FBB

Hola,

Back to posting.

Since we decided to play a tiny bit with the fraction of breaking waves code, I dug out our old data (the one with soap in the tank), synchronized it with our useful data (as to have the same length and almost the same starting time) and run the code on it.

The following plot is the fraction of breaking waves comparison between the clean data (blue lines, Runs with numbers above twenty) and the foam data (two runs per condition (red and green, with numbers below 10). Results are generally in good agreement, specially offshore. The clean data shows more waves breaking near the shoreline, for reasons that I explain later. Here is the figure



To understand the behavior, take a look at the following images. The first run is of the set associated to Run 25, where the agreement is generally good. It can be seen that the larger difference is near the shoreline, but other than that the lines almost overlap.



A more complex situation occurred for Run 27, where its twin Run 07 had a ton of soap on the tank, as shown below:



The extra amount of soap (and hence persisting foam) is clear, and in consequence the counting is messed up, typically underpredicting.

This behavior is due to the same situation as in the shoreline observations, where the signal is overall very bright and there is a large amount of persistent foam, which induces a very low contrast ratio between the breaking and non breaking part of the signal. In consequence, is hard to identify all the fronts and quite often two or more fronts coalesce into one, and reduces the counting.

In addition, a huge amount of soap is being transported forward by the waves, in regions where the roller did not seem to be present for the clear case. In consequence, a larger fraction of events is counted here for the very foamy case. This is an interesting situation, because it would indicate mass transport by a roller that is not sufficiently bright to be identified on the clean case. Perhaps it corresponds to the case of a roller confined to the crest of the wave, with limited spatial extent.

However, for moderate amounts of soap (hence foam), the code does not seem to be fooled by the orbital foam, which is a good thing.

Now, the following step is to run the code in the Mini data.