Monday, 8 January 2018
Exhibit Hall 3 (ACC) (Austin, Texas)
General aviation (GA) operations can be significantly impacted by low ceilings and visibilities
(C&V). Flights into unexpected instrument conditions can lead to costly diversions and is a
leading cause of accidents for pilots that are unprepared or lack appropriate training. In remote
regions, such as Alaska, it can be difficult for the aviation community to obtain accurate current
or forecast conditions of the ceiling and/or visibility. In response, the FAA has deployed a large
set of rapid update cameras across Alaska that provides pilots with a live view of the current
conditions via an FAA maintained website. However, the imagery from these cameras can only
be interpreted manually with internet access, and the data they provide are not integrated into the
current suite of automated C&V forecasting products.
This paper will discuss algorithms that have been developed by Lincoln Laboratory to
automatically translate camera imagery into calibrated C&V metrics. The focus will be on
estimating the visibility in Alaska where low-light and low sun angle conditions make a
challenging environment in which to extract visibility estimates. Also, the frequency and
duration of low visibility conditions in some of the remote Alaska mountain passes require
special considerations in the algorithms that typically require periodic clear weather conditions to
aid in calibration.
This paper will also discuss a proof-of-concept evaluation conducted during the summer of 2017
in collaboration with the FAA Alaska Aviation Weather Camera (AvCams) program office for
display of visibility estimates on the FAA camera imagery website. In the future, providing
automated rapid updates of changing C&V conditions to GA pilots may allow for better
situational awareness and decrease incidents.
(C&V). Flights into unexpected instrument conditions can lead to costly diversions and is a
leading cause of accidents for pilots that are unprepared or lack appropriate training. In remote
regions, such as Alaska, it can be difficult for the aviation community to obtain accurate current
or forecast conditions of the ceiling and/or visibility. In response, the FAA has deployed a large
set of rapid update cameras across Alaska that provides pilots with a live view of the current
conditions via an FAA maintained website. However, the imagery from these cameras can only
be interpreted manually with internet access, and the data they provide are not integrated into the
current suite of automated C&V forecasting products.
This paper will discuss algorithms that have been developed by Lincoln Laboratory to
automatically translate camera imagery into calibrated C&V metrics. The focus will be on
estimating the visibility in Alaska where low-light and low sun angle conditions make a
challenging environment in which to extract visibility estimates. Also, the frequency and
duration of low visibility conditions in some of the remote Alaska mountain passes require
special considerations in the algorithms that typically require periodic clear weather conditions to
aid in calibration.
This paper will also discuss a proof-of-concept evaluation conducted during the summer of 2017
in collaboration with the FAA Alaska Aviation Weather Camera (AvCams) program office for
display of visibility estimates on the FAA camera imagery website. In the future, providing
automated rapid updates of changing C&V conditions to GA pilots may allow for better
situational awareness and decrease incidents.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner