Use of cameras for monitoring visibility impairment. (February 2018)

• Record Type: Journal Article
• Authors: Malm, William; Cismoski, Scott; Prenni, Anthony; Peters, Melanie
• Is Part Of: Atmospheric environment. Volume 175(2018)
• Page Start: 167
• Page End: 183
• Publication Date: 2018-02
• Subjects: Webcams -- Visibility -- Image indexes -- Extinction; Air -- Pollution -- Periodicals; Air -- Pollution -- Meteorological aspects -- Periodicals; 551.51
• DOI: 10.1016/j.atmosenv.2017.12.005
• Languages: English
• ISSNs: 1352-2310

Abstract

Abstract: Webcams and automated, color photography cameras have been routinely operated in many U.S. national parks and other federal lands as far back as 1988, with a general goal of meeting interpretive needs within the public lands system and communicating effects of haze on scenic vistas to the general public, policy makers, and scientists. Additionally, it would be desirable to extract quantifiable information from these images to document how visibility conditions change over time and space and to further reflect the effects of haze on a scene, in the form of atmospheric extinction, independent of changing lighting conditions due to time of day, year, or cloud cover. Many studies have demonstrated a link between image indexes and visual range or extinction in urban settings where visibility is significantly degraded and where scenes tend to be gray and devoid of color. In relatively clean, clear atmospheric conditions, clouds and lighting conditions can sometimes affect the image radiance field as much or more than the effects of haze. In addition, over the course of many years, cameras have been replaced many times as technology improved or older systems wore out, and therefore camera image pixel density has changed dramatically. It is shown that gradient operators are very sensitive to image resolution while contrast indexes are not. Furthermore, temporal averaging and time of day restrictions allow for developing quantitative relationships between atmospheric extinction and contrast-type indexes even when image resolution has varied over time. Temporal averaging effectively removes the variability of visibility indexes associated with changing cloud cover and weather conditions, and changes in lighting conditions resulting from sun angle effects are best compensated for by restricting averaging to only certain times of the day. Highlights: Indexes derived from webcam images can be used to estimate atmospheric extinction. In clear atmospheric conditions indexes must be averaged for at least one month. Average image contrast is least sensitive to changes in image resolution. Gradient operators are very sensitive to image resolution.