It is all so cloudy ...

People often wonder why i get so worked up about clouds. Why they are the bane of my life.

I like them, don’t get me wrong. Most of my research time has been spent on clouds. I don’t mean i a way that will tell you (or the world) anything about clouds they don’t already know either. I mean that i spend a lot of my time trying to write programs that identify/remove clouds from my satellite data. Not such a problem surely?

Well let me tell you something about polar clouds. Imagine your looking down on the Antarctic from high above. You can see land ice (the continent), fast ice, sea ice and clouds....imagine what colour they are? Sure after a bit of training a human can usually figure out which bits are what. A computer is not so lucky. A human brain is amazing. The way we can make assumptions, gather information and make decisions on the fly is astounding.

I have trained computer code to do the same thing and let me tell you they really aren’t that bright. So literal. Yes. I am being facetious. I have good programs now but they are not 100% and it took a long time to move past that and concentrate on the end goals. The thing is that i have thousands and thousands of satellite images and using a computer to do most of the work is the only way to do it. It also removes the human element. I know what i said previously but humans tend to make mistakes and have bias. Besides i once had to look through a lot of imagery for a book i wrote with my supervisor and it wasn’t as much fun as you think. Not even a little bit.

Now this is not as much of a problem really. Eventually i figure out that i was only really interested in the open water areas with the sea ice, mostly along the coast. I know where the coastline is and fortunately water is significantly different. Besides it was about looking for one variable rather than trying to identify multiple variables. But that is how you learn. Sometimes you take a few wrong turns. But you always learn something in the end.

Now take a look at the following sat image (AVHRR thermal band 4) and see if you can spot the cloudy bits without using the map!

pink and orange are cloud, blues are ice/water and green is land [yes. i know that land in Antarctica is not green. in anyway. it is my little attempt at humour.]