Colour in scientific graphs – part 1

In the past years, there is a surge in the amount of colour that appears in scientific papers. There are several reasons for it. The ease of creating coloured figures increases with every new release of a software version. Journals are accepting (and encouraging) colour more than ever. Additionally, in the world of online publishing, there are no added costs for colour.

In the past, people had only black and white (and grey scale if lucky) to use. This forced them to think about line types (dashed vs dotted), symbols (circles and squares), etc. to make the graph easy to understand. Today, many people submit their pretty colour figures without putting too much thought into the act.

Let’s take a look at an example:


Seemingly, there’s nothing wrong with that graph. There are two data series, both of which are coloured (red and green) and appear in the legend. However, this figure fails in two critical aspects:

Black and white photocopying and printing

Even though it’s possible to view articles online nowadays, many people still print out papers. Just take a look at any desk here at RSES. While you may think that the green looks brighter than the red so when printing one will be dark grey and the other will be light grey, it is not the case. The two colours have the exact same luminosity and thus will appear identical in print. It is then left to the reader to try and figure out which are better, kittens or spiders. And it is open to subjective interpretation.

There is a simple solution to that: use different symbols in addition to colour:


This way, people with no access to colour printing can still understand your figure, while people with colour printing or screen viewing can benefit from your use of colour. If you prefer to use only lines and no symbols, there is a solution for you as well – line types:


Accessibility and colour blindness

Colour blindness affects mostly males (in some countries close to 10% of the population) but also females. This means that you can be confident that some of your readers are certain to be colour blind. Most commonly, people cannot distinguish red from green. This makes our figure a great example of how not to do it. While people can use the symbols to distinguish the two data series, the entire point of the colour is reduce the time the reader needs to understand the figure. Which colours should we use then? This is a hard question because of the multitude of colour blindness forms. Generally, blue and orange are a good combination:


However, with increasing data series this quickly becomes a hard task. Fortunately, there are online tools to assist in colour choice. One of my favourites is ColorBrewer: but there are others (just search online for ‘colorblind safe colors’).

Remember – colour in scientific figures is a tool to assist understand the figure better and faster. It is not to make it pretty, although that is usually a welcome side effect.

There are two ways to test your figures after you’ve created them. First of all, print it black and white. Can you still understand the different data series? Second, you should run your figure through a simulator (or ask a colour blind friend). My favourite is the one at, but it requires that your figure already exists somewhere online. Finding other simulators should be easy using a search engine. Try to run the red-green and the blue-orange versions through the simulator!

Final note about legends

Legends are terrible in my opinion: your eyes keep moving around from the actual data to the legend. I always try to avoid using them. A better alternative is quite obvious in our case:


That’s it for now. Stay tuned for part 2 where I will discuss colour gradients. There plots were produced using R. If anyone is interested in seeing the code I used, feel free to ask me.

The Great Mantle Plume Debate

The Great Mantle Plume Debate has been simmering aggressively – in a similar fashion to soup on an unattended stove- throughout all facets of Earth Science since the 1960’s. Seismologists, geophysicists, experimental petrologists, and analytical geochemists alike have invested serious time and money trying to solve the mystery that is the mantle plume. Since starting my PhD on the geochemistry of Hawaiian volcanism, I have been thrust- like a spoon in a tub of icecream- to the complex topic that is mantle plume theory, and have been surprised to learn just how much is still in dispute, and how much is still to be discovered. Sounds dramatic, but -much like a good Agatha Christie novel- a lot of headway has been made by some unassuming and cool-headed detectives (scientists).

The aggressively simmering pot of soup that is the mantle plume debate.

The aggressively simmering pot of soup that is the mantle plume debate.

First of all, what are mantle plumes?

‘Mantle plume’ is the name given to buoyant material that rises from depth in the Earth through the mantle in a narrow conduit. This material melts, and produces volcanoes at the Earth’s surface which take on a linear geometry if a tectonic plate is moving overhead- like a linear row of oozing pimples…(?)

‘I don’t care about plumes!’ I hear you say.

Plume theory, now widely accepted (although there are many influential scientists who do not subscribe to the theory in its current form), is used to explain the curious types of volcanoes that form independent of tectonic forces, unlike the majority of volcanism on Earth. For example the volcanoes which make Earth’s continental crust are produced by the collision of two tectonic plates, and most of the Earth’s oceanic crust is produced by volcanism at two diverging plates. Some volcanoes however, occur far away from plate boundaries, and the mechanism for their formation is still shrouded in mystery! Much like a man in a trench-coat would be, if he were lurking on a street corner at night.

Mantle plumes if they were a person.

Mantle plumes if they were a person.

These mysterious volcanoes are known by many different names, depending on who you talk to, including:

  • Hotspots
  • Wetspots
  • Ocean Island Basalts (OIB)
  • Or vaguely: ‘melting anomalies’

The reason for all the names is that there are still many different theories circulating in the field of Earth sciences as to how mantle plumes form, what they are exactly, and if they even exist! Yes, many Earth Scientists maintain that plumes do not exist, and that hotspot volcanism is passive, caused by surficial processes such as shallow cracks and fissures in the crust- controversial! This group prefer to use the term ‘melting anomaly’. I will continue the rest of the article assuming that plumes do exist, however plume denialists raise a good point that the current one-size-fits-all plume theory struggles to account for the many unique features of each hotspot.

At this stage it needs to be pointed out that there are over 60 hotspots/wetspots/OIBs documented globally, and every single one on Earth is unique in some way.

Many appear to start at different mantle depths, have different isotopic trace element signatures, and even highly variable major element signatures. They all have different volume fluxes, uplift and subsidence rates, some form tracks while others don’t; some have flood basalts whereas others don’t. In general though, they all tend to be located around Africa or the Pacific Ocean- kinda weird!? Some well-known hotspots include:

  • Hawaii
  • Iceland
  • Reunion
  • Yellowstone
  • Samoa
  • Kerguelen
  • Ascension
  • Cape Verde
  • Canary Islands
  • Easter Island

The two big questions that are the crux of the Great Mantle Plume Debate are:

1) How deep in the Earth do plumes originate?

The best place to start with this question, is seismic evidence. The speed of seismic waves through the mantle enables us to see – as one would see a lady’s intimate apparel by using X-ray vision goggles- material of contrasting density, and thus to see plumes. Attenuation tomography allows us to see slices through the mantle and reveals differences in porosity, permeability, and viscosity of the material through which it is slicing. Studies have shown that, yes! we can see plumes below hotspots, and interestingly most originate from highly variable depths in the mantle: some are restricted to the upper mantle, some to the lower mantle, and a few have even been tracked to the core-mantle boundary (2800 km deep in the Earth… that’s as deep as an LSD-fuelled philosophical epiphany!). This method is good, but it seems the resolution is not good enough yet to have persuaded everybody in the Earth Sciences.

Seismic waves (X-ray goggles) travelling through the Earth's mantle (lady’s peticoat).

Seismic waves (X-ray goggles) travelling through the Earth’s mantle (lady’s peticoat).

A good way to show you how seismic imagery may not be high resolution enough is to show you what mantle plumes are supposed to look like, based on fluid dynamic experiments, and then compare it to what we actually see:

Left: what plumes are supposed to look like. Right: actual tomographic images of the plume beneath Hawaii.

What plumes are supposed to look like.

plume hawaii

Actual tomographic images of the plume beneath Hawaii.









As you can see, some colourful imagination is required to convert between the two. What we can say for sure though, is that large areas of mantle below Africa and the Pacific contrast seismically to surrounding ambient mantle. As I said before, this also happens to be where most of the hotspots are located. Coincidence…? I think not!

In addition, geochemists –represent- know that the isotopic signature of hotspot lavas is very different to that of mid-ocean ridge lavas, indicating they must have come from much deeper in the mantle than the strongly depleted mid-ocean ridge lavas. Further geochemical evidence for a core mantle boundary source for some hotspots comes from the helium isotopic signature, although this isotope as a tracer has come under scrutiny because other processes may affect the ratio.

Some disagree with a deep source however, and suggest that the source depth may be restricted to the upper mantle (shallower than mid-ocean ridges in some cases), or at least to the 660km discontinuity, and that upwelling occurs as a result of mineral phase change there.

2) What causes the material to upwell?

This question appears to divide people the most. There are broadly two possibilities: it is a thermal anomaly at some boundary layer that causes material to become warm and buoyant, or it is a geochemical compositional anomaly that causes buoyancy.

Another possibility which does not get discussed nearly enough in my opinion is that it could be some combination of the two. Both work on the same principal that warmer material or material with a different chemical composition would become buoyant and upwell to create a plume- in the same way you can melt ice by either warming it up or by adding salt to it.

A few methods have been used to investigate this; one is using calculation of buoyancy flux to see what temperatures would be required to produce large amounts of melting- many people believe that for any large amount of melting to occur an increase in temperature must exist. This is where the name ‘hotspot’ originates from. Similarly, topography can be used to infer the amount of heat entering the plume, and subsequently the depth the plume originates.

Another method is the use of geothermometry to determine mantle temperatures. One that is commonly used is the olivine geothermometer, however the use of this method has produced excess temperatures of over 200̊C in some studies, but zero excess in others. Either way the thermal argument would require long-lived thermal anomalies at the core-mantle boundary, and heat conducting from the core heterogeneously.

Geochemists and petrologists are more in favour of a chemical argument for plume formationfunnily enough.

The general idea is that subducted crust may sink down to a boundary layer where it sits, warms up and becomes buoyant, and melt is produced because the reaction between this crust and mantle forms material with a lower melting point than ambient mantle. There is growing evidence that the source rock for OIB lavas may be pyroxenitic, and not peridotitic, based on experimental petrology and the ratios of major elements present in the lava.

Others speculate that the amount of water and carbon in this subducted material may be the cause for melting. ‘This is why ‘hotspots’ have started to be called ‘wetspots’ by some scientists.

Other studies have rejected the need for a deep source for this crustal material, suggesting that it may simply be delaminated lithosphere that has metasomatized the asthenosphere to cause melting.

Seismic studies have found ‘blebs’ of material sitting at the core-mantle boundary (called large low shear velocity provinces), but it cannot distinguish the difference between thermal or compositional ‘blebs’, so unfortunately this method is less useful for answering this question.

In the end, we have many different unique volcanoes that vary in so many different ways, and we don’t have a mantle plume model that explains all of them. We also don’t know the exact mechanism that causes plumes, and we don’t know for sure at what depth in the Earth they form. This is good news for us geologists as it gives us something to do with our attention deficit brains.

See What I Mean?

By Pat

Eleanor’s post last week on scale provided the perfect segue for my first ever blog (!). Like many people when I think about my work, I find it useful to visualise processes to better understand them. I do this when I think about the interactions of atoms, tectonic plates and planets for example. In geoscience it seems we often work on scales that are either too small to see or too large see all at once (or at all) and thus an imagination is vital. It is a misconception that creative people study arts whilst regimented people study science. Studying geology, or any science for that matter, requires imagination to both visualize interactions on different scales and to hypothesize new interactions or new angles to old problems. In science, our imagination is grounded in fact; we have anchor points or truths that we have to incorporate. In a vague metaphorical way it’s like we are creating our own “connect-the-dot” pictures, where the dots represent our current knowledge but it’s our imaginations that dictate the lines (see figure below).

Two interpretations of a simple connect-the-dots game. The artist on the right may be cheating by visiting the same point more than once, but i think my point is still valid.

Two interpretations of a simple connect-the-dots game. The artist on the right may be cheating by visiting the same point more than once, but I think my point is still valid.

Indeed it takes a great deal of confidence to be imaginative in the sciences – where physical laws can be so restricting. However I believe the greatest breakthroughs in geoscience probably originate from the greatest imaginations. Similarly, for me at least, the most exciting science is where there are fewer facts and thus the most room for imagination and innovation (e.g. the creation of the universe, search for life, or the origin and evolution of mineralising fluids in tin systems as indicated by tourmaline and cassiterite chemistry). At times however, in particular when we are trying to understand the basic principles behind these facts, I believe it is helpful to be able to see the processes, rather than having to imagine them. In my studies, most of this is happening at the atomic scale which, as you can imagine (haha) is hard to see. A large component of my PhD involves radiometric dating using the U-Pb decay system. We can date rocks in this way because we know that 238U decays through a series of different isotopes at a known rate¹, until eventually it stabilises as 204Pb (see figure below). Thus if we measure the amount of 238U and 206Pb in a rock or mineral we can estimate its age. Incidentally I noticed that this blog page already has a summary of radiometric dating here.

Decay chain of U and Th isotopes. Time taking to decay from each isotope is in brackets. a= ,

Decay chain of U and Th isotopes. Time taking to decay from each isotope is in brackets. Thanks USGS for the image.

Until last week I regarded this decay as one of those processes I could only imagine. However recently, whilst doing some YouTube ‘research’ I came across the following riveting video:

It is an experiment in what’s known as a “Cloud” or “Wilson” chamber which is supersaturated in water vapour or alcohol and sealed. Any form of ionizing radiation within the chamber will cause local ionisation of the gas and result in condensation and the formation of mist. If you put a large chunk of highly radioactive U metal in this chamber the radiation emitted every time an isotope decays through the sequence (as shown in above figure) it emits radiation, resulting in the formation of a trail of mist. The video is ~50 minutes long, and I highly recommend getting yourself a bag of popcorn and a choc top and enjoying the entire show. To make it an extra special occasion bite off the outer chocolate layer and then dunk the choc top into the popcorn so that the popcorn sticks to the ice-cream. If you then eat the popcorn with the light ice-cream spread you will remember the true meaning of happiness.  But if you’re in a rush, just skip to ~5 minutes in to see the main action scene where radiation is being emitted continuously and in every direction. Incredible. Whilst this experiment used mostly pure U metal and thus is a lot more intense than anything you would see from a natural sample, it helps visualise a few important aspects of my work.

Firstly this is why we don’t go near radioactive stuff.

Secondly, each trail flying out from the sample effectively represents another tick (or tock) in the U-Pb radiometric clock. As the isotope decays it releases particles and radiation. In this highly radioactive sample decay is happening vigorously. Theoretically those trails would continue to form for billions of years, however over the course of the experiment alcohol begins to condense on the U metal.

Thirdly, and building on from my first point, it illustrates the amount of heat that can be generated by radioactive decay. A large proportion of Earths heat is generated in this way in the mantle and crust. This heat has huge implications for both mantle convection² and many crustal processes (including forming hydrothermal mineralising systems – which is why I care).

Directly observing all the processes we are or will be interested in is (surely?) impossible, and in these situations it is useful to visualise what is going on. When we do get the opportunity to witness something I believe it is an incredible tool for both understanding and educating others – which ultimately is what can spark the imagination and lead to breakthroughs in our knowledge. As Albert Einstein put it,

“To raise new questions, new possibilities, to regard old problems from a new angle, requires creative imagination and marks real advance in science.”

In conclusion, never underestimate the Youtube database.

¹As it turns out, this known rate is not so well known.

²As I understand this is still up for debate.

All that you can learn from mud

By Jess

My research is in paleoclimate, which involves looking for clues in the world around us and putting these clues together to investigate past climatic changes and build a picture of the way Earth used to be. It’s a bit like being a detective (well at least it is in my maybe slightly overactive imagination).

Slightly less glamorous is that all of the clues I am looking at for my research are found in mud that has been buried under the Mediterranean Sea for thousands if not millions of years.

Some sediment cores – who would have thought mud could be so colourful

Some sediment cores – who would have thought mud could be so colourful

So marine sediment cores might not look incredibly exciting at first glance, but it’s amazing how much information can be discovered in what would usually be dismissed as muck.

Sediment cores contain thousands of little microscopic shells. Tiny organisms called foraminifera which live in the water column, form these calcite shells as they grow, and when they die the shells sink to the sea floor and become preserved in sediments. As it is formed, the shell composition records information about the seawater around it, and if the sediment is dug up, this information can be retrieved.

Foraminifera shells, very close up. These are actually only about a quarter of a millimeter in size.

Foraminifera shells, very close up. In real life these are about the same size as a grain of sand.

For example, I measure stable oxygen isotope ratios of foraminifera shells, which are strongly influenced by global ice volumes. So from these tiny little shells, it is possible for me to reconstruct sea level changes going back millions of years in time.

Even leaf wax molecules from vegetation washed into the sea can become preserved in sediments. In January I visited the University of Manchester in the UK and learnt how to extract these from sediment samples. By measuring different isotopes from the leaf wax remains it is possible to track past changes in rainfall and vegetation type for the region.

These methods take a bit more effort, but you can learn things even by looking at the colour of the sediment cores. In the Mediterranean cores I work on, there are frequent intervals of black mud. These are ‘sapropels’ and form under anoxic conditions, when the deep sea is depleted of oxygen. This allows organic matter to be preserved, giving the sediments their distinct colour. These anoxic events occurred regularly in the Mediterranean’s history, when Earth’s orbit is such that there are warmer northern hemisphere summers (a minima in the precession cycle). We know how Earth’s orbits changed in history and therefore these sapropels can be used to accurately date the cores.

These are only a few of the things you can study in sediment cores to investigate past climatic and oceanic conditions, but there are lots, lots more. It is also possible to obtain records of temperature, ocean pH and levels of biological productivity, to name just a few.

So next time you’re swimming in the sea take a second to think of all the processes going on around you which are preserving clues about our world as it is today, all in the mud on the seafloor!

Small science, big picture

By Eleanor

Something I find bizarre and amazing about science is the juxtaposition of scales.

You can have a big problem, like “how did our planet form?” and attack it by understanding very small things, like how atoms arrange themselves in magma, or tiny differences in the amounts of elements present in different rocks. By understanding many small things, you can build a huge, yet rigorous and detailed picture of the world. Like a high-resolution panoramic photograph.

As a scientist, thinking about my research every day, I get desensitised sometimes to how cool it is. But sometimes I read something that blows my mind a little bit, and gives me a new appreciation for what it is I’m doing, and the achievements of the generations of scientists before us.

A moment like this happened last week. I was reading about a new technique that I will be using soon, which involves measuring something on the scale of ‘parts per million’.

What does a concentration of parts per million mean? Well, imagine a water tank that can hold ten thousand litres of water… like this one:

Now imagine that you fill it up with water, and then (assuming there’s a little more space in the top), pour in 100 mL (a bit under half a cup) of apple juice.

The concentration of apple juice in this water tank is ten parts per million.

Now imagine you have two water tanks, and an evil mastermind has put 100 mL apple juice into one, and 110 mL into the other. This evil mastermind will tell you all the secrets to the universe if you can figure out which one is which (and no guessing allowed; that’s not scientific!).

How would you do it? You could try to taste the difference, but I’d be willing to bet that you wouldn’t even be able to tell there was apple juice there at all, since it’s so dilute, let alone the difference between the tanks.

This is exactly the type of problem that earth scientists are confronted with on a regular basis. The evil mastermind is nature, who will share her secrets only if we try very very hard. The tank of water could be rock, and the apple juice is a certain element that we try to measure in the rock.

Fortunately, scientists have, over decades of work, developed techniques and machines that help us to make measurements like these. It’s something I often take for granted because these measurements can seem so routine, but it’s nice to remind myself every now and then that some of the simplest things I do are actually pretty cool. And that some of the smallest details that we discover are helping to complete that high resolution, panoramic photograph of the world.




Fish Ears at an Ancient Lake

By Kelsie Long

Hi all! I am part of the small but unique Archaeogeochemistry group here at the Research School of Earth Sciences (RSES).

We are focused on bridging the gap between analytical sciences and archaeological research. For my part I completed my undergraduate degree in archaeology and then for my honours project I moved into the realms of palaeoenvironmental reconstruction at the archaeologically and geologically significant site of Lake Mungo. I am continuing this work for my PhD. In the months to come I want to rope in some of the other archaeogeochemists to tell you about the cool and interesting stuff they are doing but for this post I will introduce you to Lake Mungo and the wonderful realm of fish otolith geochemistry.

Lake Mungo is one of seven now dry lakes that together make up the Willandra lakes world heritage area situated in south western, New South Wales. These lakes once formed the major overflow outlet for the Lachlan River and are enormous, in total covering some 240,000 hectares! I visited Lake Mungo for a couple of days in 2012 and again in 2014 and can attest to the areas almost ethereal beauty. As you walk along the ancient lakes’ shorelines, giant lunette outcrops tower over you. Exposed over time by wind and water, these structures provide not only welcome shade on the rather barren dunes, but also evidence for cyclic lake level fluctuations stretching back over the last 100,000 years.

Mungo 1

A view across Lake Mungo from the lunettes, two humans for scale (extra points if you can spot them).

In the 1970s human remains discovered eroding out of Mungo’s shoreline lunettes pushed dates for human occupation of Australia to beyond 20,000 years BP, radically changing our understanding of Australian pre-history at the time. Since then the development of new dating technologies and the discovery of further human remains have extended the timing of human occupation to beyond 40,000 years BP (not very old in a geological sense but for Australian archaeologists this was pretty epic).  Other evidence for human occupation found in the lunettes includes stone tools and fish bone fireplaces.

The lakeside lunettes of Lake Mungo are the focus of ongoing excavations and geological mapping by Nicola Stern and colleagues at the La Trobe University in Melbourne. Evidence for human occupation has turned up during some periods but not others. Why? I want to build up more information about what the lakes were like when humans were visiting them thousands of years ago and explore why certain lake conditions were more attractive than others. To this end I am assessing how the geochemistry of fish otoliths can be used as recorders of ambient water conditions.


Excavations at one of the fish bone fireplaces at Lake Mungo (or maybe an area with stone tools, this was taken in 2012 and I can’t quite remember)

Otoliths are structures that form within the inner ear of bony fish by the incremental deposition of calcium carbonate forming annual rings similar to those of trees. As calcium carbonate is laid down elements from the surrounding water, some with distinctive isotopic compositions, become incorporated into the otolith matrix. By cutting transversely through fish otoliths their age rings can be exposed. Then we can analyse their unique elemental and isotopic compositions  using high resolution in situ techniques, such as the Sensitive High resolution Ion Microprobe (SHRIMP) here at the RSES.

otolith images 1

Clockwise from top left: an otolith and some burnt fish bones at Lake Mungo (the scale bar has 1 cm increments); two otoliths from Mungo encased in resin; a modern otolith in resin sectioned to expose annual rings; a modern otolith thin section showing those beautiful annual rings.

At the moment I am determining how the geochemistry of modern fish otoliths record water composition change, such as during evaporation or flooding, and whether or not we can pick up a temperature signal. Once I have established this I will then apply the same techniques to  ancient otoliths from Lake Mungo. Preliminary results from the modern fish studies show that large changes in the ambient water chemistry during the fish’s life, such as during evaporation or flooding, can be easily identified  but temperature effects are not as easily discernible.

I will leave you with the following comic because it made me laugh and now I wish we had interdepartmental pun-offs!

geology pun

Archaeologists vs Geologists in a pun-off

The Foram’s Song

Are you fed up with boring graphs? What about listening to your data instead of starring at them for hours?

It is possible thanks to the sonification of data.

Above, an example of the sonification played by physicists of CERN (European Organisation for Nuclear Research) for the 60th birthday of the organisation. This piece of music is made of different data set from the LHC (Large Hadron Collider), giving an idea of the complexity of the Universe.

How did they do that?

A brief explanation on sonification from BBC

The idea is simple. You assign a musical note to a value. The greater your value is, the higher the pitch gets. If your data are increasing the higher your composition will become and vice versa. Their are several free software programs on the internet that enables you to do this like Puredata among others.

Sonification has been used for a long time without creating such nice melodies. For instance, the Geiger counter uses the same principle as the more radiation there is, the strongest the sound you hear.

Why should we use it?

Because it sounds fun. And because, sonification of data enables scientists to hear more information at once than when looking at graphs. You can put together several data set and recognize patterns and harmony between them. In fact, it’s possible to hear many dimensions.

Their are many other examples on the internet. This one is a real current issue:

I am only an amateur but I know that there are few people in RSES who are very good music players. Maybe we could team up and try to work on an RSES sonification project with geology data?