这是一个语言的科学吗

Since the end of WW II, English has been the dominant language of science.  This was not always the case.  The late 19th century industrial and scientific explosion in Germany made German a potential contender before geopolitical events depopulated Germany of its scientists.  And earlier in the 19th century French, and originally Latin, were the languages of the day.

The reasons for this are simple.  England has long been a leader in scientific inquiry, and the post-war assimilation of European scientists by the USA and subsequent technological revolution there during the space race and information revolution has kept English on the forefront.

None-the-less, many scientists do still publish in their native languages.  And even when they do publish in English, there are many Journals, such as the Journal of South American Geology Earth Sciences, which offer abstracts in other languages, such as Portuguese and Spanish, the dominant languages of that continent.  Similarly, Geostandards and Geoanalytical Research publishes French Abstract, since is is based in France and published by a French research organisation.

None-the-less, I was surprised to see that the Australian Journal of Earth Sciences is now publishing abstracts in Chinese for its English articles.  Australia is an English speaking country, and although there are small but locally important groups of immigrants who speak various Chinese languages, they are not over-represented in the Earth Sciences.  And while Chinese geologists compete internationally better than their scientists on other fields, and Chinese investment is important in the Australian mineral export industry, it is still a bold move by the AJES editors to pick Chinese as the next language of science.

p.s. If you can't read the title, check that your operating system has Asian characters enabled.

Grains of sand

How many grains of sand are there on earth?  That is a good question.  But a ball-park estimate is fairly simple.

We will look at fine sand (grain size = 100 microns), and coarse sand (grain size = 1 mm).

So a cubic mm can hold 1000 grains of fine sand, or 1 grain of course sand.  Obviously grain size is important.

There are 1x10¹⁸ cubic millimeters in a cubic km.

How many cubic km of sand, sandstone, etc we have Is a tricky question.  But if we say the average thickness of all sand for the globe is 200m (a thin number in any sedimentary basin, but most of the Earth is not a basin in the traditional sense.  The surface area of earth is 5x10 ⁸ km2, so a 0.2 km layer gives 10 ⁸ cubic km of sand.

This brings the total grain count to somewhere between 10³ x 10 ¹⁸ x 10 ⁸ = 10 ²⁹ for the fine sand, and a thousand times less than that, or 10 ²⁶ grains of the coarse sand.  If you want to know how that compares to the number of stars in the sky, ask an astronomer.

How many of your co-authors have you actually met?

In my meandering career from academia to government to private sector, and back into all the grey areas in between, I've been an author on a few journal articles, government reports, and other publications. Usually, these are collaborations between groups of separated people, not all of whom interact with every other member of the team.  For example, in the academic literature, I have a total of 21 co-authors, of whom I have met 9.  If we include government reports as well as papers, then I have 42 co-authors, of whom I have met 17.  I find it interesting that this ratio is so similar between the two types of reporting (about 40%).  So I was wondering: for those of you who read this blog and publish, is your ratio about the same?

Universities Australia sticks it to the Australian high technology industry

Universities Australia has launched a recent ad campaign decrying proposed funding cuts to university research.  This ad showcases the products of off-shore corporate giants which are trying to destroy the Australian high tech industry. 

The complicated scientific instrument pictured in the ad from 0:12 to 0:17 is something called a IMS-1280, manufactured by the American technology amalgamation Ametek under the brand name of Cameca, a European tech company which Ametek took over last decade. Ametek is perhaps the most aggressive corporate giant around in trying to leverage the recent high Australian dollar to destroy the Australian technology industry. 

Obviously, Australia is only a mid-size country, and most instrumentation in Australian universities is sourced from off-shore suppliers.  But many of these suppliers are good corporate citizens, who set up Australian subsidiaries, employ Australian graduates, and work closely with Australian agents, subcontractors, and scientists to sustain the high technology industries that define advanced economies in the 21^st century.  Indeed, one of these companies, the Japanese technology group JEOL, has an electron probe installed just across the hall from where the picture in the ad was taken.

Ametek is not a good corporate citizens.  Instead of collaborating with Australian manufacturers, they hire foreign lawyers to block sales around the globe.  While other companies reinvest in Australian research they hire slick Morden-like spokespeople to belittle the achievements of Australian academics.  And instead of helping Australian universities improve productivity and reduce costs through co-developed hardware and software modifications, they lock their customers into exorbitant service contracts, the proceeds of which allow them to underbid Australian companies whose instruments are generally preferred by researchers all over the world.

Every time one of the instruments pictured in this ad is purchased instead of an Australian equivalent, Australian universities lose hundreds of thousands dollars in direct payments from Australian companies and their international customers. It also means that Australian companies cannot create jobs for university graduates, such as those pictured in the first part of the ad.

The government is proposing cuts to university funding because of a revenue shortfall.  Revenue is down because aggressive corporate tactics by companies like Ametek are denying work to Australian companies, resulting in fewer hours worked, reduced income for the employees, and reduced income tax payment to the government.  So the approach of Universities Australia to showcase one of the most aggressive job-killers in their ad asking for government money is incredibly callous to all Australian trying to earn a living outside of the Ivory tower.

The Wool Sock’s Carbon Footprint

Four years ago, I blogged about the cognitive disconnect between the ecological perceptions of wearing wool and eating beef.  However, I did not actually calculate out exactly what the carbon footprint of a wool sock is.  Here it goes:

According to Wikipedia’s wool bale article, a bale contains about 60 fleeces, and weights 150 ± 50 kg.  This gives a fleece weight of about 2.5 kg.

This wool sock weighs about 100g, meaning that you can get about 25 socks per fleece.  A sheep produces one fleece per year.

A ballpark estimate from the NSW department of primary industries suggests that a medium sized (45 kg) adult sheep in warm weather needs about 500g of dry feed per day to survive.  If this feed is mostly cellulose, it will metabolize to produce about 800g of CO₂ per day, or 297 kg/ year. Assuming 25 socks per year, that gives about 12 kg of respired CO₂ per sock.

However, in addition to respiration, sheep also produce a fair amount of methane, which is generally considered to be 25 times more potent a greenhouse gas than carbon dioxide.   This paper estimates a methane yield of about 20 grams / day/ sheep, or about 7.3 kg of methane per year.  Using the 25 times multiplier, we get a CO₂ equivalence for that methane of about 180 kg / sheep/ year, which is a bit over half the direct respiration emissions.  Dividing by 25 socks/sheep gives is a CO₂ equivalent of 7.3 kg per sock (300 grams methane).  In total, our CO₂ equivalent emissions from the sheep are about 19 kg of CO₂ per wool sock- 12 from respiration, and 7 from methane.  This figure only includes the CO₂ footprint for growing the wool.  It does not include additional emissions from shearing, transporting the wool, spinning it into yard, and manufacturing the sock.  This is the same amount of CO₂ released by burning about 8 liters of gasoline (which is enough to drive a mid-size car 100 km), or one sixth the emissions of a top fuel drag race (with 2 cars in it).  So a hackey sack game with more than three pairs of new socks in it is worse for the atmosphere than this.

In contrast, a 50 gram synthetic sock (synthetics weigh less than wool) probably has a carbon footprint of 10-25 grams*.  It production is one THOUSAND times less carbon intensive than a wool sock.  So the next time some green evangelists starts looking down their noses at your car or your plate, check out their feet.

* In both the case of the plastic sock and the wool sock, the carbon in the sock itself is sequestered in the sock drawer for the lifetime of the sock, and in a landfill for several decades afterwards.  Unless you burn your old socks, which smells, or recycle your used synthetic socks into drink bottles, which is disgusting.

Why deflecting asteroids is a really bad idea

In the aftermath of the Chelyabinsk fireball last month, there have been increasing calls to identify asteroids on a collision course with Earth and develop technologies to deflect them.  This would be a very stupid thing to do.

The reason for this can be seen in figure 1, below.  In part A, this figure shows the minimum deflection necessary to make an asteroid on a collision course with  Earth to miss.  The deflection angle depends on how far from Earth this deflection occurs; the farther away, the smaller the angle.  In practice, very small angles from very far away would be used. 

The green line shows the minimum translational distance an asteroid must be deflected in order to miss the Earth.

Figure 1. An illustration for how the deflection needed to make an asteroid miss can be used to make many more hit.

The problem with such a system is shown in part B of the figure.  Here, an identical deflection is applied to a harmless asteroid that never would have hit Earth.  However, by deflecting it towards the Earth, this harmless rock ends up exploding in the atmosphere.  For a rock the size of the Chelyabinsk bolide, this is similar in force to a large nuclear weapon.

The area of the red circle- the smallest radius necessary to protect the earth- is three times the cross section of the earth.  So for every rock you deflect, there will be at least three harmless rocks that can be turned into weapons of mass destruction.  By definition, a “planetary defense system” turns every rock that passes close to the Earth into a potential weapon of mass destruction. 

Who would actually crash a space rock into a populated area of the Earth?  The same people who crash airplanes into skyscrapers of course.  And while only a few rouge countries can launch satellites, any spacecraft in radio contact with Earth can potentially be hijacked by a hacker on Earth with enough chicken wire to erect a makeshift dish in a desert.  Amateurs already pick up signals from our most distant space probes; an asteroid deflection mission would be a magnet for every doomsday cult, terrorist fanatic, delusional hacker, and other misanthropes whose imagination had previously been limited to shooting up schools.  Obviously nobody is going to design a space deflector to be hackable, but then the drone the Iranians hijacked wasn’t supposed to be vulnerable to those sorts of attacks either. 

The threat of an asteroid impact is miniscule.  More people were killed in floods this week than were killed by impacts in the known history of the human race.  A quick glance at the morphology of our planet will explain why.    Even the giant extinction-causing impacts are less common than large flood basalt eruptions of similar ecological lethality.  But developing the technology to deflect asteroids potentially gives all the wrong people access to a weapon the size of a large hydrogen bomb for a fraction of the development cost.  This is not a smart thing to do.

High mass resolution mass spectrometry

Mass spectrometry is the dark art of separating objects by mass.  The name comes from the alchemal days of photographic plate detectors; just like a prism separates white light into a spectrum of colors, a magnet can separate a beam of ions into their component masses, which will then form an image on a plate.

These days, electronic counting systems have replaced chemical emulsion ion detectors, but the name lives on.  In the case of atomic and molecular charged particles, the masses are not continuously distributed, like the energy distribution of white light.  Rather, different ions have discrete masses.  To a first approximation, the nominal mass of an atom (or an atomic ion, if the atom is charged) is simply the sum of its protons and neutrons.  Thus, an atom with 26 protons and 30 neutrons (Iron fifty-six, abbreviated by scientists as “⁵⁶Fe”) has a nominal mass of 56. 

The whole point of mass spectrometry is to separate things with different masses.  So, for example, most mass spectrometers can separate ⁵⁶Fe, with 26 protons 30 neutrons, from ⁵⁴Fe, which also has 26 protons, but only has 28 neutrons. The ability to distinguish atoms of the same element with different mass- isotopes- is one of the main uses of mass spectrometers.  In this case, separating ⁵⁴Fe from ⁵⁶Fe requires a mass resolution of 1 part in 28.  The mass resolution, defined by IUPAC as M/ΔM = 56/2 = 28.  This is quite low. It is about nine times worse than what is needed to separate ²⁴⁰Pu from ²³⁹Pu, for example. And even then, a mass resolution of 240 is still generally considered low.  There is no formal definition of high and low mass resolution.  However, as a general rule, mass spectrometers which can only measure the nominal masses of inorganic ions are generally known as low mass resolution instruments.

Note the use of the word ‘nominal’ when describing ionic masses so far.  As it turns out, exact masses are not the same was nominal masses.  For one thing, protons and neutrons do not have the same mass; their mass differs by about one part in a thousand.  More importantly, combining them into nuclei changes some of their mass into energy via Einstein’s famous equation, E=mc².  This ‘binding energy’ makes the nucleus lighter than its component protons and neutrons, and different nuclei have different binding energies, and therefore different exact masses.  So, for example, the mass resolution required to resolve a molecule of hydrogen, ¹H₂, from deuterium (²H) atom (a hydrogen atom with a neutron in its nucleus) is about 1350.  As a good working definition, high mass resolution is mass resolution high enough to resolve ions with the same nominal mass (called “isobaric interferences” by mass spectrometrists) as a result of the small differences in real mass caused by their binding energy.

The trouble with this definition, of course, is that the mass resolution required to separate isobaric interferences, depends on what they are. For example, see the isobaric interferences in figure 1.

Figure 1. SHRIMP mass spectrum of atomic and molecular peaks at mass 56 in San Carlos Olivine (Mg₂SiO₄ with Fe and Ca substituting for Mg). Green and purple: nominal M/dM = 5000; orange and blue: nominal M/DM = 15000; green and orange: Faraday cup; blue and purple, electron multiplier.

This figure contains mass scans taken on both high (~5000) and higher (~15000) mass resolution.  The mass resolution required depends on the interference.  For example, the resolution required to resolve ²⁸Si₂ from ⁴⁰Ca¹⁶O is more than ten times higher (~15000) than the mass resolution required to resolve ⁵⁶Fe from ²⁴Mg¹⁶O₂ (~1500). Note that increasing mass resolution also decreases signal intensity.

The easiest way to avoid interferences is to not create them in the first place.  This can be done by chemists who purify the element whose isotopes they wish to measure, or by material scientists who make pure compounds without trace elements (such as Ca in the above figure).

Another technique is to use an ionization source that doesn’t produce many molecular ions.  So high mass resolution is most useful when a compositionally complex material is ionized using a method that creates all sorts of complex species.

This is why high mass resolution mass spectrometry is popular in geological SIMS analyses.  Minerals generally contain a wide variety of minor and trace elements, and the SIMS ionization produces all sorts of molecular fragments.  So being able to resolve species based on their mass defects is extremely useful.

Do you really need a Nobel prize to know...

that heat is most easily lost from the head on cold January mornings...