PreprintPDF Available

On the Anthropogenic and Natural Injection of Matter into Earth's Atmosphere

Authors:

Abstract and Figures

Every year, more and more objects are sent to space. While staying in orbit at high altitudes, objects at low altitudes reenter the atmosphere, mostly disintegrating and adding material to the upper atmosphere. The increasing number of countries with space programs, advancing commercialization, and ambitious satellite constellation projects raise concerns about space debris in the future and will continuously increase the mass flux into the atmosphere. In this study, we compare the mass influx of human-made (anthropogenic) objects to the natural mass flux into Earth's atmosphere due to meteoroids, originating from solar system objects like asteroids and comets. The current and near future significance of anthropogenic mass sources is evaluated, considering planned and already partially installed large satellite constellations. Detailed information about the mass, composition, and ablation of natural and anthropogenic material are given, reviewing the relevant literature. Today, anthropogenic material does make up about 2.8 % compared to the annual injected mass of natural origin, but future satellite constellations may increase this fraction to nearly 40 %. For this case, the anthropogenic injection of several metals prevails the injection by natural sources by far. Additionally, we find that the anthropogenic injection of aerosols into the atmosphere increases disproportionately. All this can have yet unknown effects on Earth's atmosphere and the terrestrial habitat.
Content may be subject to copyright.
On the Anthropogenic and Natural Injection of Matter into Earth’s Atmosphere
Leonard Schulza,
, Karl-Heinz Glassmeiera,b
aInstitut ur Geophysik und extraterrestrische Physik, Technische Universit¨at Braunschweig, 38106 Braunschweig, Germany
bMax-Planck-Institut ur Sonnensystemforschung, 37077 ottingen, Germany
Abstract
Every year, more and more objects are sent to space. While staying in orbit at high altitudes, objects at low altitudes
reenter the atmosphere, mostly disintegrating and adding material to the upper atmosphere. The increasing number of
countries with space programs, advancing commercialization, and ambitious satellite constellation projects raise concerns
about space debris in the future and will continuously increase the mass flux into the atmosphere. In this study, we
compare the mass influx of human-made (anthropogenic) objects to the natural mass flux into Earth’s atmosphere due
to meteoroids, originating from solar system objects like asteroids and comets. The current and near future significance
of anthropogenic mass sources is evaluated, considering planned and already partially installed large satellite constella-
tions. Detailed information about the mass, composition, and ablation of natural and anthropogenic material are given,
reviewing the relevant literature. Today, anthropogenic material does make up about 2.8 % compared to the annual
injected mass of natural origin, but future satellite constellations may increase this fraction to nearly 40 %. For this case,
the anthropogenic injection of several metals prevails the injection by natural sources by far. Additionally, we find that
the anthropogenic injection of aerosols into the atmosphere increases disproportionately. All this can have yet unknown
effects on Earth’s atmosphere and the terrestrial habitat.
Keywords: Atmosphere, Satellite constellations, Mass influx, Human-made injection, Anthropogenic effect,
Meteoroids, Ablation, Meteorite composition
1. Introduction
Earth’s atmosphere is subject to a constant bombard-
ment by various objects from space. Most are of natural
origin, i. e. meteoroids (in the following, the term mete-
oroids refers to objects of natural origin without any size
limit) from comets, asteroids, and even differentiated bod-
ies. With the exploration of space, anthropogenic objects
like spacecraft and rocket bodies in orbit around Earth
also enter the atmosphere. Upon reentry, bodies heat up
and ablate depending on their physical and chemical prop-
erties. This way, matter in form of atoms and aerosols is
injected into the atmosphere.
With the steady growth of spaceflight activities with
evermore nations operating space programs and the in-
crease of commercialization, more and more objects are
launched into orbit around Earth. This has raised major
concerns about space debris (Klinkrad,2006). As a result,
standards have been introduced to minimize the amount
of orbital debris (ISO Central Secretary,2019) and space
agencies like ESA and NASA have introduced guidelines
and requirements, largely accepting those standards (see
for example ESA,2008;NASA,2019). A consequence of
these guidelines is that payload launched into low Earth
Corresponding author
Email addresses: l.schulz@tu-bs.de (Leonard Schulz),
kh.glassmeier@tu-bs.de (Karl-Heinz Glassmeier)
orbit (LEO) has to be disposed of within 25 years after
end of operation. This is achieved by reentry into the at-
mosphere. Hence, more and more anthropogenic material
is injected into the atmosphere, raising questions about its
significance in comparison to the natural injection caused
by the ablation of meteoroids, and about possible impacts
on the atmosphere itself.
Several companies have proposed large satellite con-
stellations of hundreds to thousands of small spacecraft in
LEO providing global internet and other telecommunica-
tion services (Liou et al.,2018). The amount of spacecraft
to be launched combined with their limited lifetime will
dramatically increase the anthropogenic amount of mass
reentering Earth’s atmosphere in the future. Thus, the
future influx caused by those satellite constellations needs
to be considered in more detail.
In this study, we provide a first overview and compar-
ison of the natural and anthropogenic injection of matter
into Earth’s atmosphere. We focus on the mass influx, its
elemental composition, and the resulting ablation prod-
ucts injected into the atmosphere. This is done separately
for natural injection (Section 2) and anthropogenic injec-
tion (Section 3). The necessary information to qualify the
mass influx, the elemental composition, and the ablation
processes has been acquired from many published studies,
partly providing conflicting numbers and information. We
try a best effort summary of all the available information.
Preprint submitted to Advances in Space Research August 29, 2020
With this information we provide a review of the nat-
ural injection and three different scenarios for the anthro-
pogenic injection. These scenarios include a present day
analysis as well as two near-future scenarios taking into
account different planned projects for large satellite con-
stellations. This allows to compare the relative contribu-
tions of human-made objects and natural objects entering
the atmosphere.
2. Natural injection
Many meteoroids originating from asteroids, comets,
material of planetary origin, interplanetary and even in-
terstellar dust (see for example Jewitt,2000;Plane et al.,
2017) enter Earth’s atmosphere every day. In this section,
we look at the mass, composition and ablation of these me-
teoroids and estimate the resulting injection with respect
to ablation products and the elemental composition.
2.1. Yearly natural mass influx
The knowledge of the total natural mass flux into Earth’s
atmosphere is of high importance. The mass influx distri-
bution is sort of bimodal with a maximum at a particle
mass of about 108kg (e. g. Flynn,2002;Carrillo-S´anchez
et al.,2015;Plane et al.,2017) and a second maximum at
high particle masses, although the mass influx increases
continuously for large objects. Objects in the size range of
a few millimeters to meters, which are the main source of
meteorites found on Earth, only contribute a small frac-
tion of the whole mass (Flynn,2002). The mass influx
distribution used in this study is displayed in Figure 1.
The distribution has two mass ranges with rather high
mass input, a small dust particle contribution in the sev-
eral microns to mm-size range and large meteoroids in the
tens of meters range. Differentiation of these ranges is
crucial because of the different origin, composition and
ablation of those two groups (see Sections 2.2,2.3). The
peak at small meteoroid sizes of several microns to mil-
limeters is caused by the large amount of interplanetary
dust particles (IDPs) in the solar system. The particles
themselves have very low masses, but are high in num-
ber. They have several sources, mainly various types of
comets and the asteroid belt, whereas the contribution of
interstellar material is negligible (Plane et al.,2017).
Dust particles are normally defined to be smaller than
tenth of microns (Rubin and Grossman,2010;Koschny
and Boroviˇcka,2017). However, based on the analysis
and modelling of the observations of the zodiacal dust
cloud by the Infrared Astronomical Satellite (IRAS) and
ground based radars as well as various other observations
(Nesvorn´y et al.,2010,2011) we adopt a cutoff size of
2 mm. Therefore, the upper mass limit of the IPD popu-
lation is roughly 105kg (Fig. 1).
In contrast, the mass flux peak at high impactor sizes
is caused by their high mass, while their impact rate is
quite low and decreases with increasing size. Bodies heav-
ier than hundreds of tons (larger than several meters in di-
ameter) hit Earth once a year, while impacts with objects
several ten meters in diameter occur only once in a thou-
sand years (Chapman and Morrison,1994;Zolensky et al.,
2006a). However, upon entering Earth’s atmosphere, large
impactors ablate and disintegrate, leaving behind a trail
of aerosols and particles of molecular size. Ablation ma-
terial in the atmosphere in form of dust particles seems
to sediment within several months (Klekociuk et al.,2005;
Gorkavyi et al.,2013), while we can not rule out that ions
and particles of molecular size remain for a longer time in
the upper atmosphere. Thus, we include large bodies that
impact Earth at least every 10 years to account for such
ablation and disintegration processes.
Drolshagen et al. (2017) have calculated the mass in-
flux of meteoroids in a size range of 1021 to 1011 kg.
Their mean model for masses below 102kg is based on
the widely used interplanetary flux model of Gr¨un et al.
(1985), which is used by NASA (Moorhead,2020) and is
close to the newest ESA meteoroid flux model IMEM2
(Soja et al.,2019). It is derived from different spacecraft
in-situ measurements of meteoroids and zodiacal light as
well as lunar impact measurements. The model provides
an analytic function of the particle flux at 1 AU. For the
intermediate mass range, 102to 106kg, the power law
model of Brown et al. (2002) is used, which is based on
spacecraft fireball data. For large bodies heavier than 106,
a similar power law adapted from Stokes et al. (2003) is
used.
Beside the named studies, Drolshagen et al. (2017) have
included measurements from the Hubble Space Telescope
solar array impacts of meteoroids (McDonnell,2005) as
well as visual data from meteor entries (Koschny et al.,
2017) to verify the model of Gr¨un et al. (1985). Addition-
ally, studies from Halliday et al. (1996) (fireball data) and
Suggs et al. (2014) (lunar impact flashes) were used to find
the best way to connect the models of Gr¨un et al. (1985)
and Brown et al. (2002). As Drolshagen et al. (2017) only
briefly reviewed other studies that also provide estimates
on the annual mass influx we shortly discuss these other
studies.
For the large impactor size range, we regard the models
by Brown et al. (2002) and Stokes et al. (2003) as the cur-
rent best models. However, for the IDP mass range, there
is a large difference between the various estimates pro-
posed. Plane (2012) reviews several studies, experiments,
and models regarding the influx of IDPs into Earth’s at-
mosphere. The mass influx estimates vary from 1,800 to
100,000 t/yr for the respective mass range. Thus, the IDP
mass range needs a more critical discussion to derive a suit-
able estimate for the purpose of our study. Four studies
are important here.
Nesvorn´y et al. (2010) modelled the zodiacal dust cloud
using IRAS data. They estimated an influx of 100,000 t/yr.
A later update (Nesvorn´y et al.,2011), using refined or-
bital characteristics of the IPDs based on meteor radar
2
10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 100101102
Meteoroid diameter (m)
10-20 10-15 10-10 10-5 1001051010
Meteoroid mass (kg)
0
500
1000
1500
2000
2500
3000
Mass influx (t / yr / mass decade)
IDPs
Meteorites & Large impactors
Large impactors (impact
frequency < 0.1/yr)
Grün et al. (1985)
Brown et al. (2002)
Stokes et al. (2003)
Figure 1: Variation of the yearly mass influx distribution of Earth’s atmosphere with meteoroid mass. The particle diameter is calculated by
using a density of 2500 kg/m3and assuming a spherical shape. The blue data, mainly covering the IDP mass range, are based on the flux
law of Gr¨un et al. (1985). The orange data, covering the meteorite and large impactor mass range, are calculated using the power law from
Brown et al. (2002) with a mean impact velocity of 20km/s. The grey data is calculated using the power law of Stokes et al. (2003), given in
Drolshagen et al. (2017), again using the same average velocity. It covers large impactors which impact Earth less than every 10 years. The
influx distribution reflects the mean model from Drolshagen et al. (2017) and was calculated as shown in Appendix A. Exact values are given
there, too. The annual mass input is 11,509 t for the Gr¨un et al. (1985) model, 863 t from Brown et al. (2002), and 4,349 t from Stokes et al.
(2003).
data, proposes a more realistic value of about 15,000 t/yr.
Hughes (1978, pp. 148–157) used satellite, radar and
visual data of IDP and micrometeor entries to reach an
estimate. The IDP mass influx rate of 16,100 t/yr in a
mass range of 1017 to 101kg is widely accepted and
part of the mass influx distribution presented by Flynn
(2002).
Mathews et al. (2001) analyzed observations of microm-
eteor entries into the upper atmosphere measured by the
high power low aperture (HPLA) radar at the Arecibo ob-
servatory. They determine the mass and speed of entering
particles, reaching estimates of 1,600 and 2,700 t/yr for the
mass range 108to 101kg. These are the lowest annual
influx values of all those studies reviewed in Plane (2012).
Their mean geocentric velocity of 50 km/s, endorsed by
later measurements with Arecibo (Janches et al.,2006),
is much higher than in other observations (Hughes,1978,
pp. 150–155; Gr¨un et al.,1985;Nesvorn´y et al.,2010,2011;
Koschny et al.,2017). Such a large velocity raises ques-
tions as it implies that the majority of the dust particles
moves retrograde in the solar system. This could be due to
HPLA radars being unable to detect very slow (<15 km/s)
and small particles. The lack of detection of smaller parti-
cles could explain the quite low influx estimate. von Zahn
(2005) discusses further possible shortcomings and biases.
Here, we assume that Mathews et al. (2001) significantly
underestimate the amount of incoming meteors.
Love and Brownlee (1993) measured IDP and mete-
oroid impacts on the Long Duration Exposure Facility
(LDEF). Impact crater size, depth and number were de-
termined in order to obtain information about the mass of
each impactor. Assumptions had to be made concerning
3
the particle properties, velocity and impact angle. The
integration of the derived mass distribution yields approx.
27,000–40,000 t/yr (Love and Brownlee,1993;Taylor et al.,
1998;Mathews et al.,2001).
The estimate of Love and Brownlee (1993) has several
uncertainties as it is based on a mean geocentric particle
velocity of 16.9 km/s. For further detail on the velocity
we refer to Gr¨un et al. (1985); Taylor (1995,1996); Taylor
and Elford (1998); Brown et al. (2005); Drolshagen et al.
(2008); ECSS (2008); Nesvorn´y et al. (2010,2011), and
Carrillo-S´anchez et al. (2016). After reviewing all the men-
tioned literature, we use the higher geocentric mean veloc-
ity (normalized on mass) of 20 km/s for IDPs, as the eval-
uations of Taylor and Elford (1998); Brown et al. (2005)
seem to incorporate the best combination of relevance and
experimental falsification. A similar velocity is also used
by Drolshagen et al. (2017).
Increasing the mean velocity implies a reduction of par-
ticle mass. This results in a shift of the mass distribu-
tion towards lower masses and reduces the mass influx
value. Also considering Borin et al. (2009); Cremonese
et al. (2012), a realistic mass influx estimate is a value
below 20,000 t/yr.
2.1.1. Resulting mass influx and mass distribution
Results of the studies discussed roughly agree with the
mass influx estimate by Drolshagen et al. (2017). The
Gr¨un et al. (1985) flux model is still widely accepted, and
we use it to determine an annual mass influx for the differ-
ent mass decades. For the higher mass ranges, the power
laws by Brown et al. (2002) and Stokes et al. (2003) are
used (Figure 1). For further details see Appendix A.
Thus, considering the mass range of objects impact-
ing Earth less than every 10 years, 1021 to 106kg, we
yield an annual mass influx of 12,372 t with most of the
mass (11,509 t) caused by particles with masses lower than
102kg. The mass flux is dominated by the IDP contri-
bution of 10,856 t/yr. For all the mass ranges studied con-
siderable differences in the estimation of the mass influx
exist. With an error factor 2 we yield a range of 6,186 to
24,746 t/yr for the mass influx.
2.2. Composition of the natural material
IDPs and larger meteoroids have quite different compo-
sitions due to different origins. The major contribution to
the IDP flux is thought to originate from Jupiter Family
Comets (JFCs) (Zolensky et al.,2006b;Nesvorn´y et al.,
2010;Jenniskens,2015, pp. 282–283; Yang and Ishiguro,
2015;Carrillo-S´anchez et al.,2016). The composition of
IDPs has been examined comprehensively in two studies:
Schramm et al. (1989) examined 200 IDPs on their major
element composition and Arndt et al. (1996) gathered data
on 89 IDPs covering elemental abundances also for mi-
nor elements. Both studies only give relative abundances,
normalized to Si, Fe, and CI carbonaceous chondrite class
abundance. The CI abundance is representative of the so-
lar system abundances of elements (Anders and Ebihara,
1982;Anders and Grevesse,1989) and used for normaliza-
tion. The only element with a significant mass fraction
not determined by the two above mentioned studies is hy-
drogen. Here, we use the 2.02 wt% value from Anders and
Grevesse (1989). The absolute elemental mass abundances
used in following discussions are listed in Table 1. For fur-
ther considerations we use mean values. To summarize,
a significant fraction of the mass contribution are metal
(31 %) and metalloid elements (13 %), while the majority
is nonmetallic (47 %). About 8% of the mass could not be
assigned to an element.
Table 1: Elemental composition of IDPs and meteorites.
Z El. Unit Arndt
et al.
(1996)
Schramm
et al.
(1989)
Mean
IDP
Meteo-
rites
1 H µg/g 20200a186
3 Li µg/g 2
4 Be ng/g 31
5 B ng/g 377
6 C wt% 9.7 9.7 0.3
7 N µg/g 69
8 O wt% 30.7 30.7 35.2
9 F µg/g 97
11 Na µg/g 3747 5503 4625 5936
12 Mg wt% 9.3 11.0 10.1 13.6
13 Al wt% 1.2 0.9 1.1 1.2
14 Si wt% 12.9 12.9 17.1
15 P µg/g 1746 1746 1292
16 S wt% 3.6 5.3 4.4 2.1
17 Cl µg/g 1229 1229 188
19 K µg/g 540 540 764
20 Ca wt% 0.3 1.0 0.6 1.3
21 Sc µg/g 12 12 8
22 Ti µg/g 549 549 676
23 V µg/g 74 74 71
24 Cr µg/g 2199 3590 2894 3466
25 Mn µg/g 1644 1644 2355
26 Fe wt% 17.3 17.9 17.6 25.9
27 Co µg/g 337 337 808
28 Ni wt% 0.5 0.7 0.6 1.7
29 Cu µg/g 186 186 91
30 Zn µg/g 405 405 57
31 Ga µg/g 18 18 7
32 Ge µg/g 42 42 13
33 As µg/g 15 15 4
34 Se µg/g 32 32 8
35 Br µg/g 81 81 1
36 Rb µg/g 6 6 2
37 Sr µg/g 16 16 10
39 Y µg/g 2 2 2
40 Zr µg/g 18 18 7
>40 µg/g 16
Total (%) 46.4 78.0 91.3 100.0
Column 4 to 6 are derived from Arndt et al. (1996) and Schramm
et al. (1989) with the sixth column representing the arithmetic mean
of both studies. Column 7 is the meteorite composition, taking into
account the meteorite portion of each group (Table 2) and the re-
spective meteorite group elemental mass abundances. For further
details on the derivation see the text and Appendix B.
aThe abundance of hydrogen is estimated as described in the text.
By contrast, large meteoroids are mostly of asteroidal
origin (Bottke et al.,2002;Binzel et al.,2015), only a
4
small portion originates from comets (Binzel et al.,2004;
Fern´andez et al.,2005;DeMeo and Binzel,2008) or dif-
ferentiated bodies (Grady,2000;Boroviˇcka et al.,2015, p.
258; Russell et al.,2015, p. 419). A significant amount of
material can survive upon entry and reaches the ground
as meteorites. Therefore, we estimate the average elemen-
tal mass abundance of large meteoroids by calculating the
composition of meteorites found on Earth. Meteorites are
divided into different classes, based on their mineralogy
and thus elemental composition. By weighting the com-
position of each meteorite class with their respective fre-
quency of finds and falls on Earth, we calculate an average
meteorite composition. Due to the large amount of clas-
sified meteorites of more than 22,000, this statistical ap-
proach is possible. We use the data given in Grady (2000)
to yield the frequency of each meteorite class (see Table 2).
For the average elemental mass abundances we use data
from Wasson (1974); Lodders and Fegley (1998); Mittle-
fehldt et al. (1998); Demidova et al. (2007). The derived
average composition of each meteorite class as well as de-
tails on the data are provided in Appendix B. By weighting
the elemental mass abundances with the class frequencies
of Table 2, we yield the overall elemental mass abundance
of meteorites listed in Table 1.
The metal (45 %) and metalloid (17 %) elemental abun-
dance is higher than in IDPs, while the non-metallic por-
tion (38 %) is lower. All in all, IDPs and meteorites show
considerable differences, but are similar in the abundance
of some elements.
The above described method of determination of the
composition of the large meteoroids is biased, as mete-
oroids show different ablation rate and behaviour depend-
ing on their composition upon entry into the atmosphere.
Thus, the amount of produced meteorites and the final
composition to some extend depends on the properties
of the initial meteoroids. Additionally, some meteorite
classes are easier to find, e. g. iron meteorites are easier
to distinguish from the environment due to their metal-
lic look. Also, the number of meteorite finds and falls for
some groups is quite low and the statistical sample might
be insufficient in that case. Thus, there are uncertainties
in our approach.
2.3. Atmospheric processing of the natural material
Results from various studies are used to derive ablation
products of IDPs and larger meteoroids. Three ablation
products are thought to be important: material due to
deposition in the atmosphere in form of atoms, ions or
molecules; material deposited as aerosols, e.g. particles of
microns to nm size; material directly reaching the ground,
thus not contributing to atmospheric injection.
2.3.1. Small meteoroid ablation
For small meteoroids, several studies suggest that there
is a cutoff size below which no ablation is taking place,
due to insufficient heating of the particle. This cutoff size
Table 2: Frequencies of meteorite classes derived from finds and falls
(Grady,2000).
Class Portion (%)
Chondrites 91.95
Ordinary Chondrites 86.35
H 42.27
L 37.72
LL 6.36
Carbonaceous Chondrites 3.40
CH 0.08
CI 0.04
CK 0.54
CM 1.18
CO 0.62
CR 0.57
CV 0.36
Enstatite Chondrites 1.22
EH 0.93
EL 0.28
Other Chondrites 0.99
K (Kakangari) 0.13
R (Rumurutiite) 0.85
Achondrites 3.69
Acapulcoites 0.07
Angrites 0.02
Aubrites 0.28
Brachinites 0.04
Lodranites 0.09
Ureilites 0.56
Winonaites 0.07
From Vestaa2.36
Diogenites 0.57
Eucrites 1.22
Howardites 0.57
Lunar 0.11
Lunaite 0.11
Martian 0.09
Shergottites 0.03
Nakhlites 0.03
Chassignites 0.03
Stony Irons 0.52
Mesosiderites 0.29
Pallasites 0.22
Irons 3.85
IAB 0.76
IC 0.06
IIAB 0.60
IIC 0.05
IID 0.09
IIE 0.11
IIF 0.03
IIIAB 1.34
IIICD 0.24
IIIE 0.08
IIIF 0.04
IVA 0.37
IVB 0.08
Total 100
aExpected to originate from Vesta.
ranges between meteoroid masses of 1011 to 1015 kg de-
pending on the study (Jones and Kaiser,1966;Nicol et al.,
1985;Popova,2004;Vondrak et al.,2008). Meteoroids be-
low this size can be treated as part of the aerosol mass
5
fraction as they are slowed down to cm/s velocities. It
takes them weeks to years to reach the ground depending
on their size (Kasten,1968;Rietmeijer and Jenniskens,
1998;Rietmeijer,2002).
Looking at masses higher than the cutoff mass, there
are three models to be considered. Rogers et al. (2005)
present a numerical model of the ablation of small mete-
oroids in the mass range 103to 1013 kg for discrete ve-
locities and different meteoroid densities. With the knowl-
edge of the velocity distribution of meteoroids in the re-
spective size range, an overall estimate of the ablation can
be made. Taylor (1996) state that the velocity distribu-
tion of meteoroids from 102to 1015 kg is similar. We
use the velocity distribution from Taylor (1995), tabulated
in ECSS (2008), and recalculate the distribution to an in-
cident height of 100km, thereby taking into account the
acceleration due to Earth’s gravity. Values from Rogers
et al. (2005) are weighted with that velocity distribution.
To come closest to an IDP density of 2,200 kg/m3(Carrillo-
anchez et al.,2016), the mean of the ablated mass for
two different particle densities of 1,000 and 3,300 kg/m3
is taken. Thus, we yield a final value for the fraction of
ablated mass for the different mass bins (see Figure 2).
Small meteoroids with masses above 108kg show more
than 90 % ablated mass. Towards lower masses, the frac-
tion of ablated material decreases to almost zero. For this
model, we adopt a cutoff size of 1014 kg and interpolate
the data (also shown in Figure 2). By weighting with the
Gr¨un et al. (1985) mass influx, an ablated mass fraction
of 86 % is obtained. We assume that all of the ablated
material enters the atmosphere in atomic form, as recon-
densation of the vaporized material to dust is unlikely due
to the small particle masses.
Love and Brownlee (1991) have performed a similar
study, simulating the atmospheric entry of over 50,000
meteoroids. Using their data on the amount of vapor-
ized mass, equal to the mass of atoms ablated, and trans-
forming the particle diameter to mass by using a den-
sity of 3,000 kg/m3, we yield the values and interpolation
depicted in Figure 2. Here, we adopt a cutoff mass of
1013 kg. Weighting with the Gr¨un et al. (1985) mass in-
flux, an ablated mass fraction of 69 % is obtained, which
is considerably lower than the Rogers et al. (2005) esti-
mate. Both, Love and Brownlee (1991) and Rogers et al.
(2005) show a very small to zero survivability of particles
in the mass range 106to 102kg. This is supported by
Rietmeijer (2002).
A third study, Carrillo-S´anchez et al. (2016), utilizes
the chemical ablation model CABMOD, introduced by
Vondrak et al. (2008) incorporating differential ablation
of different elements along with the model of the zodiacal
cloud (Nesvorn´y et al.,2010,2011). They use a different
mass distribution and a different velocity distribution with
a lower mean velocity than used in our study. For their
mass range of 106kg to 1012 kg, only 18.2 % of the ma-
terial are ablated atoms. Taking the same mass range, the
two other study interpolations yield a fraction of 85 % and
10-20 10-15 10-10 10-5
Meteoroid mass (kg)
0
0.2
0.4
0.6
0.8
1
Atomic mass fraction
Rogers et al. (2005)
Love and Brownlee (1991)
Figure 2: Fraction of mass ablated in atomic form in dependence of
meteoroid mass. The crosses depict the data points derived from the
studies of Rogers et al. (2005) and Love and Brownlee (1991), while
the line is the interpolation.
65 %, respectively. Thus, differences to the other two mod-
els are large, partly to be explained by the slower average
velocity and the different mass distribution. In the fol-
lowing we use the simulated values of Love and Brownlee
(1991).
2.3.2. Large meteoroid ablation
For large meteoroids, ablation largely reduces the mass
of entering meteoroids. However, a substantial fraction
can survive atmospheric entry (Rietmeijer,2002, p. 236;
Boroviˇcka et al.,2015, p. 258). The survival fraction,
the ratio of the meteoroid’s terminal and initial mass, is
highly dependent on velocity, density, composition and ini-
tial mass itself. The dependence of the survival fraction
on mass is displayed in Figure 3, which is based on results
of Halliday et al. (1996); Klekociuk et al. (2005); Popova
et al. (2011,2013). Data are widely scattered due to the
various dependencies mentioned above. An average mass
dependent survival fraction is derived by fitting a scaled
Rayleigh distribution to all the given data. The starting
point of the distribution is chosen to be at 101kg with
no mass survival from 102to 101kg considering Bald-
win and Sheaffer (1971). This fits the results by Love and
Brownlee (1991); Rietmeijer (2002); Rogers et al. (2005)
and also matches our model of the small meteoroid abla-
tion. The resulting survival fraction model S(m) with
S(m) = (0,2log m < 1
0.67 log(m)+1
1.342e(log(m)+1
2·1.34 )2
,1log m8
(1)
where Sdenotes the survival fraction, and log mthe decadic
logarithm of the meteoroid initial mass, is displayed in Fig-
6
10-2 10-1 100101102103104105106107108
Meteoroid mass (kg)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
Survival fraction
Average (fitted)
Halliday et al. (1996)
Popova et al. (2011)
Klekociuk et al. (2005)
Popova et al. (2013)
Figure 3: Survival rate of large meteoroids. Shown are fireball data from different studies. Errors are depicted if available. The black line
depicts our fitted estimate; for further details see the text.
ure 3.
In a further step one needs to clarify how much of the
ablated material is deposited in the atmosphere in atomic
or aerosol form. Here, ablated material can recondense to
dust and also dust particles can leave the fireball as it is
ablated. Observations of dust clouds are very rare and in-
corporate large errors. The Chelyabinsk object created a
dust cloud of roughly 24 % of the initial mass of approx.
1.3·107kg (Popova et al.,2013). Klekociuk et al. (2005)
provide observational results from an entry of a massive
meteoroid (roughly 1.4·106kg). The dust cloud amounted
to roughly 79 % of the total meteoroid mass, at least 47 %.
TC3, a smaller bolide of around 5 ·104kg produced a dust
cloud of about 20 % (at least 15 %) of the initial mete-
oroid mass (Boroviˇcka and Charv´at,2009). Detailed mod-
elling of a fireball entry by Boroviˇcka et al. (2019) indi-
cates that more fragmentation leads to more dust being
released. This would point towards an increase in the dust
fraction for larger meteoroids as fragmentation events are
more likely for larger meteoroids. Therefore, we assume a
linear increase in the aerosol fraction with increasing log-
arithm of mass. For meteoroid masses of 102kg, all the
material is ablated in atomic form in accordance with the
findings in the previous section. The aerosol mass fraction
increases to 50 % for a mass of 107kg.
2.3.3. Resulting mass dependent ablation
The mass dependent fraction of the three different abla-
tion products is displayed in Figure 4, based on the results
from the previous two sections. For nearly every meteoroid
mass, the material is injected into the atmosphere either
in atomic or aerosol form. Only for a very limited mass
range significant amounts of the entering meteoroids reach
the ground directly upon entry into the atmosphere.
2.4. Overall natural injection
With the estimates of the mass distribution, the com-
position, and the ablation of incoming meteoroids a com-
plete picture of the injection of natural matter into Earth’s
atmosphere is available. The following estimates emerge.
12,325 t natural material are entering Earth’s atmosphere
every year. Only 48 t/yr of meteoroids are reaching the
ground (0.4 % of the whole mass) directly upon entry. The
rest is injected into the atmosphere, 8,421 t (68 % of the
whole mass) in atomic form, 3,904 t (32 %) as aerosols.
Most of the material is non-metallic (5,674 t), but metals
are also significant (4,047 t). Metalloids take the small-
est portion (1,655 t). The most abundant metals are iron
(2,295 t) and magnesium (1,300 t), other metals only con-
tribute with minor fractions, e. g. aluminum (131 t), nickel
(90 t), calcium (88 t) and sodium (59 t). Non-metallic and
metalloid elements with high injection masses are oxygen
(3,851 t), silicon (1,654 t), carbon (1,054 t), sulfur (513 t)
and hydrogen (220 t).
3. Anthropogenic injection
Since the beginning of the space age, anthropogenic
injection into the upper atmosphere occurs. Decommis-
sioned spacecraft, rocket bodies, and other debris are en-
tering Earth’s atmosphere. This is due to the aerodynamic
7
10-20 10-15 10-10 10-5 100105
Meteoroid mass (kg)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
Mass fraction
Aerosol Atomic Meteorites (Ground-reaching)
Figure 4: Estimated fractional mass ablation of meteoroids in dependence of the meteoroid mass. Values are given for each mass decade. For
details on the derivation, see the text. Note that all of those values are rough estimates.
drag of the atmosphere, which is reducing the speed of or-
biting objects even at altitudes as high as 1000 km. Most
of the space components are made of metals.
With the ongoing use of space, more and more space
debris is inserted into orbits around Earth. Space debris
has become a serious problem as it is a hazard to operat-
ing spacecraft and even the International Space Station.
All the debris in LEO can lead to a cascade effect of de-
bris impacting satellites causing even more debris and so
forth, possibly rendering whole orbits unusable for decades
to hundreds of years. In order to reduce the amount of fu-
ture space debris and ensure safety in space, guidelines
have been introduced (ESA,2008;ISO Central Secretary,
2019;NASA,2019). For LEO, satellites and upper stages
have to be de-orbited within 25 years after their end of
lifetime. Due to the increasing use of space, these require-
ments might tighten in the following years and decades.
Thus, nearly everything launched into LEO nowadays will
burn up in the atmosphere, eventually.
3.1. Today’s annual anthropogenic mass influx
We derive an estimate of the annual mass influx from
the altitude depending mass distribution around Earth as
provided by Liou et al. (2018). Due to the atmospheric
drag, there is an altitude below which, on average, all ob-
jects reenter the atmosphere within one year. With that,
the mass entering every year is the sum of the mass of all
objects below this altitude. Using Boykin and Mc Nair
(1966) together with data from Bowman (2002); Saunders
et al. (2012), we come up with an average reentry altitude
of 450 km, roughly matching calculations by Braun et al.
(2013). As a result, the annual mass influx amounts to
about 190 t/yr, a value comparable to that one estimated
by Pardini and Anselmo (2013). About 60 % of the mass
are spacecraft, 40 % rocket bodies. This mass influx value
will increase in the future as the amount of mass in orbit is
rising continuously (Liou et al.,2020). It should be noted
that space debris is large in numbers but contributes only
a negligible part to the anthropogenic mass influx.
In a further step, the core stages of launch vehicles
need to be considered. Although these are accelerated to
considerable velocities, they remain suborbital and reen-
ter the atmosphere right after liftoff. These objects are
not tracked by space agencies and therefore are not in-
cluded in the studies by Pardini and Anselmo (2013) and
Liou et al. (2018). To estimate the mass contribution by
core stages, we consider the launch history of 2019. From
each orbital launch, we consider every rocket stage that
is jettisoned into a suborbital trajectory. Using available
data on launch profiles, the approximate entry velocity of
each stage can be calculated. Only stages with entry ve-
locities higher than 3.8 km/s are taken into account as we
consider lower entry velocities to be insufficient for signif-
icant ablation and contribution to the injection rate into
the atmosphere. Additionally, we neglect launch vehicles
with a payload mass lower than 1 t as well as suborbital
rocket launches. A complete list of the data and sources
is given in Appendix C. Using the mass of each stage,
we estimate about 702 t of rocket stage mass reentering
Earth’s atmosphere in 2019, with speeds from 3.8 to km/s
to 7.6 km/s. This mass value will increase in the near fu-
ture, too. Summarizing, today’s (2019) annual mass influx
from anthropogenic sources amounts to about 890 t. The
largest contribution (87 %) are rocket bodies.
8
3.2. Satellite constellations
With evermore companies engaging in commercial space-
flight, satellite constellation projects have been proposed
and some of them already started. Mostly, these projects
aim at providing global telecommunication services for the
global internet (Liou et al.,2018). Therefore, hundreds to
thousands of satellites used as relays will be brought to
LEO and eventually, after reaching their end of lifetime,
burn up in the atmosphere. In Table 3, proposed and
(partially) realized large satellite constellation projects are
listed with additional information such as characteristics
of the satellites or the current status. In total, constella-
tions of nearly 110,000 satellites have been proposed. The
two constellations of SpaceX and OneWeb are about to op-
erate soon (Arianespace,2020;Krebs,2020d), others will
most certainly follow.
Constellation satellites are launched into relatively low
orbits because of the effective range of antennas and la-
tency. Some satellites even have to raise and retain their
orbit by themselves using on-board propulsion. Most likely,
due to the large number of satellites, some of the satel-
lites will fail in orbit due to electronic failure or propul-
sion problems. From the first Starlink launch, 3 out of
60 satellites did not seem to work properly. Thus, we as-
sume a failure rate of 5 % for the satellites, which have
to be replaced. This increases the mass estimate. Ad-
ditionally, we expect most of the satellites at high orbits
(around 1,000 km) to be de-orbited after service, as many
companies have already vowed to do so.
In order to estimate the total mass influx caused by
constellations, upper stages and core stages of the launch-
ing rockets have to be incorporated, too. Depending on
the launch vehicle, the upper stage mass relative to the
payload mass is different. Typical launch vehicles are the
Falcon 9 and the Soyuz 2.1 Fregat rockets. The Falcon 9
payload is 15.6 t (derived from Krebs,2020d), that one of
the Soyuz 2.1 Fregat 5 t (Arianespace,2020). With their
respective upper stage and core stage mass (Kyle,2018;
Arianespace,2012) one can estimate a typical ratio of pay-
load to upper (core) stage mass of 0.24 t (0.89 t) per ton
launched and reentering within years. This is included into
the subsequent calculations. Based on statements from
some companies, we expect the lifetime for each satellite
to be around 5 years. This means, every 5 years a whole
constellation has to be replaced.
In the following, we consider two scenarios for a possi-
ble future mass influx:
3.2.1. Future Scenario 1 Probable influx
This scenario assumes today’s, 2019 based mass influx.
To this influx we add that one due to satellite constella-
tions most probably installed in the future (constellation
projects from Table 3with bold and normal font type num-
ber of satellites). We expect all satellites in LEO to reen-
ter Earth’s atmosphere. However, we do not expect the
satellites of OneWeb’s 8,500 km height constellation (see
Table 3) to enter the atmosphere; only their upper and
core stages will reenter and are taken into account. All in
all, in 5 years additional 19,411 satellites as well as 585
upper stages and 443 core stages will be brought to orbit.
With the above mentioned lifetime, failure rate and upper
stage and core stage per payload mass, every year, 960 t of
satellites, 291 t upper stages, and 1,491 t core stages will
reenter the atmosphere. All in all, the annual mass influx
amounts to 2,742 t/yr. However, a significant portion of
that material may reach the ground, as discussed further
down.
3.2.2. Future Scenario 2 Maximum influx
This scenario assumes a doubling of the 2019 mass in-
flux. To this influx all of the planned satellite constellation
projects are added. However, from the projected addi-
tional 30,000 Starlink and nearly 48,000 OneWeb satel-
lites we merely assume 50 % of them being realized. Thus,
within 5 years, nearly 75,000 additional satellites, 1,984
upper stages, and 1,503 core stages are considered in this
scenario. The total mass influx is estimated at 8,114 t/yr,
consisting of 3,153 t satellites, 880 t upper stages, and 4,081 t
core stages.
3.3. Composition of the anthropogenic material
The composition of satellites and rocket bodies differs
largely from the composition of meteoroids. Generally, the
metal abundance is way higher as most structural compo-
nents are made of alloys. We distinguish between rocket
bodies (core stages that reenter right after launch and up-
per stages which reach orbit) and spacecraft.
Rocket bodies mainly consist of propulsion tanks and
rocket engines. For composition information of the propul-
sion tanks we use the information provided by Henson
(2018). Tanks have to withstand the pressure of the loaded
fuel and thus are made out of durable alloys. Most com-
mon is AA2219 aluminum alloy but newer rockets like
SpaceX’ Falcon 9 are also made out of Al-Li alloys like
AA2198 (Wanhill,2014). On the other hand, the Centaur
tank is manufactured out of 301 stainless steel. Those
rockets using solid rocket motors (e.g. Antares or Vega
launch vehicles) use D6AC steel or similar for their tanks.
Other parts like feedlines, other pressured vessels, and un-
pressurized structure are made out of steels, Al-alloys and
Ti-alloys. Non-metallic materials are rare. They are, for
example, found in thermal insulation and adapters. Based
on the available information and respecting the large vari-
ety of materials we assume the following relative composi-
tion values as a first estimate: 80 % AA2219 Al-alloy, 5 %
AA2198 Al-Li-alloy, 5 % D6AC steel, 5 % 301 steel, 5 %
others.
Liquid fuelled engines make up the other part of a
rocket body. They need even more resistant alloys, es-
pecially due to the high temperature load in combustion
chambers. According to Halchak et al. (2018), different
groups of alloys are used. For the nozzle, the most heavy
9
Table 3: Satellite constellation projects with more than 100 satellites and satellite masses greater than 10 kg.
Project Country Satellite
mass (kg)
Proposed
sat. number
Height
(km)
Project Status Sources
Starlink
(SpaceX)
USA 260 7518
1548+2824a
30000
335–346
540–570
300–600
538 launched, of that 60
prototypes (Krebs,2020c,d),
filing for 30000 more
satellites to FCC prepared
(Koksal,2019)
FCC (2018a,b,2019);
Space Exploration
Technologies
Corporation (2020)
OneWeb UK 147 720b
1280
1280 c
47128
1200
8500
8500
1200
74 satellites launched
(Arianespace,2020), LEO
extension to 47844 satellites
requested in 2020
FCC (2017,2020);
WorldVu Satellites
Limited (2018,2017,
2020)
Telesat LEO
(Telesat)
Canada 168d117+292 1000 1 prototype launched Grant (2019); FCC
(2018c)
Kepler
(Kepler Com-
munications)
Canada 12–15 140 500–600 2 prototypes (3U Cubesats)
launched
FCC (2018c); Grant
(2019)
Project
Kuiper
(Amazon)
USA 3236 590–630 FCC filing submitted in 2019 Kuiper Systems LLC
(2019)
Hongyun
(CASC)
China 247d320 1000 1 prototype launched Grant (2019)
Xingyun
(CASIC)
China 93 156 570 2 satellites launched in 2020 Grant (2019)
Satellogic Argentina 37 300 500 8 satellites launched since
2016
Lal et al. (2017)
Boeing USA 3000 1200 Boeing withdrew several
FCC filings and
Grant (2019); Lal
et al. (2017)
Samsung South
Korea
4600 300–2000 No perceptible action since
2015
Khan (2015)
MCSat
(Thales)
France 800 +4000 Beside filings at the
International
Telecommunication Union
(ITU), no information
available
Grant (2019)
3ECOM-1 Lichtenstein 288 Beside ITU filings, no
information available
Grant (2019)
The list might not be complete as companies in this area emerge and disappear rapidly. Reliable information are hard to gather. The font
used for the proposed satellite number (column 4) indicates the probability of realization: an italic font means that the realization is not safe
or improbable, normal font implies a high probability for the project realization, and a bold font means that the constellation is granted by
the FCC of the USA and will therefore most likely be materialized or is already in process. Additional information is acquired from company
websites and satellite launch data. The projects are ordered after their probability of realization and the number of spacecraft.
aSpaceX currently has approval for 1,548 satellites at 550 km altitude and 2,825 satellites at 1,110–1,325 km altitude. They have filed for a
modification of the orbit altitude to 540–570 km and reduction to 2,824 satellites. As earlier modifications of this kind were successful, it is
likely this gets granted.
bOneWeb has authorization of the launch of 720 satellites at this altitude but company statements suggest only 648 are needed.
cOneWeb filed at the FCC for a doubling of that number, but has withdrawn that request.
dPrototype mass.
part of an engine, Ni-alloys like Inconel 718 (used in the
Space Shuttle main engine RS-25) or Inconel 600 (Vulcain
2) are the most common alloys. The much used RD-107
and RD-108 engines of the Soyuz launch vehicle incorpo-
10
rate Copper-Chromium alloys (3 % Cr). A similar alloy
called Narloy-Z has been used in the combustion chamber
of the RS-25. RD-107 and RD-108 nozzles are additionally
made of stainless steel, nickel-based alloys or other metals.
Other components of engines like propellant pumps, com-
ponents of turbopumps, valves, and feed lines are made of
different alloys, typically Al-, Ti-, or Ni-based due to the
required durability. Only a minor portion is made of non-
metallic components like silicon carbide. Again, based on
the variety of different materials being in use, we use the
following estimate for the composition of the rockets’ en-
gines: Inconel 718 Ni-alloy, Inconel 600 Ni-alloy, and A286
alloy each 25%, 10 % Cu-Cr-alloy (alloy with 3% Cr), and
5 % Ti, 5 % Al, and 5 % Ni.
We additionally differentiate between core stages and
upper stages. For upper stages, the engine mass portion
of the whole rocket body normally is lower than for core
stage. Looking at the Ariane 5, Soyuz 2.1 Fregat, and
Atlas V launch vehicles, we estimate 8 % mass (dry mass)
of upper stages taken up by the engine, while for core
stages it is 18 %.
Satellites differ largely in composition and size depend-
ing on their field of use, mission lifetime, etc. Finckenor
(2018) provides some insight into materials used in space-
craft. The structure mainly consists of Al-alloys, Ti-alloys,
or stainless steel. Even Ni-alloys are in use, if more dura-
bility is needed. Aluminum is a preferred material as it
is lightweight. Generally, most structures are made out of
metals. However, some parts of the spacecraft, e. g. the
outer hull exposed to the sun, are constructed out of ma-
terial capable of withstanding thermal expansion. These
materials are mostly non-metallic, for example polyimide,
graphite or fiberglass. Non-metallic materials can also be
found in orbital debris and meteoroid shielding, although
Al-sheets are common. As many spacecraft in LEO are
Earth observation satellites, optical materials like quartz
or mirror material also play a role in the overall mass com-
position. A very important mass contributor of satellites
are solar arrays. We estimate their mass contribution to be
16 %. We assume half of the solar arrays to be old-fashion
Si solar arrays, whereas the other half use newer technol-
ogy with multijunction solar cells made of layers of Ge,
Ga-As, and Ga-In-P. All in all, we assume the following
mass composition: 40 % Al, 5 % Ni, 5 % Ti, 10 % Fe, 8 %
Si, 4 % Ge, and Ga, In, P, and As each 1 %. The remaining
24 % are other materials. We expect the large constella-
tion satellites to also match this composition reasonably
well.
3.4. Atmospheric processing of the anthropogenic material
Spacecraft and upper stages entering the atmosphere
have a much longer interaction time than meteoroids due
to their shallower entry angle and small entry velocity of
about 7 km/s. This is due to these anthropogenic ob-
jects approaching almost circular orbits during the reentry
phase. Therefore, their ablation is largely different from
that one of meteoroids. Anthropogenic material reaches
temperatures of 850 to 1950 K (Rochelle et al.,1997;Ailor
et al.,2005;Lips et al.,2017) while meteoroids and fireballs
can reach temperatures above 3000 K (Boroviˇcka,1993;
Jenniskens,2004). This strong difference in temperature
and the different materials imply different ratios of the
atomic and aerosol ablation for meteoroids/fireballs and
anthropogenic material. In need of better data, we as-
sume a higher aerosol fraction (75 %) for the anthropogenic
material while that one of meteoroids is only about 30 %
(compare with Figure 4). This is due to higher tempera-
tures causing transition into the gas and/or plasma phase.
Due to the lower ablation temperature and the high
mass (in the order of tons) of the anthropogenic material,
a significant fraction of their mass reaches the ground. To-
day, survival rates of human-made objects are expected to
range between 5 to 40 % (Ailor et al.,2005;Anselmo and
Pardini,2005;Pardini and Anselmo,2019). Simulations
with reentry software indicate values in this range, too
(Anselmo and Pardini,2005;Klinkrad et al.,2006;Kelley
et al.,2010). The thermal ablation of spacecraft and up-
per stages differs as well due to the different structure. For
spacecraft, we assume an average survivability of 20%, for
upper stages 35 %, and for core stages 70%. Large constel-
lation satellites are estimated to burn up completely in the
atmosphere (Space Exploration Technologies Corporation,
2016,2018;WorldVu Satellites Limited,2016).
3.5. Overall anthropogenic injection
Combining all the information about the annual an-
thropogenic mass influx, composition, and atmospheric
processing provides the following estimates for today’s in-
jection and the two different future scenarios emerge.
3.5.1. Today’s influx
Currently, 892 t of anthropogenic material enters Earth’s
atmosphere every year of which 88 t are injected in atomic
form; aerosols make up 263 t. The remaining material
(541 t) reaches the ground. From the injected elements,
aluminum is most abundant with 211t, followed by iron
(36 t), nickel (23 t), and copper (15 t). Metals make up at
least 86 % of the injected material.
3.5.2. Future Scenario 1
For Scenario 1, the annual anthropogenic mass influx
increases drastically to 2,742 t. 1,573 t are injected into
the atmosphere, 1,180 t as aerosols, 393 t in atomic form.
Again, aluminum is the largest part of the injection with
807 t, followed by iron (159 t), nickel (89 t), and silicon
(76 t). Again, most of the injected material is metal (at
least 75 %).
3.5.3. Future Scenario 2
For Scenario 2, the annual anthropogenic mass influx
increases even more to 8,114 t from which 4,904t are in-
jected into the atmosphere. Aerosols contribute 3,678 t,
material of atomic form 1,226 t. The order of the most
11
Table 4: Anthropogenic and natural injection for the different ablation products. Masses are given in t/yr. Numbers in parenthesis are the
percentage compared to the value of the natural material in the respective column.
Atomic Aerosol Total injection Ground-reaching
Anthropogenic
Today 88 (1.0) 263 (6.7) 351 (2.8) 541 (1,138)
Scenario 1 393 (4.7) 1,180 (30.2) 1,573 (12.8) 1,168 (2,458)
Scenario 2 1,226 (14.6) 3,678 (94.2) 4,904 (39.8) 3,210 (6,753)
Natural 8,421 3,904 12,325 48
Table 5: Anthropogenic and natural injection per element group. Masses are given in t/yr. The numbers in parenthesis depict the percentage
compared to the natural injection value in the respective column.
Metals Metalloids Non-metals Not assignable Total injection
Anthropogenic
Today 305 (7.5) 12 (0.7) 1 (0.02) 33 351 (2.8)
Scenario 1 1,189 (29.4) 123 (7.4) 10 (0.2) 252 1,573 (12.8)
Scenario 2 3,643 (90.0) 406 (24.6) 32 (0.6) 822 4,904 (39.8)
Natural injection 4,047 1,655 5,674 949 12,325
injected elements is the same as in Scenario 1 with alu-
minum (2,467 t), iron (496 t), nickel (272 t), and silicon
(251 t). The metal portion is at least 74%.
4. Final results and comparison
With all the available information, the natural and
anthropogenic injection can be tabulated and compared.
Three aspects have been evaluated in this study: the in-
jection by ablation products (Table 4), by element group
(Table 5), and the injection of selected elements (Table 6).
Today, the injection into the atmosphere is dominated by
natural material. About 2.8% of the mass is of human
origin. Although metals are highly abundant in space-
craft and rocket bodies, the anthropogenic metal injection
is also well below the natural metal injection. However,
there are elements which are injected mainly by human-
made objects, for example aluminum or copper. The an-
thropogenic injection can also prevail the natural injection
for some specific elements that are not very abundant in
the solar system and therefore in meteoroids, e. g. germa-
nium.
With the incorporation of large satellite constellations,
the injection situation changes strongly. The near future
Scenario 1 predicts 1,573 t of anthropogenic material in-
jected into the atmosphere, which is already 12.8% of the
natural injection. For the extreme Scenario 2 we infer
an anthropogenic mass injection rate of about 39.8 % of
the natural rate. For metals, the injection is even higher
with 29.4 % (Scenario 1) and even 90.0 % (Scenario 2)
of the natural metal injection, respectively. Additionally,
there are more elements for which the anthropogenic in-
jection surpasses the natural injection: for example tita-
nium (2459 %), chromium (131 %), and nickel (304%) for
Scenario 2. Satellite constellations also lead to a massive
enhancement of the injection of aluminum and copper.
The anthropogenic injection also increases the injection
of aerosols disproportionally as we estimate the entry of
human-made objects to produce more aerosols than atoms.
Today, human-made bodies make up 6.7 % compared to
the natural injection, while atomic material is only at
1.0 %. For future satellite constellations, the aerosol frac-
tion increases to 30.2 % and 94.2 % for the two scenarios.
Table 6: Anthropogenic and natural injection of some selected ele-
ments. Masses are given in t/yr. The numbers in parenthesis depict
the percentage compared to the natural injection value in the respec-
tive row. Note that percentages larger than 100 % indicate that these
elements are mainly of anthropogenic origin. For some elements, no
anthropogenic abundances were calculated.
El. Anthropogenic Natural
injectionToday Scenario 1 Scenario 2
H 220
C 0.1 (0) 0.2 (0) 0.5 (0) 1,054
O 3,851
Mg 0.04 (0) 0.1 (0) 0.3 (0) 1,300
Al 211 (161) 807 (614) 2,467 (1,877) 131
Si 8 (0) 76 (5) 251 (15) 1,654
S 513
Ti 7 (100) 52 (754) 171 (2,459) 7
Cr 7 (20) 17 (47) 48 (131) 37
Fe 36 (2) 160 (7) 496 (22) 2,295
Ni 23 (25) 89 (99) 272 (304) 90
Cu 15 (720) 38 (1,747) 106 (4,923) 2
Ge 4 (776) 37 (7,973) 124 (26,435) 0.5
5. Conclusion
The extensive review, analysis, and estimates presented
in this study provide an overview on the natural and an-
thropogenic injection of matter into Earth’s atmosphere.
At the present time, the anthropogenic injection already
12
contributes a non-negligible amount of mass to the injec-
tion. With large satellite constellations, proposed and
started from companies all over the world, the anthro-
pogenic injection will become significant compared to the
natural injection. Although many of the values used to es-
timate the injection inhibit uncertainties due to different
scientific results on many topics or insufficient data, the
results of this study should raise attention and also con-
cern towards the alteration of Earth’s atmosphere due to
the reentry of human-made spacecraft and rocket bodies.
Especially looking at metals, the anthropogenic injection
may well exceed 30% of the whole material deposited in
the upper atmosphere every year. Overall, in the near
future we need to be prepared that the injection of an-
thropogenic material will increase to 12.8 % 39.8 % of
the natural injection. Those values clearly show that the
anthropogenic injection is not negligible in the near future
and requires further consideration with respect to their
impact on Earth’s atmosphere.
The uncertainties involved demonstrate that more re-
search needs to be done to clarify the significance of the
effects of the human use of space on Earth’s habitat. There
are many different possible effects on the atmosphere that
may be caused by an increased injection. For example,
the large amount of aerosols injected by the ablation of
anthropogenic material may have an effect on Earth’s cli-
mate as aerosols in the high-altitude atmosphere have a
negative radiative forcing effect Lawrence et al. (2018).
Beside the intensively discussed problem of space de-
bris (e. g. Klinkrad,2006) we conclude that the re-entry of
human-made objects into the upper atmosphere may have
a significant effect on our habitat and needs more attention
in future studies. Advances in technology and a stronger
and stronger use of Earth’s environment always have side
effects that are most often not perceived at the beginning
of innovation and progress.
Acknowledgements
The authors thank Carsten Wiedemann, Martin Sippel,
Sven Stappert, Gerhard Drolshagen, and urgen Blum for
helpful discussions.
Appendix A.
The annual mass influx per mass decade shown in Fig-
ure 1is derived separately for the three different mass
ranges. For masses between 1021 to 102kg, the inter-
planetary flux model from Gr¨un et al. (1985) with the flux
at 1 AU given as
F(m) = (2.2·103m0.306 + 15)4.38
+ 1.3·109(m+ 1011 m2+ 1027 m4)0.36
+ 1.3·1016 (m+ 106m2)0.85 (A.1)
is used. The mass influx in a mass range between the
masses m1and m2can be calculated by integrating over
the flux F(m) and multiplying with Earth’s surface SE=
4π·(6.471 ·103m)2and the gravity enhancement factor
G= 1.445 (see Drolshagen et al.,2017). Here, an incident
atmospheric altitude of 100 km is used. Additionally, the
number of seconds in a year T= 3.1536 ·107has to be
multiplied to yield the annual mass influx, then given as
MGr¨un =SE·G·T·Zm2
m1
F(m) dm. (A.2)
In the mass range from 102to 106kg and 106to
1011 kg, the power laws from Brown et al. (2002) and
Stokes et al. (2003)
NBrown(E)=3.7E0.9(A.3)
NStokes(E)=2.4E0.79 (A.4)
are used, respectively. The latter one is obtained from
Drolshagen et al. (2017). N(E) represents the cumulative
number of meteoroids with a kinetic energy greater than
Eimpacting Earth every year, where Eis in units of kt
TNT equivalent. With an average meteoroid velocity of
20 km/s and 1 kt TNT equivalent = 4.184 ·1012 J, the en-
ergy dependence can be transformed to a dependence of
mass:
NBrown(m) = 2.86 ·104m0.9(A.5)
NStokes(m) = 6.22 ·104m0.79 (A.6)
with mthe mass in kg. So N(m) represents the number of
meteoroids of a mass greater than mhitting Earth every
year.
To yield the annual mass influx from these power laws,
further calculations are necessary (e. g. compare with Bland
et al.,1996, Appendix A). The number of meteoroids im-
pacting per year in a mass range from m1to m2can be
expressed by
N(m1)N(m2) = Zm2
m1
dN(m)
dmdm. (A.7)
To yield the annual mass influx Min the respective mass
range, the mass has to be incorporated in the integral by
multiplication:
M=Zm2
m1
mdN(m)
dmdm. (A.8)
This way, we yield the annual mass influx for both models
in the mass range from m1to m2
MBrown =Zm2
m1
2.57 ·104m0.9dm(A.9)
MStokes =Zm2
m1
4.9·103m0.79 dm. (A.10)
Taking the respective valid mass ranges of each model
(given above and in Figure 1), the integration yields MGr¨un =
13
Table A.7: Annual mass influx of meteoroids into Earth’s atmosphere
for each mass decade (values of Figure 1).
log m(kg) M(t/yr) log m(kg) M(t/yr)
-21 0.02 -5 384
-20 0.03 -4 184
-19 0.04 -3 86
-18 0.09 -2 42
-17 0.26 -1 53
-16 1.0 0 67
-15 4.3 1 84
-14 18 2 106
-13 85 3 133
-12 367 4 168
-11 1,125 5 211
-10 2,193 6 (264)
-9 2,671 7 (429)
-8 2,213 8 (696)
-7 1,410 9 (1,129)
-6 768 10 (1,831)
The mass influx Mis given for the interval of the object mass log m
to log (m) + 1. For example, 384 t of meteoroids in the mass range
from 105to 104kg impact Earth every year. The values derived
from Stokes et al. (2003) are in parenthesis as they are not included
in the annual mass influx in this study.
11,509 t/yr, MBrown = 863 t/yr, and MStokes = 4,349 t/yr.
The values for each mass decade, depicted in Figure 1, are
given in Table A.7. Remember that only the mass values
from MGr¨un and MBrown (so masses in a range from 1021
to 106kg) are counted to the annual mass influx used in
this paper.
Appendix B.
The elemental mass abundances of meteorites are given
in Table B.8. Compositions are derived from the follow-
ing sources: Chondrites, namely ordinary chondrites (H,
L, and LL), carbonaceous chondrites (CH, CI, CK, CO,
CR), Kakangari and Rumurutiites chondrites (K and R),
and enstatite chondrites (EH and EL) from Lodders and
Fegley (1998, Tables 16.10 and 16.11); most achondrites,
namely Acapulcoites (Acap), Angrites (Angr), Aubrites
(Aubr), Brachinites (Brac), Diogenites (Diog), Eucrites
(Eucr), Howardites (How), Lodranites (Lodr), Shergot-
tites (Sher), Nakhlatites (Nak), Chassignites (Chas), Ure-
ilites (Ur), and Winonaites (Wino) from Lodders and Fe-
gley (1998, Tables 16.11, 16.17, and 16.18) and Mittle-
fehldt et al. (1998, Tables 6, 8(4), 19, 21, 22, 26, 34, 35,
and 40); lunar achondrites, so called Lunaites (Luna) from
Demidova et al. (2007); stony irons, namely Mesosiderites
(Meso) and Pallasites (Pal) from Mittlefehldt et al. (1998,
Tables 13 (main group), 14 (main group), 15 (main group),
16 (all except Eagle Station), 17 (all except Eagle Station),
45(1), 46 (all except the last three)) and Wasson (1974, Ta-
bles II-5 and II-7(Ni)); and for Irons (IAB to IVB) from
Mittlefehldt et al. (1998, Tables 3 and 8 (1 & 9)) and Was-
son (1974, Table II-5). In a few cases, some abundances
were estimated (mainly oxygen) considering similar mete-
orite subgroups to complement the data. The total mass
abundance of all meteorites is shown in the last column.
It is the product of the meteorite abundance with the re-
spective elemental composition normalized to 100 % of the
mass. Thus, it represents the overall elemental composi-
tion of meteorites found on Earth. This is used as the
average elemental composition of meteorites given in the
last column of Table 1.
14
Table B.8: Elemental compositions of meteorite groups.
Z El. Unit H L LL CH CI CK CM CO CR CV EH EL K R Acap
Fraction wt% 42.27 37.72 6.36 0.08 0.04 0.54 1.18 0.62 0.57 0.36 0.93 0.28 0.13 0.85 0.07
1 H µg/g 20200 14000 700 2800
3 Li µg/g 1.7 1.85 1.8 1.5 1.4 1.5 1.8 1.7 1.9 0.7
4 Be ng/g 30 40 45 25 40 50 21
5 B ng/g 400 400 700 870 480 300 1000
6 C µg/g 2100 2500 3100 7800 34500 2200 22000 4400 20000 5300 3900 4300 580
7 N µg/g 48 43 70 190 3180 1520 90 620 80 420 240
8 O wt% 35.7 37.7 40 30 46.4 38 43.2 37 38 37 28 31 30 34 35.6
9 F µg/g 125 100 70 60 20 38 30 24 155 140
11 Na µg/g 6110 6900 6840 1800 5000 3100 3900 4200 3300 3400 6880 5770 6800 6630 6619
12 Mg wt% 14.1 14.9 15.3 11.3 9.7 14.7 11.5 14.5 13.7 14.3 10.73 13.75 15.4 12.9 15.8
13 Al wt% 1.06 1.16 1.18 1.05 0.865 1.47 1.13 1.4 1.15 1.68 0.82 1 1.3 1.06 1.24
14 Si wt% 17.1 18.6 18.9 13.5 10.64 15.8 12.7 15.8 15 15.7 16.6 18.8 16.9 18 18.0
15 P µg/g 1200 1030 910 950 1100 1030 1210 1030 1120 2130 1250 1400 1706
16 S wt% 2 2.2 2.1 0.35 5.41 1.7 2.7 2.2 1.9 2.2 5.6 3.1 5.5 4.07 2.76
17 Cl µg/g 140 270 200 700 260 430 280 250 570 230 100
19 K µg/g 780 920 880 200 550 290 370 360 315 360 840 700 710 780 516
20 Ca wt% 1.22 1.33 1.32 1.3 0.926 1.7 1.29 1.58 1.29 1.84 0.85 1.02 1.22 0.914 1.20
21 Sc µg/g 7.8 8.1 8 7.5 5.9 11 8.2 9.5 7.8 10.2 6.1 7.7 7.9 7.75 8.63
22 Ti µg/g 630 670 680 650 440 940 550 730 540 870 460 550 700 900 505
23 V µg/g 73 75 76 63 55 96 75 95 74 97 56 64 73 70 87
24 Cr µg/g 3500 3690 3680 3100 2650 3530 3050 3520 3415 3480 3300 3030 3600 3640 4229
25 Mn µg/g 2340 2590 2600 1020 1940 1440 1650 1620 1660 1520 2120 1580 2400 2960 2852
26 Fe wt% 27.2 21.75 19.8 38 18.2 23 21.3 25 23.8 23.5 30.5 24.8 24.7 24.4 22.8
27 Co µg/g 830 580 480 1100 505 620 560 680 640 640 870 720 750 610 789
28 Ni wt% 1.71 1.24 1.06 2.57 1.1 1.31 1.23 1.42 1.31 1.32 1.84 1.47 1.46 1.44 1.44
29 Cu µg/g 94 90 85 120 125 90 130 130 100 104 215 120 110
30 Zn µg/g 47 57 56 40 315 80 180 110 100 110 290 18 145 150 205
31 Ga µg/g 6 5.4 5.3 4.8 9.8 5.2 7.6 7.1 6 6.1 16.7 11 8.2 8.1 8.99
32 Ge µg/g 10 10 10 33 14 26 20 18 16 38 30 16
33 As µg/g 2.2 1.36 1.3 2.3 1.85 1.4 1.8 2 1.5 1.5 3.5 2.2 2.4 1.9 2.19
34 Se µg/g 8 8.5 9 3.9 21 8 12 8 8.2 8.7 25 15 20 14.1 9.75
35 Br µg/g 0.5 0.5 1 1.4 3.5 0.6 3 1.4 1 1.6 2.7 0.8 0.9 0.55 0.2
37 Rb µg/g 2.3 2.8 2.2 2.3 1.6 1.3 1.1 1.2 3.1 2.3 1.7 0.2
38 Sr µg/g 8.8 11 13 7.3 15 10 13 10 14.8 7 9.4
39 Y µg/g 2 1.8 2 1.56 2.7 2 2.4 2.6 1.2
40 Zr µg/g 7.3 6.4 7.4 3.9 8 7 9 5.4 8.9 6.6 7.2
41 Nb ng/g 400 400 250 400 400 500 500
42 Mo µg/g 1.4 1.2 1.1 2 0.92 0.38 1.4 1.7 1.4 1.8 0.9
44 Ru ng/g 1100 750 1600 710 1100 870 1080 970 1200 930 770 850 960 670
45 Rh ng/g 210 155 140 180 160 170
46 Pd ng/g 845 620 560 560 580 630 710 690 710 820 730
47 Ag ng/g 45 50 75 200 160 100 95 100 280 85 50
48 Cd ng/g 5.5 30 40 690 420 8 300 350 705 35 30 20
49 In ng/g 0.8 10 10.5 80 50 25 30 32 85 4 3 4
50 Sn ng/g 350 540 1700 490 790 890 730 680 1360
51 Sb ng/g 66 78 75 90 135 60 130 110 80 85 190 90 150 72 83
52 Te ng/g 520 460 380 2300 800 1300 950 1000 1000 2400 930 2000 1100
53 I ng/g 60 70 430 200 270 200 160 210 80
55 Cs ng/g 98 240 150 190 110 80 84 90 210 125 50
56 Ba µg/g 4.4 4.1 4 3 2.35 4.7 3.1 4.3 3.4 4.55 2.4 2.8
57 La ng/g 301 318 330 290 235 460 320 380 310 469 240 196 320 310 468
59 Ce ng/g 763 970 880 870 620 1270 940 1140 750 1190 650 580 830
59 Pr ng/g 120 140 130 94 137 140 174 100 70
60 Nd ng/g 581 700 650 460 990 626 850 790 919 440 370
62 Sm ng/g 194 203 205 185 150 290 204 250 230 294 140 149 200 180 223
63 Eu ng/g 74 80 78 76 57 110 78 96 80 105 52 54 80 72 96
64 Gd ng/g 275 317 290 290 200 440 290 390 320 405 210 196
65 Tb ng/g 49 59 54 50 37 51 60 50 71 34 32 58
66 Dy ng/g 305 372 360 310 250 490 332 420 280 454 230 245 29 468
67 Ho ng/g 74 89 82 70 56 100 77 96 100 97 50 51 59
68 Er ng/g 213 252 240 160 350 221 305 277 160 160
69 Tm ng/g 33 38 35 40 25 35 40 48 24 23
70 Yb ng/g 203 226 230 210 160 320 215 270 220 312 154 157 215 216 241
71 Lu ng/g 33 34 34 30 25 46 33 39 32 46 25 25 33 32 36
72 Hf ng/g 150 170 170 140 105 250 180 220 150 230 140 210 150 161
73 Ta ng/g 21 21 14 19
74 W ng/g 164 138 115 150 93 180 160 150 110 160 140 140 180
75 Re ng/g 78 47 32 73 38 60 50 58 50 57 55 57 43 60
76 Os ng/g 835 530 410 1150 490 815 670 805 710 800 660 670 550 690 693
77 Ir ng/g 770 490 380 1070 465 760 580 740 670 730 570 560 550 610 798
78 Pt µg/g 1.58 1.09 0.88 1.7 1 1.3 1.1 1.24 0.98 1.25 1.29 1.25 1 1.3
79 Au ng/g 220 156 146 250 145 120 150 190 160 153 330 240 220 183 217
80 Hg ng/g 30 22 310 60
81 Tl ng/g 0.5 2.4 15.5 142 92 40 60 58 100 7 3 20
82 Pb ng/g 240 40 2500 800 1600 2150 1100 1500 240
83 Bi ng/g 5 14 12.5 110 20 71 35 40 54 90 13 25 27
90 Th ng/g 38 42 47 29 58 41 80 42 58 30 38 50
92 U ng/g 13 15 15 8 15 12 18 13 17 9.2 7 25
Total % 101.9 100.8 101.6 99.7 100.4 99.1 100.0 100.7 99.3 99.6 97.2 96.8 98.1 98.4 100.6
15
Table B.8 (continued)
Z El. Unit Angr Aubr Brac Diog Eucr How Lodr Luna Sher Nak Chas Ur Wino Meso Pal
Fraction wt% 0.02 0.28 0.04 0.57 1.22 0.57 0.09 0.11 0.03 0.03 0.03 0.56 0.07 0.29 0.22
1 H µg/g
3 Li µg/g 5.4 3.5 3.7 3.9 1.4
4 Be ng/g
5 B ng/g
6 C µg/g 3000 500 265 200 850
7 N µg/g
8 O wt% 40.9 46.4 36 43.3 42.5 43.0 32.0 43.8 40 39 37 39.4 35 23.5 20.9
9 F µg/g 43.5 57 15
11 Na µg/g 179.2 3692 2805 896 3095 1568 760 2595 9700 3200 920 743 3720 1327 113
12 Mg wt% 7.84 22.7 17.3 14.6 4.8 9.2 16.8 3.8 6.19 7.55 19.2 20.6 14.6 5.2 12.6
13 Al wt% 5.66 0.58 0.60 0.97 6.84 4.50 0.23 11.60 3.42 1.1 0.42 0.26 0.7 2.43 0.024
14 Si wt% 18.8 26.8 17.4 24.4 22.9 23.3 15.1 21.0 23.8 22.3 17.5 18.2 15 12.4 8.3
15 P µg/g 556 1200 724 720 1787 1484 370 2715 1230 275 1463 10355
16 S wt% 0.59 0.61 1.52 0.12 0.138 0.077 0.62 0.16 0.0335 0.026 2 1.78 1.31
17 Cl µg/g 23 19 122.5 72.5 34
19 K µg/g 179 370 29.5 331 138 64 536 1305 985 300 66 337 166 5.4
20 Ca wt% 11.5 0.62 1.68 1.43 7.27 4.72 1.34 10.46 7.18 10.05 0.47 0.93 0.00082 2.22 0.26
21 Sc µg/g 43.8 10.9 17.5 51.8 20.7 7.2 24.5 53.5 54.5 5.3 8.1 8.1 16.5 0.64
22 Ti µg/g 7209 644 1160 862 3432 2775 320 5687 4810 2280 480 600 859 15
23 V µg/g 93 115 75 117 78 23 300 180 40 118 55 40
24 Cr µg/g 1078 182 4045 8889 2657 5490 4536 1221 1805 1525 5240 8391 1950 5992 5622
25 Mn µg/g 1433 480 2565 4294 4079 4002 2739 1042 3945 3850 4120 3024 2070 2478 1887
26 Fe wt% 13.9 1.0 22.8 13.3 12.6 13.4 31.1 7.7 14.6 16.4 21.2 13.8 19.8 46.3 46.2
27 Co µg/g 23.5 299 20.4 7.0 23.4 783 25.6 38 45.5 123 102 761 1199 3.5
28 Ni wt% 0.0047 0.41 0.0054 0.0012 0.0244 1.3 0.0119 0.00635 0.0093 0.05 1.02 1.21 4.52 10.00
29 Cu µg/g 2.5 8 16 12 2.6
30 Zn µg/g 1.47 239 0.71 1.24 26 120 64.5 66 72 230 131 3.0
31 Ga µg/g 4.9 0.18 1.7 0.75 6.0 3.6 15 3 0.7 4.9 10.9
32 Ge µg/g 0.020 0.14 0.765 2.75 0.01 37.5
33 As µg/g 0.37 0.197 1.5 0.0375 0.0825 0.008 269 2.57
34 Se µg/g 8.0 0.4 0.23 0.25 6.3 0.35 0.075 0.04 1.6 11.2 9.6
35 Br µg/g 0.41 0.10 0.21 0.855 2.435 0.088 0.27
37 Rb µg/g 2 0.1 0.13 0.32 3.1 6 3.3 0.73
38 Sr µg/g 15 1.7 65.9 31 143 46.5 67 7.2
39 Y µg/g 1.2 17.8 19 3.85 0.6
40 Zr µg/g 2.7 3 30 17 80 65 9.1 2.1
41 Nb ng/g 2700 5050 1530 340
42 Mo µg/g 0.015 0.37 0.086
44 Ru ng/g
45 Rh ng/g
46 Pd ng/g 2 0.4 10 1.75 15.85 0.15
47 Ag ng/g 11 30.0 19 49 2.6
48 Cd ng/g 21 13 47 94 14 19
49 In ng/g 3.2 0.92 25 20 4
50 Sn ng/g 10 600
51 Sb ng/g 56 11 7.2 62 45 7.1 40 0.9 121
52 Te ng/g 5 5.3 2.45 4.75 50
53 I ng/g 25 40 97 24 140 10
55 Cs ng/g 200 1.1 3.85 20 400 355 37
56 Ba µg/g 12 34.1 14 65 30 28 7.6
57 La ng/g 3545 365 154 2373 1214 80 5373 1835 1960 530 69 190 1366
59 Ce ng/g 10350 1600 315 7185 2668 14065 4600 5345 1120 4698
59 Pr ng/g 970 280 810 735 130
60 Nd ng/g 860 110 4960 1400 9228 3475 3160 620
62 Sm ng/g 2828 135 147 1440 680 58 2785 1290 805 140 26 90 308
63 Eu ng/g 956 57 41 560 276 27 925 537.5 230 45 10 48 173
64 Gd ng/g 240 2348 905 744 2540 890 110
65 Tb ng/g 803 65.5 409 200 23 653 380 120 30 82
66 Dy ng/g 175 2990 893 313 930 2850 860 200
67 Ho ng/g 48 758.75 230 211 710 162.5 44
68 Er ng/g 140 1740 590 1740 385 90
69 Tm ng/g 50 20 280 1000 300 52
70 Yb ng/g 2590 154 281 1526 790 173 2343 1450 360 110 73 157 408
71 Lu ng/g 380 43 23 228 124 30 336 227.5 53.5 15 12 24 61
72 Hf ng/g 1800 205 1317 1760 80 2044 1850 275 100 255
73 Ta ng/g 240 29 193 84 148 225 40045 20
74 W ng/g 6.5 30 558 535 260 46
75 Re ng/g 0.06 0.0053 0.0375 0.033 0.06 99
76 Os ng/g 0.7