Appendix B
Astronomers and Units

 B.1 Brief History of Unit Systems
 B.2 Working with Units

I hope all Americans will do everything in their power to introduce the French metrical system. (...) I look upon our English system as a wickedly, brain-destroying system of bondage under which we suffer. The reason why we continue to use it, is the imaginary difficulty of making a change.
 
(William THOMSON, Lord Kelvin;  Thomson, 1889)

B.1 Brief History of Unit Systems

The necessity to unify disparate measures. During Antiquity and the Middle Age, measures of weight, length and duration were varying from one place to another. The first government in History to try and homogenize measures was under king Henri III, in England. The 1297 version of the Magna Carta (originally signed in 1215), expressed the will to define standards for measuring weights and distances. “Article 25. One measure of Wine shall be through our Realm, and one measure of Ale, and one measure of Corn, that is to say, the Quarter of London; and one breadth of dyed Cloth, Russets, and Haberjects, that is to say, two Yards within the lists. And it shall be of Weights as it is of Measures.” The rest of the world kept using different yards and pounds, for several centuries.

Measuring the Earth. In France, the Académie royale des sciences was funded in 1666 by Jean-Baptiste COLBERT, under king Louis XIV, influenced by his secretary for sciences and arts, Charles PERRAULT (Débarbat & Quinn, 2019). In 1667, the royal observatory was created and astronomers were tasked with providing more accurate maps of the realm. Clergyman Jean PICARD performed a series of measures by triangulation, with a unique measuring board, and estimated the size of the Earth (Mesure de la Terre, 1671). During the XVIIIth century, several scientific expeditions in Latin America or the North Pole (by La Condamine, Maupertuis, et al.) refined the measurement of the size of the Earth and confirmed its flattening around the poles.

The introduction of the metric system. It is only during the French revolution (1789–1799) that the metric system was introduced. It was consistent with its time. It was aimed at erasing differences in an abstract way, with new standards independent of the old human references (foot, inch, etc.). One of its important features was that it was a decimal system, simplifying calculations. It can be traced back to the months before the revolution, in 1789. The cahiers de doléances (register of grievances) expressed the wish to have unified measures throughout the realm. In 1790, Charles-Maurice DE TALLEYRAND-PÉRIGORD, a bishop elected at the recent national assembly, submitted a memoir to adopt a new system of weights and measures, contributed by Marie-Jean-Antoine CRITAT DE CONDORCET and Joseph-Jérôme LEFRANÇOIS DE LALANDE (Débarbat & Quinn, 2019). In 1792, Jean-Baptiste DELAMBRE and Pierre-François MÉCHAIN were charged with measuring the length of the meridian between Dunkerque and Barcelona (Alder, 2015). The definition of the meter was then 110000000 of the distance between the North Pole and the equator. The kilogram was defined as the mass of one cubic decimeter of water.

Difficulty of adoption. The metric system was not adopted right away, even in France. It was mocked by Napoléon, although Laplace promoted the advantage of its decimal system to him. For a while, we kept “mesure usuelles”, which were a standardization of imperial units. In Germany, around 1830, Carl-Friedrich GAUSS formalized the metric unit system in physics, and proposed to add the seconds to meters and kilograms, leading to the CGS system (Centimetre-Gram-Second). During the 1851 World’s fair in London, France promoted the metric system to the world. It led to the “Treaty of the Metre”, signed in Paris in 1875, by seventeen countries adopting the metric system. Great Britain, the Netherlands and Portugal were opposed. England, in particular, felt that adopting the French system would be a political defeat. A diplomatic solution was proposed in adopting the Greenwich meridian, during the 1883 Geodetic Congress: “The Conference hopes that, if the whole world is agreed upon the unification of longitudes and hours in accepting the Greenwich meridian as the point of departure, Great Britain will find in this fact an additional motive to take on her side new steps in favour of the unification of weights and measures, by joining the Metrical Convention of May 20, 1875” (The Geodetic Congress, 1883). The Greenwich meridian was finally adopted during the 1884 International Meridian Conference, but England did not adopt the metric system

The international system. Since 1960, the International System of units (SI) is the Meter-Kilogram-Second-Ampere system (MKSA or MKS). Astronomers are one of the last communities to use CGS units (and the Gaussian system for electrodynamics; Table B.4). The continued use of a mix between imperial units and the CGS system is counterproductive. The most dramatic example is the crash of the 1999 Mars Climate Observer probe, because of a conversion mistake between imperial and metric units (Popular Mechanics, 2017). Table B.1 gives the correspondence between the MKS and CGS systems, and Tables B.2 – B.3 list the fundamental constants in both systems. Table B.4 compares the MKSA and Gaussian systems for electrodynamics.






Quantity

International units (MKS)
Astronomer’s units (CGS)





GENERAL





Length

1 m (meter) 102 cm (centimeter)





Force

1 N (Newton) 105 dyn (dynes)





Energy

1 J (joule) 107 erg





Power

1 W (watt) 107 erg/s





Flux (1 Jansky)

1026 W/m2/Hz 1023 erg/s/cm2/Hz





ELECTROMAGNETISM





Charge

1 C (Coulomb) 2.99792458 × 109 esu (electrostatic unit)





Current

1 A (Ampere) 2.99792458 × 109 esu/s





Electric potential

1 V (volt) 1299.792458 statV (statvolt)





Electric field

1 V/m 129979.2458 statV/cm





Magnetic field

1 T (Tesla) 104 G (Gauss)





Magnetic flux

1 Wb (Weber) 108 G.cm2





Auxiliary field H

1 A/m 4π × 103 Oe (Oersted)





ANGULAR DISTANCE





1 arcsec

4.848 × 106 rad





1 arcsec at 1 Mpc

4.848 pc





Table B.1: Unit conversion.






Quantity

International units (MKS)
Astronomer’s units (CGS)





UNIVERSAL





Light speed, c

2.999792458 × 108 m/s 2.999792458 × 1010 cm/s





Newton constant, G

6.67428 × 1011 m3/kg/s2 6.67428 × 108 cm3/g/s2





Planck constant, h

6.62606896 × 1034 J.s 6.62606896 × 1027 erg.s

h2π

1.054571628 × 1034 J.s 1.054571628 × 1027 erg.s





Magnetic constant,

μ0 4π×107

1.2566370614 × 106 N/A2





Electric constant,

𝜖0 1μ0c2

8.854187817 × 1012 F/m





Elementary charge, e

1.602176487 × 1019 C 4.8065295 × 1010 esu





ATOMIC





Electron mass, me

9.10938215 × 1031 kg 9.10938215 × 1028 g

mec2

8.18710438 × 1014 J 8.18710438 × 107 erg

mec2e

0.510998910 MeV





Proton mass, mp

1.672621637 × 1027 kg 1.672621637 × 1024 g

mpc2

1.503277359 × 1010 J 1.503277359 × 103 erg

mpc2e

0.938272013 GeV





Rydberg,

R α2m ec2h

10973731.568527 m1 109737.31568527 cm1

R c

3.289841960361 × 1015 Hz

R hc

2.17987197 × 1018 J 2.17987197 × 1011 erg

Rhce

13.60569193 eV





MACROSCOPIC





Boltzmann constant, k

1.3806504 × 1023 J/K 1.3806504 × 1016 erg/K

ke

8.617343 × 105 eV/K

keh

69.50356 × 1010 Hz/K





Atomic mass unit,

mu m(12C)12

1.660538782 × 1027 kg 1.660538782 × 1024 g





Avogadro number, NA

6.02214179 × 1023 mol1





Molar gas constant,

R kN A

8.314472 J/mol/K 8.314472 × 107 erg/mol/K

Table B.2: Fundamental constants.






Quantity

International units (MKS)
Astronomer’s units (CGS)





GENERAL





Astronomical unit, a.u.

1.495979 × 1011 m 1.495979 × 1013 cm





Parsec, pc 1a.u.1

3.085678 × 1016 m 3.085678 × 1018 cm





SOLAR SYSTEM





Solar radius, R

6.9599 × 108 m 6.9599 × 1010 cm





Solar mass, M

1.9889 × 1030 kg 1.9889 × 1033 g





Solar luminosity, L

3.846 × 1026 W 3.846 × 1033 erg/s





Earth radius, R

6.378140 × 106 m 6.378140 × 108 cm





Earth mass, M

5.974 × 1024 kg 5.974 × 1027 g





GALAXY





Solar velocity around G.C., Θ

220 km/s





Distance sun-G.C., R

8.0 kpc





Local disk density, ρdisk

3 12 × 1021 kg/m3 3 12 × 1024 g/cm3

n disk

1 5 × 106 m3 1 5 cm3





Local halo density, ρhalo

2 13 × 1022 kg/m3 2 13 × 1025 g/cm3

n halo

10 60 × 104 m3 0.1 0.6 cm3





COSMOLOGY





Hubble expansion rate, H0

71 km/s/Mpc





Critical density, ρc 3H028πG

1.399062 × 1011 M /Mpc3 9.472 × 1030 g/cm3





Pressureless matter density,

ΩM ρMρc

0.15 ΩM 0.45





Baryon density, ΩB ρBρc

0.019 ΩB 0.046





Cosmological constant,

ΩΛ Λc23H 02

0.6 ΩΛ 0.8

Table B.3: Astronomical constants.




Quantity

Rationalized MKSA Gaussian units



Lorentz Force

F = q E + v B F = q E + v c B

dF dV = ρE + j B dF dV = ρE + j c B




Dielectric Constant & Permeability

𝜖0 = 107 4πc2μ0 = 4π107 𝜖 0 = 1μ0 = 1




Displacement & Magnetic Field

D = 𝜖E + P D = 𝜖E + 4πP

H = B μ M H = B μ 4πM




Maxwell Equations

.D = ρ .D = 4πρ

H = j + D t H = 4π c j + 1 c D t

E + B t = 0 E + 1 c B t = 0

.B = 0 .B = 0




Poynting Vector

P = E H P = c 4πE H




Electromagnetic Power

P = SP.dS P = SP.dS




Energy Density

U = 1 2 𝜖E2 + B2 μ U = 1 8π 𝜖E2 + B2 μ




Table B.4: Classical electrodynamics.

B.2 Working with Units

My personal experience in working with units led me to the following advices.

Adopt specific units for each problem,
so that the quantities one have to deal with are close to unity (in orders of magnitude). This is particularly important to avoid numerical problems. In ISMology, the μm is a good wavelength unit, and the cm3 a good density unit, and the pc a good distance unit.
Deciding which quantity should be logarithmic
must be based on the way the uncertainty on this quantity has been estimated. The conversion between linear and logarithmic quantities indeed is not straightforward. Keeping the same 1σ range is a good practice, but it is not rigorously equivalent:
log X ± σlog X X(1 10σlog X )X+(10σlog X 1)X. (B.1)

The underlying probability law is different in both cases. It is preferable to choose a representation of the quantity so that its uncertainty is the closest to a normal law. If we are estimating the uncertainty on a parameter, the dynamical range is important. If a quantity varies by more than one order of magnitude, it is often a good choice to treat the logarithm of this quantity in a Bayesian model. Concerning fluxes, the magnitude system is logarithmic, but it is not a decimal system (it uses a cumbersome 2.5 factor), and it relies on arbitrary zero-point fluxes (cf. Table B.5). The only useful formula concerning magnitudes is to get out of them:

Fν(λ0) = Fν,0 × 100.4m(λ0). (B.2)
Perform conversions through the SI,
and avoid CGS units, which are are already deprecated outside astronomy. CGS are indeed boomer units, so is the Gaussian unit system in electrodynamics. In terms of computing, it is important to have a reliable conversion module in the different programming languages that one uses, to avoid stupid mistakes.




Central wavelength, λ0 Zero-point flux, Fν,0



U

0.36 μm 1884 Jy



B

0.44 μm 4646 Jy



V

0.55 μm 3953 Jy



R

0.66 μm 2875 Jy



I

0.80 μm 2241 Jy



J

1.25 μm 1602 Jy



H

1.60 μm 1010 Jy



K

2.18 μm 630 Jy



L

3.45 μm 278 Jy



M

4.75 μm 153 Jy



N

10.6 μm 36.3 Jy



Table B.5: The magnitude system.