Improvements in forward scatter-type visibility sensors in the past decade have resulted in a growing interest in automating this important meteorological parameter. Applications for visibility sensing include airport weather systems (synoptic and runway visual range), road weather systems, air quality studies, and fog warning networks.

Visibility sensing differs from many of the typical meteorological parameters because of the complex nature of the measurement. No standard test methods or measurement practices are available for visibility. Other sensors, however, are held to rigorous standards. For example, a temperature sensor is tested to the American Society for Testing and Materials (ASTM) E644-98 Standard Test Methods for Testing Industrial Resistance Thermometers, implemented with ASTM D6176M-97 Standard Practice for Measuring Surface Atmospheric Temperature with Electrical Resistance Temperature Sensors (Metric) and traceable to standards at the National Institute of Science and Technology (NIST).

The lack of industry standards for visibility leads to several problems,including misunderstandings of how to prepare sensor specifications and requirements on the part of the buyer and how to specify accuracy and key parameters on the data sheet on the part of the seller. Being a savvy user of visibility sensors and data begins with understanding what is realistic. This paper will examine several key areas in an attempt to help define the accuracy of visibility sensors. First, a review of the accuracy believed achievable by leading meteorological organizations and the visibility accuracy requirements of several weather systems in use today will be presented. Second, sensor manufacturer statements and claims about their own sensor accuracy will be compared. Third, test data from organizations including the World Meteorological Organization (WMO), Federal Aviation Administration (FAA), and National Weather Service (NWS) will illustrate real world results. Fourth, other factors effecting the measurement including the psycho-physical nature of the measurement by humans and the inherent inaccuracies of the reference sensors will be discussed. Using these four key areas, the author will propose a reasonable level of accuracy you can expect from electro-optical visibility sensors and the need for standardization in their testing and reporting.