Koen,
At a particular period - around the time of Matthew Paris, more or less - a convention for which there is no linguistic or cultural foundation among the Latins begins to appear in certain types of imagery. Some types only.
In maps, for example, we find that a copper-green distinguishes the outer ocean from the deep blue of the sky, and that the Red Sea is coloured red.
An interesting example, whose derivation from an Egyptian precedent is not widely known, is found in the Higden Map. (1350)
A Latin precedent occurs (with less emphasis on the green) in the Paris Psalter map (1260s)
The Anglo-Saxon world map uses green in a different way (1025-1050)
and if you go further back you find no use of green to signify water - salt or otherwise.
So the introduction of a non-indigenous custom of representing salt water by using green appears at a defineable time in the west, but ... I won't go into huge amounts of detail ... it finds its original lineage in the eastern Mediterranean, and there too we find an informing habit of speaking about seas and oceans as green.
The general opinion, then, is that this habit natural in the eastern Mediterranean, and having a linguistic basis which pre-dates even the Hellenistic Greeks and survives through the Islamic era, and occurs also in Persian, came to the west with materials gained during the crusading era, and very possibly - the question remains debated - directly as a result of works commissioned in the holy land or in Egypt, or perhaps in Anatolia by the occupying Latin forces. There's also the Byzantine question, but that's another still being explored by scholars.
In addition to that introduction of the non-indigenous custom (which, by the way, didn't last long because it ran contrary to the practice and languages of the west), there is also the type of line used over the pigment.
Now, in the eastern Mediterranean in earlier times, and due to Egyptian set practice, water was denoted by parallel lines of even zig-zags. We see this echoed in some earlier medieval Christian works, especially e.g. the mosaic at Bobbio.
However, the practice which developed in medieval Latin works was, as a rule, to denote bodies of water with parallel gently-waved parallel lines and when the temporary fad for green water passed, this remained constant and is seen over the blue colour natural to western ways of speaking and thinking. It is only to be expected, because the intense copper-greens which occur elsewhere - even around Greece - are rarer in the more turbulent seas and colder climates of the west.
So, when we find green water used in copies of the 'Balneis' it doesn't tell us that 'green water is not unusual in Latin works' at all.
What it tells us is the time when the illustrations were made for a particular copy - a textual tradition's origin is not necessarily contemporaneous with an iconographic tradition. That's one of those things that's self-evident when you think about it, but which is rarely thought about at all.
As you move down the time-line of copies of the Balneis, you see the shift occurs also in the tone and colour of the pigment used for the water, and the way the overlaid lines are formed.
But the bottom line is that using green for water is a non-Latin practice, which for a time was adopted into Latin art. It certainly cannot be used to suggest that because some of the waters in the Vms are coloured blue, and others are green that we can presume the imagery itself originated in Europe. Indeed, the opposite is indicated in my opinion, because it shows knowledge of a distinction which is entirely in keeping with the eastern tradition by which the 'Great Green' (to translate the original Egyptian term' was the salt water, and the blue signified fresh - drinkable, river- water.
It is certainly a reasonable argument that, in a work such as the Vms which is thought (not known) to have been made in Latin Europe early in the fifteenth century, then
IF its imagery (not text) were
known to derive from wholly Latin exemplars (which is not known and rarely even investigated), then
those exemplars are very probably ones that we may date to the thirteenth or fourteenth centuries, when the fashion for green water was strongest.
What cannot be done is to wave away all other possibilities - including the obvious possibility that the bathy- section's imagery has been gained from elsewhere - on the grounds that if Latins did it, then it was a Latin cultural possession, and 'not unusual'.
It is unusual for western Christian art (notice how narrow the range of compaisons offered from Latin works); it is certainly not an indigenous practice there, whereas outside Europe language and cultural traditions made it natural and a long-standing practice.
It's one of those 'roses are flowers, but not all flowers are roses' situations.
If anyone's interested.