Airplane photo update
Couldn't help playing a bit with one of the airplane photos from the previous entry. Thepicture below (left) shows a series of vortices trailing off of the upper edge of the island. The vortices were hard to see when the plane was just above them, but they really jumped out when the moved into the sun glint. While you can see them in the raw photo, a little work in Matlab highlights the vortices (image on the right). To get this image, I made a black & white version of the image. I then created a smoothed version (median smoother with a radius of 5 pixels). The smoothed version removes the fine scale features we're interested in, leaving only large-scale continuous features like the island and the sun. I then subtracted the smoothed version from the raw, to create an anomaly. I then squared the anomaly (to make the large differences pop) and multiplied by the sign of the anomaly (so that +/- were retained). Finally, I replaced values close to zero (absolute value <5) with nans and then overlayed on to the photo. So, the vortices highlighted in red are reflecting more light than expected (sea surface angled towards the camera or smoother water), and those in blue are reflecting less light than expected (rougher water?).