With the data you provided, I understand that you have data of different dimensions (In this case different depth values and the number stations).
The data has severalNaNvalues which needs to be handled. A possible work around would be extracting each row and use thermmissing to remove the missing data at each station for all the 3 datasets.
Then you can either use ainterp1 to interpolate all the 3 datasets depths values at aparticular stationto the desired number of depth values. Or truncated the depth values to the minimum of all the 3 datasets at aparticular dataset. Then use the scatter plot to plot the data