Why are Lakes and Rivers in the Canadian Rocky Mountains so Brilliantly Turquoise Blue?

An investigation into geology, history, and optics

I've always wondered what exactly is going on with these lakes and rivers. I was surprised to find so little written online to explain their colour. I did find a lot of people asking the same question, however. There are a few good answers, ranging from simple to very complex. I relied on them for my answer and I direct you to them here:

1) Ben Gadd, an Earth Science specialist and licensed interpretive guide, resides in Canmore, Canada. He offers an excellent not-too-technical treatment of the question in his book called Handbook of the Canadian Rockies. The geology in this book is challenging but really good if you want to know how the Canadian Rocky Mountains came to be (we call them simply The Rockies here). You can sense his passion for this gorgeous natural part of Canada.

2) An online article called Glacial Lake Colour – Get The Right Story. Nadine Fletcher, a mountain guide, offers the perfect introduction to glacial lake colour. She also mentions an out-of-print book as her source in the article, if you are lucky enough to find a copy.

3) If you want to go deep, you will find the 1988 scientific paper called Colours Of Glacial Water by Eyvind Aas and Jim Bogen (you have to pay to access it online) to be probably the best scientific answer available. It's not an easy read but physicists will appreciate it.

4) Wikipedia has a few good but brief entries to check out. Try Color of Water, Rock Flour and Glacial Lake.

5) Hyperphysics.com does a great job of explaining the concepts of light scattering. Usually applied to atmospheric optics, these are also the basic scientific principles behind glacial lake and river colour. As you will discover in this article, the answer takes you into a lesson about optics.

Introduction

Numerous lakes and rivers in the Rocky Mountains of Western Canada (my favourite stomping ground) are an otherworldly turquoise-blue hue. It's an opaque yet brilliantly bright colour that looks as if someone is continuously dumping huge vats of food colouring upstream. It is as if someone has painted the riverbeds and lake bottoms turquoise. And they are so luminescent at times that they appear as if they are under-lit as well. It is equally mysterious that rivers, which are obviously shallower than the lakes (you can see man-sized boulders peaking out of them), often appear just as brightly turquoise as the lakes, many of which are hundreds of metres deep.

A great example of this colour is Peyto Lake in Banff National Park (below) seen from a viewpoint on the Icefields Parkway. Once you see this colour yourself, you never forget it.

Tobias Alt; Wikipedia

You will likely see this particular turquoise only in the Rocky Mountains of Western Canada. These lakes and rivers are glacially fed bodies of water. There are many others around the world as well, but each glacial source tends to be unique. This means that the colours of the waters that melt from them likewise tend to be unique.

For example, Tasman Lake, a glacial water-fed lake in New Zealand, is muddy grey-green, as seen below.

It's a very young lake (around 25 years old) and it formed as ponds of glacial meltwater merged from fast-retreating Tasman Glacier. A submerged glacial ice apron projects into the lake (not visible) and icebergs periodically break off it to float across the lake.

Green, from my research, seems to be the more common glacial lake colour. A good example of a green glacial lake is the unnamed proglacial lake made of the meltwater from Schoolroom Glacier in Wyoming, US, seen below.

The water is held in a cradle of rocks and rock debris (called a moraine) left behind by the retreating glacier) (you can see the edge of the glacier in the upper right corner). Lakes formed by the damming action of moraines, or from ice dams, are called proglacial lakes. Both Tasman Lake and this unnamed lake are proglacial lakes.

Explanations for Rocky Mountain Lake Turquoise: A Stewpot of Anecdote and Science

I have heard several anecdotal explanations for the turquoise colour of our Rocky Mountain glacial lakes and rivers. One explanation is that the water surface reflects the blue of the sky. Frankly this is the laziest explanation. This is true, but it is just one small contribution to the colour, and it's blue not turquoise. If you look at the sky's reflection in the Peyto Lake photo above, you can see that the blue sky's reflection is a true blue colour on the lake, different from its more greenish turquoise inherent colour. The reflection of sunlight on lake water can make all lakes look blue at least at certain times of day on sunny days when the surface is calm and reflective. The luminescent turquoise of glacial lakes appears much different, and it doesn't require an intensely sunny day, although it appears more brilliant on bright sunny days. On stormy early summer afternoons, and there are many in the Rockies, lakes turn dark and foreboding and glacial rivers in particular seem to take on an almost eerie milky light blue.

Besides the reflection "answer," it is true that any water in general looks blue en masse. The colour comes from the water molecules themselves. The molecules tend to absorb the red, orange, yellow and green wavelengths within the full spectrum of white sunlight striking them, while they reflect the shorter blue wavelengths back to your eyes. This is called Rayleigh scattering, more on this mechanism in a moment. The water in an indoor pool that is lit by white ceiling lights but not sunlight looks blue for this reason. The deeper the water is, the bluer it looks (there are more molecules reflecting blue light).

An additional source of lake colour can come from suspended organic material, which most lakes have. This includes all the living organisms suspended in the water column. These organisms will reflect the sunlight that strikes them back upwards. The sunlight must pass through water of varying depths to reach them first, so this means that the light that strikes these particles is mostly blue light. Some organisms, especially diatoms, act like little mirrors. The light they reflect upward to your eyes is also blue. Algae often have their own intrinsic colour, and at high enough concentration (an algal bloom) they will change the colour of a lake to deep green, blood red or even bright pink depending on the species.

Shown below,  an intense cyanobacterial bloom turned Lake Erie turquoise and green in 2011. These organisms are technically bacteria, not algae (a unicellular plant).

This pink bloom is caused by a algal species of dinoflagellate. Some dinoflagellates are bioluminescent. They glow eerie blue in the water at night.

Marufish – Flickr;Wikipedia

Algae are not at play in glacial turquoise lakes, however. Glacial lakes and rivers are cold and very low in organic matter, although super-tough Watermelon algae, tiny ice worms and ice springtails can be present in glaciers. Is glacier water safe to drink? If the ice has a pink tinge don't drink from a glacier's snout as it has Watermelon algae, which is said to be a laxative.  I couldn't find any evidence that the animals are toxic if ingested. Is the rock flour itself okay to ingest? Glacial snout water will clog up a coffee filter in no time, and consider too that this stuff acts like fine sandpaper. A laxative sandpaper drink wouldn't be my choice.

Besides suspended organisms, chemicals and ions dissolved in water can colour it. The natural presence of copper was once an explanation for the turquoise colour of the glacial water. Decades ago as a teenager I was told this "fact." Perhaps folklore lingering from the late 1800's mining rush in the Canadian Rockies explains why it was natural to point to copper as the reason behind turquoise water. Copper can exist in rock as pure (insoluble) metallic ore and it can also exist in soluble compound form such as copper sulphate, copper chloride, etc. Many copper compounds are blue-green or bright blue when the copper ions (usually Cu⁺⁺) dissolve in water. Copper ions are naturally found in lake and river water in the Rocky Mountains, but their concentration is generally at or just over the current detection limit of 0.001 mg/L. This concentration number comes from a study done along Athabasca River (a Rocky Mountain glacial river) in 1995. The concentration is lowest upstream near its glacial source and increases downstream because of contributions from human development. The tiny amount of copper dissolved in glacial water comes from the weathering of mountain rock, which contains trace amounts. It is not high enough to colour the water a detectable amount.

The Canadian Rocky Mountains do (or did) contain some minable solid copper ore deposits. Quartz-carbonate veins in the limestone and dolostone rock of the mountains contain copper. These veins are found in fractures in the rock, where the copper-bearing quartz-carbonate forced its way up from deep underground. The fractures come from faults and mechanical shear zones in the sedimentary rocks that formed as they were folded and deformed during mountain building.

Composition of the Canadian Rocky Mountains

These geology terms might be new for many readers so I offer here a brief amateur lesson in Rocky Mountain geology (I can mostly thank Ben Gadd's book for my understanding): The Rocky Mountains are made of sedimentary rock. This rock was slowly deposited over time as layers of sediment built up on the floor of a shallow ancient inland sea. Later, as the plates of Earth's crust separated, shifted and collided, those plates pushed against this rock, folding it into mountains, like bunching up a rug on a floor. Limestone (CaCo₃) and dolostone (CaMg(Co₃)₂) are sedimentary rocks composed of calcium carbonate. Dolostone is similar in chemical composition to limestone except that it has the addition of magnesium (Mg). These rocks, along with two other sedimentary rocks, sandstone (formed when sand grains cement together over time) and shale (formed from much finer grained silt and clay), make up almost all of the Rocky Mountains. Quartz (silicon dioxide or SiO₂), part of the copper bearing quartz-carbonate veins mentioned earlier, is a very common mineral that either crystallizes from magma or precipitates from very hot hydrothermal veins. Quartz, hard and very resistant to weathering, is commonly found as crystals in sedimentary rocks such as sandstone, shale and in carbonate rocks such as limestone and dolostone. Beach sand itself is composed mostly of granules of quartz and calcium carbonate.

Key Differences Between the Canadian and American Rocky Mountains

Fracture zones filled with copper-rich quartz/carbonate veins are not common in the Rocky Mountains, at least not near the surface. However, some were accessible enough in the far northern part of the Rockies to be minable. There was once a productive copper mine (Churchill Copper Mine, discovered in 1943 and most productive between 1970 and 1975) in that region.

The copper rush (or more accurately a silver/gold/copper rush) took place earlier, ramping up in the late 1800's in the Canadian Rockies. Copper Mountain, a popular ski resort in the Rocky Mountains, was named after a prospector that suspected vast copper ore was present within its rock (it wasn't). There was also a frenzied copper rush at Silver City located at the foot of Castle Mountain (along with wild rumours of gold and silver, hence the name). There were minute amounts of copper and silver ores found there but not enough to be commercially viable. Just as quickly abandoned as it formed, only a marker now indicates the area where Silver City once stood.

The mining hysteria originated in the American Rocky Mountains, which extend southward from our border. Prospectors found a wide variety of valuable minerals and metals in their American extension of the North American Cordillera. It was natural to expect more of the same north of the border. Unknown at the time, however, there is a key difference between the two regions. The Canadian Rocky Mountains are composed of layered sedimentary rock that underwent mild metamorphism, while the American extension of Rockies is mostly of metamorphic and igneous origin (this is rock such as gneiss (metamorphic) and granite (igneous).

Rock of deep volcanic origin (igneous) and rock formed under tremendous heat and pressure (metamorphic) tend to be where gems are formed and where metallic ores and precious metals are deposited. The American part of the Rocky Mountains was reformed from a mountain range that existed 300 million years ago. Unlike the present American Rocky Mountains, this ancient range was built by volcanic activity. In contrast, the Rockies to the north, in Canada, were formed from sedimentary rock laid down at the bottom of a large shallow sea that existed around that same time and earlier. Both of these types of rock bunched up into the Rocky Mountains around 80 million years ago, as the Pacific continental plate pushed into the North American continental plate.

Sites in the Canadian Rockies are rich in rare Cambrian marine fossils. These creatures lived around 500 million years ago, long before any plants or animals lived on dry land. As animals died their bodies were covered in fine silt (the raw material for shale rock), which prevented oxygen from seeping into them and decaying them. For this reason even some soft body tissues are beautifully preserved, offering one of the best windows in the world into our first ancient multicellular creatures. The fossils are still intact after all this time because the Cambrian sedimentary rock in this zone, called the Stephen Formation, was protected by an adjacent ancient cliff of hard-to-compress limestone (called the Cathedral Formation) nearby. This means that the fossil-rich Stephen Formation was never heated or squeezed intensely during mountain-building later on. Besides the beautiful glacial lakes and rivers, another reason to come and visit our Rocky Mountains is to take a guided Burgess Shale hike in Yoho National Park to see them for yourself. The Burgess Shale formation is a world-famous UNESCO site.

This Geology Lesson Leads To Rock Flour, Our Secret Ingredient

In terms of our turquoise colour investigation, we can now rule out organic matter, dissolved copper, and the sky's reflection as the source. This leaves a perhaps unexpected or perhaps inevitable source: rock flour. If you have never heard of rock flour you are not alone. Rock flour, or glacial flour, consists of very fine-grained particles of rock. Like grains of wheat ground under a grindstone mill, the surfaces of mountains were ground fine as enormously heavy rock-hard glaciers crept slowly along the surface during the last ice age (yes glacial ice is a fluid – it will flow very slowly downhill due to gravity). Rock flour particles are tiny and come in all shapes. Some can be rough-edged. A layer of these particles underneath a heavy creeping glacier thoroughly sands down the landscape.

Rock flour was deposited throughout the Rockies. You can see it mixed in with larger rock debris at the snouts of glaciers today. The entire North American Cordillera was glaciated in the past. Extremely thick continuous sheets of ice covered almost all of Canada during the last interglacial period, while the southern American Rockies were, and some still are, covered with local glaciers. The Big Pine glacial lakes, for example, exist as far south as southern California. These lakes located in the Sierra Nevada exist at the southern end of the Rocky Mountains. They exhibit a lovely emerald green colour thanks to their rock flour content.

Rock dust, as you can guess, is composed of the mountain's rock, so here in Canada it is rich in limestone, dolostone, sandstone and shale. Most of the particles are so small that they can stay in suspension in glacial meltwater for a long time. The water very close to the snout (leading edge) of a glacier can be so heavily loaded with rock flour that it appears milky white or milky light grey. In rivers downstream and in glacial lakes, however, suspended rock flour magically turns the water luminous turquoise. This is where the science of optics comes in.

Rock Flour Optics

Colour is all about visible light. Visible light is a small portion of the electromagnetic (EM) spectrum of energy that humans can see. The energy of EM radiation increases as wavelength decreases. Visible light wavelengths range from 380 nm (violet) to 750 nm (red). Green, in the middle of the optical spectrum, is around 540 nm (these wavelengths are in the nanometre or billionths of a metre range). This screenshot of Wikipedia's entry will help you compare colour wavelengths as you read on.

The Intrinsic Blue Colour of Water: Rayleigh Scattering

I snuck in a little intro to light scattering, called Rayleigh scattering, earlier on. Rayleigh scattering contributes blue colour to pure water. Now we can fully explore how it works. This contribution has to do with how the water molecule vibrates in its liquid state. If you wiki Rayleigh scattering you find it is defined as "the (dominantly) elastic scattering of light or other electromagnetic radiation by particles much smaller than the wavelength of the radiation." This is a good working definition we can expand upon.

Individual atoms and molecules are the right size to produce this phenomenon. They have a diameter smaller than one tenth of an optical wavelength. When sunlight strikes a molecule of water, for example, it doesn't change its physical state but it does act on the electrical charges in the molecule (the charged electrons). Visible light is EM radiation. It is a traveling oscillating electric field. This oscillating field acts on the molecule's charges, making them oscillate at the same frequency. In effect it turns the molecule into a tiny EM radiating dipole. It is now a tiny oscillating electric field, so it radiates its own EM radiation as scattered visible light. We don't need to know a lot of complex math to understand this but the following relationship is key: The amplitude of the scattered light is proportional to the inverse square of its wavelength. This means that small visible wavelengths (blue wavelengths) will produce the most intense scattered light. The scattering process itself, in other words, is most effective at short wavelengths, or toward the blue end of the visible spectrum. As an example, the scattering of sunlight off air molecules at 400 nm is almost 10 times greater than it is at 700 nm. This is why the sky, full of small nitrogen, oxygen and other molecules, looks blue when sunlight shines through it.

In our case we are examining Rayleigh scattering in a liquid. A water molecule (H₂O) never sits still. In technical terms it is a tiny harmonic oscillator. It can vibrate in three different ways. It can vibrate along each of its two oxygen hydrogen (OH) bonds, and the molecule itself can rotate back and forth. However, it can only rotate when it is free in the gas state as water vapour (this rotation is called a v₂ vibration). As a liquid in our case, molecules are too close together to rotate freely. This leaves OH stretching oscillations (acting like tiny vibrating elastic bands) and there are two kinds possible, shown below left. Blue spheres represent hydrogen atoms and yellow sphere represents oxygen. The bonds are shown as black lines (each black/striated triangle pair is a non-bonding lone electron pair).

Each of these two fundamental modes of vibration has a frequency. V₁ (top) is 3650 cm^-1 and v₃ (bottom) is 3755 cm^-1. The absorption of EM radiation at these vibrations is in the invisible infrared part of the spectrum. The strongest reflected wavelength (the true colour of pure water if we could see it) is ultraviolet, which is also invisible. This is the fundamental mode of vibration of those two OH bonds. None of this contributes to the visible colour of pure water. This reflected UV radiation, by the way, is not the shorter wavelength high energy UV solar radiation that causes skin cancer and cataracts.

So why is water blue then? The water molecule not only has fundamental modes of vibration, it has harmonic modes as well (just as in music). A significant harmonic mode is v₁ + 3v₃, which equals 14,318cm^-1. This frequency corresponds to a wavelength of 698 nm, which is visible red. The water molecule absorbs red, so the colours it reflects are disproportionally on the blue end of the visible spectrum. Water therefore has an intrinsic blue colour.

Difference Between Rayleigh and Mie Scattering

When light strikes any molecule, some wavelengths are absorbed and others scatter off it. This is a size-dependent phenomenon that operates only on very small size objects such as molecules. It works for particle sizes up to one tenth the wavelength of the light. Much larger objects such as water droplets in clouds make clouds look white. Light scattering for them is mostly independent of wavelength in the visible range. Incoming white sunlight scatters off them as white light. The mechanism responsible here is Mie scattering.

For example, look up at a sunny sky when the sunlight is angled from the South (if you're in the northern hemisphere). If you look over to the north, you see the most intense blue colour. If you look more toward the south, near where the Sun is shining, the sky progressively looks whiter. This happens because the contribution of Rayleigh scattering decreases in favour of increased Mie scattering. This effect is more pronounced when there is a lot of particulate matter (larger than atomic size) in the sky. Smoke blown in from forest fires often makes the entire sky look white. To see the difference between Mie scattering and Rayleigh scattering in easy diagram form, click this Hyperphysics link.

Both Mei and Rayleigh scattering are examples of light scattering off a small object such as a molecule (Rayleigh) or dust grain (Mie). In both cases, light scatters elastically, which means that the energy (wavelength or frequency) of the light is not changed during the process. To describe how light interacts with larger objects, you move on to ray optics. These processes include reflection, refraction, and diffraction, where light is described as rays rather than as an oscillating EM wave. Ray optics are also at work in glacial lakes and rivers (the partial reflection of a blue sky in a glacial lake is an example).

Neither Rayleigh Nor Mie Scattering Make Glacial Water Turquoise

Rayleigh scattering does not contribute to the brilliant turquoise colour of glacial water. There is a common misconception that rock flour particle sizes generally match the wavelength of the blue-green light component in the Sun's light spectrum, so they reflect blue-green light like tiny Rayleigh scattering machines in the water. Remember, Rayleigh scattering operates when particle size is, at most, a tenth of visible wavelength size. On the other hand, Mei scattering occurs most predominantly when particle size roughly matches wavelength size. The smallest rock flour particles approach visible wavelength size but many are larger. The finest particles tend to be between 2 um (micrometres, or millionths of a metre) and 100 um. Most rock flour particles are larger than a typical optical wavelength. To compare, red, the longest visible wavelength, comes in at 698 nm, or 0.698 um, under the size of most rock flour particles, but probably not all. For Rayleigh scattering to work, particle size must be less than one tenth the size of a visible wavelength. Mei scattering, however, works in this range. The problem is that Mei scattering reflects (white) sunlight back as white, not turquoise, light. It certainly contributes to the brightness or brilliance of glacial lakes, however.

How Rock Flour Results in Turquoise Water: The Tyndall Effect

To understand the mysterious turquoise colour, we must move on to a different optical explanation called the Tyndall effect, or colloid scattering. As a mixture, a colloidal dispersion finds itself between a solution and a suspension. A solution is a homogenous mixture where a substance is dissolved in a solute. Particles in solution can be atoms, ions or molecules. They are smaller than 1 nm (0.001 um) across. A suspension consists of particles that can be evenly distributed if the contents are shaken or disturbed, but at rest the particles settle out. Medium to coarse rock flour in water (40 - 100 um rock flour particles) is a suspension. Very fine rock flour acts much like a colloid in the water. When it is mixed with water, the particles tend to stay evenly distributed throughout it for some time (but they will eventually settle out too). Colloidal particles tend to be around 1 micrometre (um) in diameter, which is a just bit larger than a visible wavelength.

The Tyndall effect is not a mechanism per se and, unlike Rayleigh and Mie scattering, it does not have a formal mathematical description. Instead, it simply describes how light tends to behave when it passes through particles suspended in a liquid.

This effect is not Rayleigh scattering but, like Rayleigh scattering, it is particle size dependent, and it is wavelength dependent. Longer-wavelengths tend to be transmitted through the colloid (suspension) and shorter wavelengths tend to be reflected by scattering. The underlying mechanism is, in fact, a special form of Mei scattering, which might seem a bit contradictory, as we are getting colour here. Colloidal scattering becomes mathematically equivalent to Mei scattering if the colloidal particles are perfectly spherical in shape. Rock flour particles are all kinds of shapes from roughly spheres to sharper fragments. Confusingly, as we talk about the Tyndall effect, the argument can seem reminiscent of Rayleigh scattering, as the following Tyndall effect analogy suggests. 

To get a feel for the Tyndall effect, think of the difference between (long) radio waves and (short) light waves. Radio waves can pass right through solid materials like walls in buildings because the length of the wave is much larger than the width of the wall. In order to absorb or reflect EM radiation, you need a charge oscillation that corresponds to the EM wavelength. In most materials, the oscillating charges in molecules can match visible EM radiation (clear glass is an exception so it is invisible to light). The radiation is absorbed and reflected (which wavelengths absorb and which wavelengths reflect determines what colour the wall appears). There are no charge oscillations big enough to correspond with radio waves so walls are transparent to them. Some metal walls are an exception. The electrons in metals form a mobile "sea" that can resonate with long radio wavelengths and reflect them.

Don't despair if this theory is confusing. Both Rayleigh and Mei scattering are simply mathematical constructions based on ideal (unreal) objects. In real life, with real objects such as a mixture of irregularly shaped rock particles in water, we use the best theoretical approximations we can. We get a close, but not perfect, fit for our situation. All light absorption/reflection/scattering/interference at all scales boils down to a simple process: EM photons matching the charge oscillations of electrons in atoms.

Colloidal scattering is most strongly observed when particle size is between 0.04 um and 0.9 um, slightly under visible wavelength range, and just under the size of very fine rock flour. While this isn't a perfect match, the effect is much more intense than Rayleigh scattering, and therefore significant. The reason for this can be obtained by comparing the diagrams for Rayleigh and Mie scattering. While the Rayleigh scattering intensity is uniform around the particle (same size arrows), Mie scattering intensity is highest away from the light source, available to reflect the light straight back from the upper surfaces of deeper particles. Why this is so has to do with interference effects that develop as particle size increases from Rayleigh particle size (the size of a chemical bond) to Mie particle size (a sphere). The reflective/absorptive surface area of the particle starts to have a significant effect. Technically speaking, wavelength interference develops through phase variations over the particle's surface.

Colloidal particles are basically what we are dealing with when we talk about glacial water - rock particles that are small and light enough to suspend in water. An easy colloidal mixture to make at home is to suspend flour from the kitchen in a glass of water. Milled flour particles are between 1 and 100 um, according to engineeringtoolbox.com, pretty close to the particle range of rock flour (2 to 100 um). Wow! It looks a lot like blue glacial water (you have to get the concentration right for this effect).

Chris 73;Wikipedia Commons

Like the flour-water above, the smallest rock flour particles (around 2 um and just under) reflect most strongly in the visible blue range. They also display a fairly high irradiance ratio, meaning it’s a brilliant blue too. The irradiance ratio is the amount of light scattered backwards compared to the amount of light that continues down deeper into the water column. Particles around 20 um reflect more toward the green range and their irradiance ratio is quite low, about one tenth that of the smallest particles. Finally, large suspended particles (around 100 um) reflect equally across the visible spectrum, so white light shining down on them is reflected back up as white light. Like the smallest particles, their irradiance ratio is strong so they are highly reflective as well.

We can imagine what all these optical effects do when we think about water melting from a glacier's snout. There will be lots of rock flour right under the snout. This means that glacial water near its source, where the largest rock flour particles haven't had a chance to settle out yet, appears bright and milky white (or grey if the melting glacier has lots of dust/debris in it). A glacial lake with a very high concentration of suspended rock flour, including lots of medium size particles will look greener. A lake in which the larger particles have sifted out and only tiny particles remain will be the most intensely turquoise.

By the Tyndall mechanism alone, the inherent colour of the tiny rock pieces themselves has nothing to do with glacial lake colour. Tan, white and grey rock flours should produce similar effects. There could be other colour-enhancing effects at play, however. Think of the shallow clear brilliant turquoise water just off tropical beaches. The colour is especially vivid when seen from a plane directly above. In this case, we are likely dealing with a simpler effect. The sand and coral beneath these turquoise shallows is often roughly tan-coloured, like our rock flour. You can think of the ocean floor as a tan sheet having a high absorption coefficient for short wavelength (blue, indigo and violet) light. It in turn reflects some blue, green, yellow, orange and red light. If it wasn't underwater, these absorption/reflection effects would add up to what we see as the colour tan. The tan ocean floor eliminates some blue light by absorbing it and the water itself absorbs reds, oranges and yellows. What remains is blue-green, aqua or turquoise light, which makes these shallows look blue-green, aqua or turquoise. Water itself, as we learned, is intrinsically blue. That colour intensifies with the depth of the water column, so as the ocean deepens away from the shoal, the blue deepens and eventually gives way to the black colour of deep-ocean water.

Conclusion

As we've seen, glacial lakes can look turquoise, green, blue or milky white, depending on how much rock flour is suspended in them and especially how coarse those particles are. It also depends on what angle the Sun is shining. To see the most intense turquoise possible, look down at a glacial lake from a plane when the Sun is at its highest in the sky on a sunny clear day in late June. At this time, the light will be right for maximal turquoise reflection and the rock flour concentration should be at its highest from maximal melt.

Sometimes you can see layers of different colours in a glacial lake after a heavy rain as rainwater flows into it from rising rivers and streams, which change the rock flour concentration and distribution of grain size. Our glacial lakes and rivers change all the time, depending on rain, melt rate in the glaciers, how sunny it is, what angle you are viewing from and what time of day it is. Even when you know the science behind it, it still seems magical. In early spring, the magician has not yet arrived for the year. Glacial lakes have settled out over winter and they are dark blue just like other lakes. By mid-June at the peak of melt season, they turn intensely turquoise and they more or less tend to stay that way all summer long. The conversion to luminescent turquoise really seems more like magic than science, as this photo of Lake Louise proves.

You can see accumulated rock flour on the bank as it enters and becomes suspended in the glacial meltwater. Once suspended in deeper lake water, the magic happens. I have seen Lake Louise appear emerald green. Other times, it looks robin-egg blue, like the photo above, which has always been my favourite colour.

"How can we see something from the origin of the universe?"

Physicists have a lot of very good evidence that the universe inflated from a singularity, a single point, approximately 13.8 billion years ago. Some of the evidence comes indirectly from mathematical models, but physicists can also detect and investigate the echo left from this time because that echo remains today. The entire universe is bathed in a sea of microwaves, called the cosmic microwave background (CMB). The CMB is direct evidence for the Big Bang theory, as it is called. If you have an old TV set you can see CMB photons contributing to the static on the screen. This echo is still here because it has nowhere else to go.

The unimaginable energy (here, think about all the current matter and energy in the entire universe squeezed into a microscopic space) of the Big Bang unleashed a fury of extremely energetic photons called gamma rays. They shot around in every direction, crashing into other particles. Some of those photons are striking Earth today billions of years later, but as microwaves rather than gamma rays (fortunately). About 300 photons are flying through very square centimeter of space right now. There are so many of them, only a tiny fraction of them have stuck other particles and transferred their energy into other forms. Where are they coming from?

To get our answer we need to take a look at the expansion of the universe. No one knows why the universe expanded from a single point or what existed, if anything, before that. What physicists do know is that the Big Bang was the origin of both space and time, as we understand it. Space and time are described as dimensions in a four-dimensional theory called general relativity. Physicists can theoretically trace events back to about 1/32nd of a second after the universe popped, or banged, into existence. Before that time, current theories about how space-time operates break down into nonsense, so it is impossible to peak into the very first moments of our universe.

Shortly after the Big Bang (a tiny fraction of a second), the universe, according to most well established theories, went through a brief period where its expansion rate increased to faster than light speed. Physicists can prove that galaxies farthest away from us are also traveling away faster than light speed because we are now once again in an era of accelerating expansion. These two periods of accelerating expansion owe themselves to different mechanisms.

All of this faster than light business does not violate the special relativity rule that says that nothing, even light, can travel faster than light speed, because space-time ITSELF is the thing doing the moving, or expanding in this case.

How can we see the CMB when it is the oldest stuff in the universe? Shouldn't all those photons be invisible because they are moving away from us faster than they could travel toward us? The CMB permeates the entire universe. That includes the part of the universe that is not moving away from us faster than light speed. This is the CMB that is detectable. Everything is traveling away from everything else in the universe but the expansion is cumulative. You can't even detect it on the scale of our own galaxy but as you get further and further away over great distances, that miniscule expansion rate accumulates so that as you observe regions very far away, you are seeing photons coming from very old stars and their motion is approaching light speed. There is evidence of a much larger invisible universe outside the boundaries of the visible universe. Photons in this outer region will never reach us so it is invisible to us. When you look at the oldest stars you are seeing stars that are long burnt out. You are seeing the light that shone from huge white-hot stars that were born when the universe was just hundreds of millions of years old. In other words you can see ghost of stars that once were. There may currently be new stars in that region that are generations younger but their light hasn't reached you yet.

Expansion of the universe had a significant effect on CMB. Each photon travelling randomly through space travelled while that space expanded. As space expanded, the wavelength of each photon was increasingly stretched across it. What started as very short wavelength gamma photons are now long wavelength microwaves. We can no longer see them because they are now in the invisible part of the spectrum. If you could put on glasses that let you see in the microwave spectrum, you would see the universe glowing all around you. You might be wondering: does this mean that those photons lost their energy to space somehow? No, although this "tired light" hypothesis had some traction decades ago, apart from colliding with other particles and transferring their energy, photons retain as much energy as when they were created. There is no "friction" in the vacuum of space.

An important point to keep in mind is that there is nothing outside the universe, at least according to most theories. A common analogy used to think about the expanding universe is an expanding balloon, but there are some key differences. While a balloon expands into the space around it, there is no space for the universe to expand into, not even a vacuum. The surface of the balloon is a two-dimensional surface expanding outward, whereas the universe is expanding spatially in three dimensions. You can get a simplified idea of what the expansion looks like if you draw a few dots on a balloon with a marker and then blow it up. The dots grow further and further apart from each other as the balloon expands. In reality, however, the dots, which are galaxies of stars, are embedded in the substance of the balloon. A better analogy might be raisins moving apart from one another in a rising loaf of bread. There is an additional quality of space-time that the balloon doesn't illustrate. All the matter dotted throughout the universe affects space-time. It stretches space-time's four-dimensional fabric. Space-time stretches as the universe expands, taking everything in space along for the ride, but matter itself also stretches space locally. Like a bowling ball on a trampoline, a massive object such as a galaxy makes a depression in space-time, except the depression is in four dimensions rather than in three.

Space-time is also a relative fabric. The theory of special relativity says that how a section of space-time looks and behaves depends on how fast you are traveling relative to it. Think of an object like a spaceship travelling close to light speed through space relative to you. In other words it could be flying through space past you as you float in a space station. Its spatial length would appear to be squished. It would look like a short squat ship. It would appear as thin as sheet of paper if it were going almost exactly light speed. People on the space ship wouldn't notice anything weird. The ship's time frame would also stretch. Those people could travel to some planet light years away, settle down and colonize, and still be decades younger than you when they returned for a visit. The movie, Interstellar, illustrates this unsettling effect very well. Exactly at light speed, relative time stops altogether. If you could hitch a ride on the back of a photon, your children would live their lives, the entire history of mankind would play out, in fact the universe would unfold to its end around you and you would not have time to even blink. It would all play out instantly to you. Time also plays this tricky maneuver at the event horizon of a black hole. In this case, time slows down relative to the space around it because space-time around a massive black hole is stretched infinitely by mass concentrated to a point of infinite density. Here, too, however, the rules of physics break down and there is speculation about exactly goes on inside one.

If we get back to CMB for a moment, there is another way to think about why those original photons are still everywhere. Because the universe is a self-contained expansion of what was originally a singular point, there is no location in the universe that you can label as the starting point of that expansion. The entire universe is the starting point. Although scientists (correctly) talk about the oldest stars being those that are farthest away from us, it can be a bit misleading. It doesn't mean that space-time is youngest in the center, and that center is surrounded by older shells of space-time. The space-time of the whole universe is the same age. The whole universe is that point in space 13.8 billion years after it started to expand. That being said, there are older and younger stars. The oldest stars, mentioned earlier, first began to shine just a few hundred million years after the Big Bang. Thanks to the universe being a giant self-contained Big Bang, we can see almost the entire evolution of the universe just by observing photons of light. All the ghosts and echoes of the past are still there. From Earth, distant stars are moving away in all directions. They are red-shifted, or Hubble shifted. The most distant stars are the oldest stars and they are moving away fastest. This is not because Earth is in the center of the universe. It is because everything is moving away from everything. If you could jump into a wormhole and travel instantly somehow to any distant planet in any distant galaxy, you could look up in the night sky there and see the same phenomena. Distant stars would be moving away in all directions and the oldest most distant ones would moving away fastest, approaching light speed.

By mapping out the ancient CMB photons across the visible sky, scientists can glimpse what the very young universe looked like. Its energy wasn't perfectly homogenous. Those slight variations in density created tiny gravitational pockets into which matter tended to clump, forming the first stars and galaxies. Those ancient photons could not travel freely when they were first created. The universe was so dense and energetic that they constantly banged into electrons and protons. When the universe was about 380,000 years old, the density of electrons and protons was low enough that photons could travel for distances between them. From then on, they were free to travel not outward into the universe but in all directions as the universe expanded outward. That is why we can see back into the universe's past only up to 380, 000 years after the Big Bang. By observing a similar map of neutrinos, however, physicists hope to see the ghost of an even younger universe, because neutrinos escaped and began to stream long before photons could. The tricky part here is that neutrinos are themselves almost ghostlike particles. They are very difficult to detect.