For the first few casts the .con files had the reversal of V0 and V1, the above con is correct.
The errant points on the oxygen graphs are artifacts of processing the two deep casts 66 and 20. The asc and csv files are correct, didn't see the necessity to go back and redo the plots. The regression cant really fit much better. Cut out BTL_CHL values above 1.5 to improve fit.