James Joyce described it as snot-green. Lord Byron went for plain old dark blue. Homer's was frequently "wine dark".
These literary greats were describing the colour of the sea, and the variation in their prose wasn't just poetic licence.
Our own experiences suggest the colour of the sea and oceans can change markedly according to both time and place, from brilliant turquoises and whitish greens, through ultramarine, navy blues, to dishwater greys and mucky browns.
So why is that so, when we all grow up thinking the sea is blue?
The variation in sea colour, it turns out, is due to both physics and biology.
Colours of a rainbow
Pure water is of course clear. However if it is deep enough, so that light cannot reflect off the sea floor, it appears dark blue. This is largely because of some basic physics.
Human eyes contain cells capable of detecting electromagnetic radiation with wavelengths between around 380-700 nanometres. Within this band, different wavelengths correspond to the different colours we see in a rainbow.
Water molecules are better at absorbing light that arrives in longer wavelengths, meaning the reds, oranges, yellows and greens. This mostly leaves the blues, which have shorter wavelengths. As blue light is less likely to be absorbed, it can penetrate to deeper depths, making deep water look bluer.
Light at a short wavelength is also more likely to be scattered or deflected in different directions, including back out of the water towards our eyes, making the sea appear blue.
However, the purity of sea water varies. Particles suspended within it can increase the scattering of light. Sand and silt carried into the sea from rivers, or kicked up from the seafloor by waves and storms, can affect the colours of coastal waters. And organic detritus such as decayed plant matter - known to scientists as colour dissolved organic matter - can also complicate the picture, by adding greens, yellows or browns.
Phytoplankton absorb electromagnetic radiation in the red and blue parts of the visible light spectrum
That's the physics. But even more important is the biology, because the biggest impact on sea colour is made by tiny organisms called phytoplankton.
Usually smaller than a pinhead, these single-celled algae use green chlorophyll pigments to capture energy from the sun to convert water and carbon dioxide into the organic compounds that make up their bodies. Through this photosynthesis, they are estimated to be responsible for generating about half of the oxygen we breathe.
Crucially, phytoplankton absorb electromagnetic radiation in the red and blue parts of the visible light spectrum, but reflect greens, which explains why seas in which they are thriving appear greener.
Determining the colour of the ocean is more than an aesthetic exercise.
Scientists have been monitoring ocean colours from satellites since 1978. These studies have yielded evocative images, including giant tentacles of blues and greens dancing in swirls around each other.
As beautiful as they are, these images have a greater purpose. They can be used to monitor pollution and phytoplankton.
Phytoplankton can multiply very rapidly in response to changes in their environment, such as temperature changes and sudden changes in nutrient levels. Scientists have shown that their populations can double in one day.
The more phytoplankton are floating about the world's oceans, the more carbon dioxide is sucked from the atmosphere
Because of their place at the base of the marine food web, this can have important knock-on implications across the whole ecosystem. They are the primary food source for zooplankton, tiny animals such as copepods, krill and jellyfish. In turn, zooplankton are eaten by fish, which are eaten by other animals, from scallops and anemones to sharks and whales.
Changes in phytoplankton populations and distributions, and their rates of growth or decline, can also provide scientists with early warnings of environmental changes. The more phytoplankton are floating about the world's oceans, the more carbon dioxide is sucked from the atmosphere.
As carbon dioxide is a key greenhouse gas, the more that is converted into organic matter that sinks to the ocean floor once phytoplankton die, the lower average future temperatures.
Phytoplankton can form so-called red tides
"Because phytoplankton take up carbon dioxide and deliver oxygen, they play an important role in the global carbon cycle," says Venetia Stuart, scientific coordinator at the International Ocean Colour Coordinating Group. "The carbon cycle can determine future CO2 concentrations, so it's information that can be used to help model climate change."
Changes in ocean colour can also signal the onset of a deadly phenomenon known as red tides, or harmful algal blooms.
Some phytoplankton species produce toxins that can kill fish, birds and mammals and cause human illness. In higher concentrations they can form so-called red tides, which are not always red. They also have nothing to do with movement of water, hence why scientists prefer the term harmful algal blooms (HABs).
Sensing the sea
So how do scientists survey the changing colours of the seas and oceans?
The main technique is to use satellites that carry instruments that measure the intensity of visible light coming off the water.
Most sunlight is scattered on the way down to the surface of the sea by particles in the air. What is left is either absorbed or scattered within the water. But about 10% is scattered back up out of the water and into the atmosphere, and potentially in the direction of a satellite, which measures what proportion is in the green or blue spectrum.
Computer algorithms then use this data to estimate how much chlorophyll is in the water below.
The desert regions of the ocean in the northern hemisphere are getting bigger
These surveys began in 1978 with Nasa's experimental Coastal Zone Color Scanner mission. In 1997, Nasa launched the Sea-Viewing Wide Field-of-View (SeaWiFS) sensor on board another satellite, which improved the quality of ocean colour monitoring. Since then the European Space Agency (Esa), India and South Korea have also launched their own sensors.
A new generation of sensors, such as the one due to be launched on Esa's Sentinel-3 satellite later this year, will allow researchers to observe light being bounced back out of the sea in greater detail, and even spot different types of plankton, says David Antoine, head of remote sensing and satellite research at Curtin University in Perth, Australia.
For example, scientists have worked out how to spot phytoplankton groups called coccolithophores and diatoms. "It is obviously more useful to be able to distinguish between the different phytoplankton types, as each of them play different functional roles in the ecosystem," says Stuart.
Surveying the colour of the seas and oceans has also yielded even more significant results.
Last year, US researchers published a study showing how levels of chlorophyll in the oceans had changed around the world between 1998 and 2012.
Scientists may need to monitor the colour of the seas and oceans for more than 40 years
There was no overall trend. But the changing hues picked up by satellites suggested chlorophyll levels fell in some northern hemisphere oceans, and increased in some ocean basins in the southern hemisphere.
That has led some to suggest that marine zones with especially low chlorophyll levels, sometimes known as "ocean deserts", are expanding as a result of rising sea temperatures.
"The desert regions of the ocean in the northern hemisphere are getting bigger, which is a concern," says Stuart. "This has been verified with data from other sensors, so there's definitely something going on."
Others believe not enough data has been collected to prove global warming is affecting levels of phytoplankton in the seas, which may naturally vary over cycles of 15 years or more.
Some studies even suggest that scientists need to monitor the colour of the seas and oceans for more than 40 years to determine if climate change is having an impact on phytoplankton. And that could mean waiting until 2038 to get results based on high-quality surveys.
Only then will we really know if the colour of the oceans has changed, and to what degree. And from that infer whether humans are impacting how much phytoplankton is there, influencing the global carbon cycle.