Aerially observed Sea Surface Temperature data during the Lagrangian Submesoscale Experiment (LASER) from 2016-01-20 to 2016-01-30

A Parthanavia P86 dual engine aircraft based in Gulf Shores, Alabama carried a FLIR A655sc long-wave infrared (LWIR) camera used to make maps of sea surface temperature (SST). The dataset consists of composite maps of SST from individual images georectified on a common grid with a resolution of 10m. Typically, there are between 1 and 4 SST maps per 24-hour, ranging in domain size from 5 km x 5 km to 50 km x 50 km. The LWIR camera uses an uncooled microbolometer technology and it is sensitive in the range 7.5-14 micrometers. Since the camera is uncooled, a Heitronics KT19II radiometer was used to correct for drift. Under ideal conditions, this combination can achieve an accuracy of 0.05 degrees C. Using a lens with a 10 mm focal length, the field of view is 33 x 45 degrees and the sensor grid 640 x 480. At a typical flight height of 3000 m, the thermal images cover an area of approximately 3000 m by 2250 m. Using a 640 x 480 sensor array, the camera had a spatial resolution of 5 m. The images were directly georectified using the position and altitude data from an onboard Applanix POS AV310. The POSAV 310, when post-processed using RTX data provides position data with an accuracy of approximately 5 cm, altitude data with an accuracy of 0.15 degrees for roll and pitch, and 0.3 degrees for heading. This allows to georectify each pixel in the thermal images with a maximum horizontal error of a few meters. The aircraft was flown in a 'lawnmower' pattern, obtaining 2 thermal images every second. All rectified thermal images were combined in a larger mosaic that typically spanned an area of approximately 50 x 50 km. It takes about 4 hours of flight time to complete a survey of such an area. Due to the considerable overlap of the thermal images, it was possible to average about 100 observations for each 5 x 5 m bin of the resulting mosaic. Each individual image of the LWIR camera is characterized by approximately 0.07 degrees C noise; the averaging reduces the noise by one order of magnitude. Since the atmosphere is not fully transparent for radiation in the LWIR range, false SST gradients may appear due to atmospheric effects which depend on the humidity and temperature profile of the atmosphere. An attempt has been made to remove these effects by analyzing apparent temperature dependencies as a function of radiation path lengths through the atmosphere. It is important to note that the LWIR camera observes the radiative skin temperature of the sea surface. The differences between the skin temperature and the bulk temperature as measured in-situ by a vessel may be as large as 1 degree. However, the SST patterns provide a unique synoptic context for ship-based measurements and drifter motion that, as of this moment, can't be duplicated by any other means.