As in the broader marine science and remote sensing communities, you will find that the marine remote sensing community uses a substantial number of different programming languages. Often the language used will be related to that used by the institutions/supervisors etc who taught the person. There is however a general trend towards the use of more open source languages (since these do not require paid for licenses, and code can be more widely shared). There are many advantages and disadvantages to each programming language and often the best choice is the one you already know/have support for, given the investment needed to learn a new language. However, if you are able to select/learn a new language, below is some advice to consider depending on your application.
Open source, flexible, quick and easy to learn (compared to C etc), huge amounts of support and existing code online, integrates with SNAP, GDAL, QGIS, ArcGIS. Widely used for many applications, so a good general choice, you will probably only need lower level languages (C etc) if scaling up to intensive operational processing/integrating with modelling work.
Open source, great pre-existing packages and community support, very popular for statistics and in biological sciences.
Matlab (octave) and IDL
Commonly taught and used in marine science/remote sensing. Needs a license (can be costly for distributed processing).
C and Fortran
Very fast! But steep learning curve. SeaDAS written in C.
SNAP is written in this, Jpy interface works via the Snappy python module.