) program developed for IVR estimation was made up of an individual measurement mode and a real-time measurement mode. The operational errors of approximately 15% excluding precipitation and other meteorological effects occurred in real-time IVR estimation using the VRM.