Mon. Not. R. Astron. Soc. 000, 1–?? (2014) Printed 13 June 2014 (MN LATEX style file v2.2)

Detecting industrial pollution in the atmospheres ofearth-like exoplanets

Henry W. Lin1?, Gonzalo Gonzalez Abad2? and Abraham Loeb2?1Harvard College, Cambridge, MA 02138, USA2Harvard-Smithsonian Center for Astrophysics, 60 Garden St., Cambridge, MA 02138, USA

Accepted MMMM. Received NNNN; in original form 2014 June 11

ABSTRACTDetecting biomarkers, such as molecular oxygen, in the atmospheres of transitingexoplanets has been a major focus in the search for alien life. We point out thatin addition to these generic indicators, anthropogenic pollution could be used as anovel biomarker for intelligent life. To this end, we identify pollutants in the Earth’satmosphere that have significant absorption features in the spectral range coveredby the James Webb Space Telescope (JWST). We estimate that for an Earth-massplanet in the habitable zone of a white dwarf, methane (CH4) and nitrous oxide (N2O)can be detected at earth-like concentrations with an integration time of ∼ 1.5 hrsand 12 hrs respectively. Detecting pollutants that are produced nearly exclusively byanthropogenic activities will be significantly more challenging. Of these pollutants, wefocus on tetrafluoromethane (CF4) and trichlorofluoromethane (CCl3F), which will bethe easiest to detect. We estimate that ∼ 1.5 days (∼ 3 days) of total integration timewill be sufficient to detect or constrain the concentration of CCl3F (CF4) to ∼ 100times current terrestrial level.

Key words: planets: atmospheres — planets: extrasolar — white dwarfs:

1 INTRODUCTION

The search for extraterrestrial intelligence (SETI) has so farbeen mostly relegated to the detection of electromagneticradiation emitted by alien civilizations (e.g., Wilson (2001),Tarter (2001), Shostak et al. (2011)). These signals could bethe byproduct of internal communication (Loeb & Zaldar-riaga 2007) or perhaps simply the result of the alien civiliza-tion’s need for artificial lighting (Loeb & Turner 2012).

On the other hand, the search for biomarkers in the at-mospheres of transiting Earths has thus far been limited tothe search for unintelligent life. Detecting biomarkers suchas molecular oxygen at terrestrial concentration in the atmo-spheres of transiting Earths around white dwarfs will be fea-sible with next-generation technology like the James WebbSpace Telescope (JWST) (Loeb & Maoz 2013). Thoughmolecular oxygen and other signals like the red edge ofphotosynthesis are strong indicators of biological processes(Scalo et al. 2007; Seager et al. 2006; Kaltenegger et al. 2002)one might ask whether there are specific atmospheric indica-tors of intelligent life that could be simultaneously searchedfor.

In this Letter, we explore industrial pollution as a po-

? E-mail: henrylin@college.harvard.edu; ggonzalez-abad@cfa.harvard.edu; aloeb@cfa.harvard.edu

tential biomarker for intelligent life. This would provide analternative method for SETI, distinct from the direct detec-tion of electromagnetic emission by alien civilizations. Fordefiniteness, we focus on Earth-size planets orbiting a whitedwarf, as Agol (2011) have argued that white dwarfs havelong-lived habitable zones and the similarity in size of thewhite dwarf and the planet will give the best contrast be-tween the planet’s atmospheric transmission spectrum andthe stellar background (Loeb & Maoz 2013).

On Earth, atmospheric pollution has been carefullystudied in the context of global climate change and airquality concerns. It is ironic that high concentrations ofmolecules with high global warming potential (GWP), theworst-case scenario for Earth’s climate, is the optimal sce-nario for detecting an alien civilization, as GWP increaseswith stronger infrared absorption and longer atmosphericlifetime. On the other hand, the detection of high concen-trations of molecules such as methane (CH4) and nitrousoxide (N2O) though suggestive, will not be conclusive evi-dence for industrial pollution, as the contribution from nat-ural sources is comparable to the contribution from anthro-pogenic sources for the present-day Earth (Anderson et al.2010). Instead, targeting pollutants like CFCs is ideal, asthey are only produced in significant quantities by anthro-pogenic activites. In addition, the lifetimes of CFCs varyfrom ∼ 101 yrs to ∼ 105 yrs. Thus, detection of a short-life

c© 2014 RAS

arX

iv:1

406.

3025

v1 [

astr

o-ph

.EP]

11

Jun

2014

2

CFC will signal an active civilization, whereas the detectionof a long-lived CFC will be more probable, as the pollutingcivilization only needs to exist at any time in the past ∼ 105

yrs.Our discussion is organized as follows. In §2 we calculate

the necessary JWST exposure time for detecting differentmolecules. In §3,, we focus on CFC-11 and CFC-14. Finally,we summarize our main conclusions in §4.

2 JWST EXPOSURE TIME ESTIMATE

First, we compute the exposure time needed to detect agiven molecule in the atmosphere of a planet, given the pres-ence of other molecules and the changing concentration ofthe molecules with altitude. Although we focus on planetsaround white dwarfs and JWST, our results are generaliz-able to other telescopes and planetary systems.

If we are interested in molecule j, then the number ofsignal photons in a given wavelength bin from λ to λ+ ∆λis given by

Sj,λ = ∆λ

(dNwddλ

)ε(λ)

2π

A⊕

∫r>r⊕

rdr Tj (1− Tj) , (1)

where 0 6 Ti(r, λ) 6 1 is the intensity transmission coef-ficient I/I0 of the i-th molecule at an altitude of r − r⊕,Tj =

∏i 6=j Ti(r, λ), and ε(λ) is the quantum efficiency of the

instrument, which we take to be 0.6 for the Mid InfraredInstrument (MIRI) on JWST (Ressler et al. 2008). Further,we assume a blackbody spectrum for a white dwarf withsurface temperature of 6000 K:

dNwddλ

∝ λ−4

exp(2400 A/λ)− 1(2)

whereNwd is the number of photons detected from the whitedwarf, with the proportionality constant chosen such that atλ = 7000 A, we have dNwd/dλ = 8 × 10−5s−1cm−1A

−1 ×tint×AJWST, where AJWST is the effective collecting area ofJWST and tint is the total integration time. Here we haveassumed that a typical half-occulted white dwarf to be about18 mag (Loeb & Maoz 2013).

To estimate the exposure time needed to detect a givenmolecule, we use a Monte Carlo approach. We generate asignal Sj,λ, add Poisson noise Nj,λ =

√∆λ dNwd/dλ, and

then fit a functional form∑ajSj(λ) to the simulated pho-

ton deficit. We take ∆λ = λ/3000 to simulate medium-resolution spectroscopy that MIRI on JWST will provideand only consider the signal in certain spectral windows tobe determined.

We repeat the process, each time finding best fit pa-rameters aj and their associated uncertainties. We averagethese parameters aj over the repeated trials and generatenew parameters (averaging again) until all the aj are within∼ 1% of unity and then take the average uncertainty sj inaj . Since aj = 0 would indicate no signal, aj = 1± 0.2 cor-responds to a 5σ detection , and in general aj = aj ± sjcorresponds to (aj/sj)σ detection.

Note that the linear fit is approximate; in practice in-creasing the concentration for Si6=j will diminish Sj signalvia the Tj factor. However the effects of the nonlinearity issmall compared to the uncertainty in the molecular concen-trations, so we ignore this issue by suppressing the factor Tj

in our estimate. For real surveys where accurate constraintson molecular concentration are desired (and not just an es-timate of the necessary integration time), a non-linear fit inconcentration should be performed and uncertainties shouldbe estimated by, say, a Markov Chain Monte Carlo method.

As a check on our calculations, we find that the neces-sary exposure time for detecting molecular oxygen (O2) isjust ∼ 3 hrs for an efficiency ε = 0.15 and a spectral resolu-tion R ≡ λ/∆λ = 700. This is roughly in agreement with the∼ 5 hr estimate given by Loeb & Maoz (2013), though ourmethod provides a more optimistic integration time as thefitting routine takes into consideration shape information inaddition to photon count statistics.

3 TARGETING CFC-11 AND CFC-14

We compile a list of industrial pollutants with small or negli-gible production from natural sources. Of this list, we iden-tify CFC-14 (CF4 ) and CFC-11 (CCl3F) as the strongestcandidates for detection. The overlap of the 2ν4 and ν3 bandsaround 7.8µm is the strongest absorption feature of CF4 inthe middle infrared region (McDaniel et al. 1991). For CCl3Fthe ν4 band around 11.8µm is the strongest absorption fea-ture (McDaniel et al. 1991). These features are fairly broad,∼ 0.1µm and ∼ 0.4µm for CF4 and CCl3F respectively.The advantage of CF4 is that it absorbs at 7.8µm, whereasCCl3F absorbs at ∼ 11.7µm. Since these wavelengths arein the Rayleigh-Jeans regime of a 6000 K white dwarf, thephoton flux will fall off rapidly with wavelength ∝ 1/λ3. Onthe other hand, CH4 and N2O are strong absorbers in theregions of interest for CF4, whereas in the 11 – 12µm regionwhere CCl3F is strongly absorbing, interference (predomi-nantly from O3 and H2O) is less relevant.

We have calculated the synthetic spectra using theReference Forward Model1. The RFM is a GENLN2 (Ed-wards et al. 1992) based line-by-line radiative transfer modeloriginally developed at Oxford University, under an ESAcontract to provide reference spectral calculations for theMichelson Interferometer for Passive Atmospheric Sounding(MIPAS) instrument (Fischer et al. 2008). In our simula-tions the model is driven by the HITRAN 2012 spectraldatabase (Rothman et al. 2013). We use the US standardatmosphere2 to simulate the temperature, pressure and gasprofiles of the major absorbers and the IG2 climatology ofCFC-11 and CFC-14 (Remedios et al. 2007) originally de-veloped for MIPAS retrievals.

Since we are interested in CF4 and CCl3F, we considerthe signal in the following spectral windows:

• 7760 nm < λ < 7840 nm for CH4, N2O, CF4

• 11600 nm < λ < 12000 nm for CCl3F

Tighter constraints on CH4 and N2O in turn improve theconstraints on CF4 due to CH4 and N2O signatures in thewavelengths where CF4 features are present. We thereforeconsider four additional windows:

• 2190 nm < λ < 2400 nm for CH4 and N2O• 3350 nm < λ < 3550 nm for CH4

1 http://www.atm.ox.ac.uk/RFM/2 http://www.pdas.com/atmos.html

c© 2014 RAS, MNRAS 000, 1–??

Detecting industrial pollution in the atmospheres of earth-like exoplanets 3

Figure 1. Windows used for detecting CF4 and CCl3F . The top figure shows the combined transmission T = I/I0 from the most

relevant molecules in Earth’s atmosphere over the JWST wavelength range. To constrain CF4 concentration, we use five intervals in

wavelength space where CF4, N2O, and CH4 features dominate. We show in shaded orange three of the five most important regions onthis plot (1-3). The fourth region is used to constrain CCl3F. The zoomed in versions of these plots are displayed at the bottom.

• 3840 nm < λ < 4130 nm for CH4 and N2O

• 4500 nm < λ < 4600 nm for N2O

Although CH4 and N2O absorb strongly at other wave-lengths, these windows minimize the interference from othermolecules.

By demanding a given signal-to-noise ratio (say, S/N =5) we can solve for the total integration time tint necessaryfor a 5σ detection of the given molecule. Figure 2 showsthese exposure times as a function of the exoplanet’s nor-malized concentration. Note that lower levels of CH4 andN2O concentrations on the exoplanet will lower the neces-

c© 2014 RAS, MNRAS 000, 1–??

4tint

@days

D

-1.0 -0.5 0.0 0.5 1.0

0.05

0.10

0.50

1.00

5.00

Log CêC≈

CH₄

N₂O

tint

@days

D

1.0 1.5 2.0 2.5 3.0

1.0

10.0

5.0

2.0

20.0

3.0

1.5

15.0

7.0

Log CêC≈

CFC-14

CFC-11

Figure 2. Exposure time needed to detect various molecules as a function of their normalized concentration C/C⊕ where C⊕ is the

current terrestrial concentration. On the left panel, N2O and CH4 are displayed. On the right, we display CCl3F and CF4. The molecules

in the left plot will be considerably easier to detect, requiring . 1 day of observing time for a 5σ detection, whereas detection of moleculeson the right will be significantly more challenging unless present at concentrations ∼ 102 times terrestrial levels.

sary exposure times for CF4 detection, since these moleculesprovide most of the interference in the ∼ 7.8µm region. Fora CF4 concentration of ∼ 100 times that of the Earth, oneneeds ∼ 3 days of exposure time, whereas ∼ 1.5 days willbe sufficient to detect CCl3F at 100 times terrestrial levels.The long exposure time needed to detect CF4 or CCl3F incomparison to the time needed to detect molecular oxygenis mostly a consequence of the factor of ∼ 1000 fewer pho-tons per unit wavelength emitted at ∼ 7.8µm compared to∼ 0.7µm, the location of the molecular oxygen signature.The situation is even worse for CCl3F at ∼ 11µm.

Given the cost of long exposure times, we suggest thefollowing observing strategy:

• Identify Earth-size exoplanets transiting white dwarfs.• After ∼ 5 hours of exposure time with JWST, water va-

por, molecular oxygen, carbon dioxide and other biomarkersof unintelligent life should be detectable if present at earth-like levels (Loeb & Maoz 2013).• While observing for these conventional biomarkers,

methane and nitrous oxide should be simultaneously de-tected, if they exist at terrestrial levels. Extreme levels ofmethane and nitrous oxide could be preliminary evidence ofrunaway industrial pollution.• If biomarkers of unintelligent life are found within the

first few hours of exposure time, additional observing timecould be used to reduce the uncertainties on the concentra-tion of the discovered biomarkers and search for additional,rarer biomarkers. Methane and N2O can be “subtractedout”, and constraints on CF4 can be obtained. Constraintson CCl3F can be simultaneously obtained.• Direct exoplanet detection experiments could then be

used to push constraints on industrial pollutants to terres-trial levels. At the same time, these experiments could searchfor less exotic biomarkers like the “red edge” of chlorophyll(Seager et al. 2005).

It is worth noting that a recent study by Worton et al.(2007) estimate the atmospheric concentration of CF4 at∼ 75 parts per trillion (ppt), whereas CF4 levels were at∼ 40 ppt around ∼ 1950. Assuming a constant production

rate C = C0 +γt, we expect as a very crude estimate that inroughly ∼ 1000 years, the concentration of CF4 will reach10 times its present levels. Coupled with the fact that thehalf-life of CF4 in the atmosphere is ∼ 50, 000 years, it is notinconceivable that an alien civilization which industrializedmany millennia ago might have detectable levels of CF4.

4 CONCLUSIONS

We have shown that JWST will be able to detect CF4 andCCl3F signatures in the atmospheres of transiting earthsaround white dwarfs if these pollutants are found at con-centrations at ∼ 100 that of terrestrial levels with ∼ 1.5 and∼ 3 days of integrated exposure time respectively, thoughN2O and CH4 can be detected at terrestrial concentrationswithin 12 hrs and 1.5 hrs respectively.

Given that conventional rare biomarkers will alreadytake on the order of ∼ 1 day to detect, constraints on CF4

and CCl3F, at ∼ 102 times terrestrial levels, could be ob-tained at virtually no additional observing costs. Detectionof high levels of pollutants like CF4 with very long lifetimeswithout the detection of any shorter-life biomarkers mightserve as an additional warning to the “intelligent” life hereon Earth about the risks of industrial pollution.

ACKNOWLEDGMENTS

This work was supported in part by NSF grant AST-1312034, and the Harvard Origins of Life Initiative.

REFERENCES

Agol E., 2011, ApJl, 731, L31Anderson B., Bartlett K., Frolking S., Hayhoe K., JenkinsJ., Salas W., 2010, Technical report, Methane and NitrousOxide Emissions from Natural Sources. EPA

c© 2014 RAS, MNRAS 000, 1–??

Detecting industrial pollution in the atmospheres of earth-like exoplanets 5

Edwards D., et al., 1992, Technical report, GENLN2: Ageneral line-by-line atmospheric transmittance and radi-ance model.. NCAR

Fischer H., Birk M., Blom C., Carli B., Carlotti M., vonClarmann T., Delbouille L., Dudhia A., Ehhalt D., En-demann M., Flaud J. M., Gessner ., 2008, AtmosphericChemistry and Physics, 8, 2151

Kaltenegger L., Fridlund M., Kasting J., 2002, in FoingB. H., Battrick B., eds, Earth-like Planets and MoonsVol. 514 of ESA Special Publication, Review on habit-ability and biomarkers. pp 277–282

Loeb A., Maoz D., 2013, MNRAS, 432, L11Loeb A., Turner E. L., 2012, Astrobiology, 12, 290Loeb A., Zaldarriaga M., 2007, JCAP, 1, 20McDaniel A. H., et al., 1991, J. Atmos. Chem., 12, 211Remedios J. J., Leigh R. J., Waterfall A. M., MooreD. P., Sembhi H., Parkes I., Greenhough J., ChipperfieldM., Hauglustaine D., 2007, Atmospheric Chemistry andPhysics Discussions, 7, 9973

Ressler M. E., Cho H., Lee R. A. M., Sukhatme K. G., DrabJ. J., Domingo G., McKelvey M. E., McMurray Jr. R. E.,Dotson J. L., , 2008, Performance of the JWST/MIRISi:As detectors

Rothman L., et al., 2013, Journal of Quantitative Spec-troscopy and Radiative Transfer, 130, 4

Scalo J., Kaltenegger L., Segura A. G., Fridlund M., RibasI., Kulikov Y. N., Grenfell J. L., Rauer H., Odert P.,Leitzinger M., Selsis F., Khodachenko M. L., Eiroa C.,Kasting J., Lammer H., 2007, Astrobiology, 7, 85

Seager S., Ford E. B., Turner E. L., 2006, AGU SpringMeeting Abstracts, p. A5

Seager S., Turner E. L., Schafer J., Ford E. B., 2005, As-trobiology, 5, 372

Shostak M. S., Zuckerman P. B., Fujishita M., Horowitz P.,Werthimer D., 2011, Astrobiology, 11, 487

Tarter J., 2001, ARAA, 39, 511Wilson T. L., 2001, Nature, 409, 1110Worton D., Sturges T., Gohar K. ., 2007, Environmentalscience & technology, 41, 2184

c© 2014 RAS, MNRAS 000, 1–??