Project update: initial images analyzed

Folks, here's a short update on where things stand with regard to the image analysis. Rudi pulled some of the SAR images that corresponded with the location and dates for which Matthew had given us photos of ice road crossings. The logic of this was that we know the road crossings are there and the photos were taken after snow melt and breakup, so that the crossing stands out, visually. We should be able to see them on the SAR imagery.

Well, yes and no. Upon initial inspection, we only found one of the crossings in the SAR image. The crossing appeared to be approximately the size of a single pixel, which would make automatic detection very difficult. We also aren't entirely sure if what we see in the image is grounded ice below the water surface (i.e. a significant amount of water is flowing over a deep ice dam) or whether it's an ice bridge that hasn't broken up at all.

Another thing was clear. From SAR, breakup may actually be a difficult time to detect the position of the ice bridge because both snow and soil are saturated with liquid water. While the two look very different visually, from SAR they look somewhat similar because the sensor is simply reflecting it's beam off the liquid water.

We decided to back up a few weeks. It may be easier to detect the position of the road and the bridge when the hard ice differs considerably from the surrounding snow in terms of reflectivity. The platform passes over the same exact spot every 24 days. We are going to look backwards from breakup every 24 days to see if we can get a strong fix on the road location. Then we'll dial back into the mosaics at the daily time step to see how the scene evolves during the snowmelt and breakup periods. This should give us a better sense of what we can see and what we can't see and if we can use this to parameterize a model for the ice melt on the tundra versus in the river.