LiA Week 4 - Pom Pom

Continuation of work dives and further development of turtle identification software.

Hi Laidlaw people,

This past week has been much of the same, with a few new developments! 

Work dives have continued along with the same rhythm, and I've found I much prefer epoxy as a substitute for glue during coral planting - it seems to be much more efficient and more easy to work with, although I don't know about its long-term viability. Unfortunately, during a run I found a dead turtle which had washed up onto the beach. Poaching is quite prevalent around the area, so whenever this occurs it's important to check for key indicators of this, such as missing flippers. Luckily, this time it seemed to have died of natural causes, so I had to take a picture of its facial scutes (scales) to send to the science team so that we may wipe it from the database. 

On Sunday, I was lucky enough to visit Sipadan, one of the best dive sites in the world. This was a great showcase on how to properly preserve a marine ecosystem, especially one so frequently visited. For example, by the end of 2004 all resorts on the island were forced to close, so divers are only allowed to visit for the day; the island is guarded by the military wing of the Malaysian police, which protects the island from fish bombing and any forms of poaching; tourists must get a relatively expensive permit (200€) to dive there for a maximum of two dives and be of at least Advanced Open Water certification, ensuring those who do get to visit are divers of a higher calibre who won't touch or damage coral. Sipadan's blooming biodiversity stems from the fact it is the only oceanic island in Malaysia, meaning its corals sit on volcanic rock which rises 600m from the deep seafloor around the island - this is unlike Pom Pom which sits on a less deep continental shelf. This unique topology attracts many pelagic fish to the island, which is why conservation of this island has been a top priority for the Malaysian government. 

Recklessly, during my second dive at Sipadan my buddy and I went past our no decompression limits (NDLs), which is basically what happens when your dive computer calculates that you've spent too much time deep underwater, and so to rid yourself of the nitrogen bubble build-up this causes, which could cause decompression sickness, you have to spend some extra time at shallower waters where the nitrogen can be purged from the body. Honestly, this is quite a rookie mistake to make as a rescue diver, but I guess we got so engrossed during the dive that we lost track. This a good lesson on temperance, where if we had been a little more cautious and not so engrossed on the marine life during the dive, we could've avoided this mistake.

Finally, I've made some progress on the turtle reidentification software using a deep neural network, to the point where it's reidentifying 100 turtles with a 97% accuracy. I have a few more days here, so I may present this to the science team if I get the chance.

This will be my last blog from Pom Pom, so I leave you with some cool pictures from Sipadan!