Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.

Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 470

CHAPTER 28
DECONTAMINATION STANDARDS FOR MARTIAN
EXPLORATION PROGRAMS
CARL SAGAN and SIDNEY COLE MAN
It has recently been established that typical samples of terrestrial soil
contain populations of microorganisms capable of surviving the rigors of
simulated average Martian environments [Scher, Packer, and Sagan, 1964;
Packer, Scher, and Sagan, 1963]. Growth of microorganisms in simulated
microenvironments having moisture and organic-matter contents higher
than the anticipated mean values of Mars has also been demonstrated
[Young, Deal, Bell, and Allen, 1964]. These results give experimental
credibility to the concern, repeatedly expressed, that the landing of unsteri-
lized space vehicles on Mars may obscure subsequent attempts to detect, in
a pristine state, indigenous life on that planet. To avoid possible biological
contamination of Mars, it is clear that entry vehicles should be carefully and
conscientiously sterilized. This recommendation is in accord with others
previously made by the COSPAR Committee on Contamination by Extra-
terrestrial Exploration [CETEX, 1958; 1959], Lederberg [I960], the
Space Science Board {Brown et al., 1962], and Imshenetsky [1963].
A policy governing the decontamination of spacecraft landing vehicles
has been adopted by the National Aeronautics and Space Administration
(see Brown, et al. [1962] ); this policy is being carried out. If contamina-
tion is to be avoided, the Soviet Union must also implement a sterilization
policy for Mars entry capsules. Methods for space vehicle sterilization are
being actively developed in the United States, and there is a good indication
470

OCR for page 470

Decontamination Standards for Martian Exploration Programs 471
that there are adequate sterilization procedures which do not compromise
the engineering and scientific requirements of the mission [e.g., Phillips and
Hoffman, 1960; Posner, 1961; Swift, 1961, 1963; Brown et al., 1962; Hall
and Bruch, 1964].
At present, the most reliable procedure for insuring a high degree of
sterility is to heat the entire landing capsule. In one scheme, the capsule
is heated in three spaced cycles to temperatures above 135°C for 36 hr.
Longer exposures to lower temperatures may possibly be equally effective.
If heat sterilization is to be employed, the initial design of capsule com-
ponents must, of course, be made with the necessity of subsequent heat
sterilization in view. This seems to be a feasible objective; indeed, the
reliability of some components—for example, certain transistors—can be
enhanced by the heat-sterilization procedure. Nevertheless, sterilization
introduces serious engineering problems, especially by decreasing compo-
nent reliability. The type and duration of sterilization procedure must
depend on some estimate of what constitutes an acceptable risk of planetary
contamination. Our intention here is to provide a means of computing the
level of spacecraft sterility as a function of this acceptable risk.
One early study [Davies and Comuntzis, 1960] arbitrarily concluded
that the probability of landing one viable microorganism on Mars should
be kept below a = 10-° per mission. A recent reestimate by Jaffe [1963]
places

472 AVOIDING THE CONTAMINATION OF MARS
v is the mean number of viable microorganisms within the landing cap-
sule at the time of landing;
a is the mean number of viable microorganisms per capsule which are
distributed outside the capsule, on the Martian surface;
N is the total number of experiments that must be performed success-
fully before the experimental program can be considered completed;
X is the mean number of experiments per capsule;
/. is the probability that a capsule lands successfully, if it lands at all;
E is the probability that an individual experiment will succeed, assuming
that the capsule containing it has landed safely;
M is the probability that a viable microorganism landed on Mars will
subsequently multiply, and contaminate a significant area of the planet
within the duration of the total experimental program and, finally,
p is the probability that the program of N experiments will be completed
without biological contamination of Mars. We will require p to be very
close to unity.
We will assume throughout that all parameters are constant during the
experimental program. If our mean values of the above parameters are
chosen properly, this assumption will introduce no significant errors into
the final results. For all cases of interest, the product of a and M is much
less than unity, and thus, the probability that a given capsule contaminates
the planet is approximately aM. On the average, N/E experiments must be
attempted for N to succeed, and m = N/^EL missions are required.
To obtain an initial orientation, we will first assume that x = 1- The
calculation of p may then be performed exactly by elementary methods.
It is possible to write down the result immediately, using standard tech-
niques (e.g., see Feller [1957]). However, for the benefit of the reader
unfamiliar with probability theory, the result will be derived from first
principles. We will then show that an excellent approximation to the exact
expression may be obtained through an approximate method. We will then
use this approximate method to derive an expression for the case x > 1.
Since it is unlikely that every mission will be successful in its search for
and characterization of life on Mars, there will be a number, Q > N,
missions to Mars before the desired information is obtained. Consider the
(N -f- /)th mission, and assume that it is the final in the series of N experi-
ments. Thus, the (N + ;')th experiment must be successful, because, by
hypothesis, we are content to risk a probability p that biological contami-
nation of Mars occurs after N + / missions. The remaining missions may
be arranged in any order whatever. Thus there are (N -f- / — 1)! permu-
tations of the remaining missions. Of these, N — 1 will be experimentally
successful, where again, the order, for which there are (N — 1)! possi-

OCR for page 470

Decontamination Standards for Martian Exploration Programs 473
bilities, is immaterial. At the same time, / missions will be failures, and
once again the ordering of the failures is unimportant. Further, there must
be no contamination events in N -\- j missions. The total number of ways
of obtaining such scenarios of successful experimentation before contami-
nation will then be
p-
(N — I)!/! (1)
j = 0
Thus, p is the probability that biological contamination does not occur
until N experiments are performed. From the power series expansion for
(1 — >>)-.*, we find
T LE(\—aM)
p = i
Taking the Nth root of each side, we find
LE(l — aM) i/y~i i n ww (3)
since (In p)/N is a small (and negative) number. Since aM and aM/LE
are also expected to be « 1, we may expand the lefthand side of (3),
finding it to be approximately 1 — (aM/LE). Equation (3) then becomes
LE Inp-. (4)
We will always wish to arrange the over-all experimental program so that
p is very close to unity. In this case,
7 p
We may obtain the approximate equation (5) in a less rigorous, though
more direct manner by making our approximations at the beginning of the
calculation rather than at the end. We argue as follows: We wish to obtain
the probability of scientific success in N experiments before biological
contamination occurs. Because aM is very small, the probability of non-
contamination is a very slowly varying function of the number of missions
landed. Also, because N is fairly large, the probability distribution for the
total number of experiments landed — successfully or otherwise — will clus-

OCR for page 470

474 AVOIDING THE CONTAMINATION OF MARS
ter sharply about its mean value, N/LE. Therefore, to a good approxima-
tion, we may replace the average probability of noncontamination by its
value at the mean—that is to say,
pc^(\— oM)*'L». (6)
Taking the logarithms of both sides, we obtain an expression which is
equivalent to Equation (4).
Our argument leading to Equation (6) can now be easily generalized
to the case that x > 1- N is then replaced by N/x and the remainder of
the argument is unchanged. Thus, we obtain
'":- (1— P). (7)
MN
Up to this point, we have discussed only vehicles intended for planetary
landings. There is a category of spacecraft, however, intended for planetary
fly-by and orbit that may, nevertheless, accidentally impact the planet. If
landers are sterilized but fly-bys and orbiters are not, the probability of
planetary contamination from the latter spacecraft may be higher than
from the former.
The number of microorganisms on a typical unsterilized spacecraft may
be v '—- 10-10- A significant fraction F of these escape from the vehicle
upon impact. Therefore, the impact of one such spacecraft has a good
chance of contaminating the planet, provided, as seems to be the case,
that FM » 1010. The heat and deceleration of impact will not kill
all the contained microorganisms. Thus, the probability of contamination
from an intended fly-by or orbiter will be I, the probability of accidental
planetary impact, a parameter ordinarily computed in the planning of
interplanetary trajectories.
We assume that, in the long run, the contribution to our knowledge of
Martian biology from fly-bys and orbiters will be negligible, compared with
the contribution from landers. Then, if n is the number of fly-bys and
orbiters launched towards Mars in the same period as the landers, our
previous arguments can be extended to give
(8>
The division of effort between sterilizing landers and controlling trajec-
tories of unsterilized fly-bys and orbiters is a problem in operations re-
search appropriate to national space agencies. However, to visualize the

OCR for page 470

Decontamination Standards for Martian Exploration Programs 475
numerical constraints on a and /, let us arbitrarily require that the con-
tamination risk from fly-bys and orbiters equal that from landers.
We then have, finally, a requirement for landers:
and for unsterilized fly-bys and orbiters:
Some representative numerical values for a as a function of N and p
appear in the accompanying graph. We have assumed that the mean num-
ber of experiments per mission over the entire exploration program is
X — 20, and that the mean probability, over the entire program, of spacecraft
failure resulting in no biologically significant information from an entire
mission is 1 — L ^ 0.1. We have also taken the mean probability of
scientific success of a given experiment to be £ — 10-1. Estimating the
value of E involves trajectory information, since presumably not all locales
on Mars are equally likely to have detectable life-forms. It also involves an
evaluation of the ability of biologists to design experimental packages to
detect Martian life-forms in the face of almost complete ignorance of the
biochemistry, morphology, or physiology of Martian organisms.
While every tested terrestrial soil sample has microorganisms capable of
surviving simulated mean Martian conditions \Scher, Packer, and Sagan,
1964], the exact fraction that survives is very difficult to determine. The
problem is complicated by the wide variety of microbial survival conditions
and the fact that no culture medium is optimal for all microorganisms.
There are many microbial types which do survive freezing and dessication
[Packer, Scher, and Sagan, 1963] . We adopt a preliminary value of the
fraction of survivors ^ 10-1, a value which, after the initial freeze-thaw
cycle, is essentially independent of time. But we emphasize that the uncer-
tainty is at least one order of magnitude; intercomparison of soil samples
shows a fractional survival range of at least this order. Survival and growth
are by no means equally easily achieved, especially under average Martian
environmental conditions.
In the following discussion, we adopt M ^ 10-2. The graph shown in
Figure 1 gives, for the choice of parameters just cited, minimum values of a.
From Equation (9) and the graph, the reader can easily make his own
choices of these parameters and redetermine

476
AVOIDING THE CONTAMINATION OF MARS
Figure 1.
Risk of planetary contamination. See
text.
0
-2
-4
--
-8
bS"°
* -,2
-14
-16
-18
-10
-I*
-24
Parameters assumed:
X- 20; L =0.9; E = '
M • ICT8! n « 30.
icriz icr10 icr"
icr6
I-P
10-* KT2 0.3
biology. If there are Martian life-forms, biologists will need adequate time
to investigate their anatomy, physiology, biochemistry, genetics, ecology,
systematics, and behavior. The wealth of desired information per variety
of Martian organism, coupled with the potential variety of Martian organ-
isms, indicates that the value of N must be large. However, it is relevant
that no existing plans for manned landings are responsive to the contamina-
tion problem. There seems to be no immediate prospect of spacesuits with
negligible contaminant leak. Indeed, the design difficulties involved may
prove insuperable. The specification of a very large N in an experiment-
centered strategy will be tantamount to requiring manned exploration of
Mars in self-contained mobile vehicles entering and leaving the spacecraft
through a sterilization lock.
An alternative and, from the present vantage point, much less desirable
procedure will be to accept a mission-centered strategy; to accept planetary
contamination after manned landings on Mars; and to maximize the effi-
ciency of the preceding unmanned exploration phase. In the spirit of this
alternative, let us adopt 1984 as the date of the first manned landing on
Mars.
Assuming an average of three launches of landing vehicles at each op-
portunity before 1984 by each of two space-faring nations, we have about
60 possible missions, of which 54 land successfully. With x — 20, this
leaves 1080 possible experiments. With E = 10-1, we have N = 108 suc-
cessfully completed experiments, a number small compared with the range
of biological prospects, but hardly insignificant. During the same period, a
number n — 30 launches of fly-bys and orbiters appears to be an upper
limit.
Suppose it were concluded that N = 108 is consistent with an experi-
ment-centered strategy. If we desire 99.9% probability that 1080 biological

OCR for page 470

Decontamination Standards for Martian Exploration Programs 477
experiments can be attempted before contamination, the mean number of
viable organisms deposited on the planet by each spacecraft must be less
than 9.3 X 10~4^ 10-3. With a mission-centered strategy the analysis fol-
lows that preceding Eq. (6), and Eq. (9) is replaced by a <(l-p)/2Mm,
where m is the total number of missions. But we have seen that m =
'N/x EL. The equation is therefore identical to Eq. (11), and whichever
strategy is used, the numerical results are the same. However, in the ex-
periment-centered strategy N will almost certainly be larger than 120, and

478 AVOIDING THE CONTAMINATION OF MARS
case, the parameters which were assumed for the computation of a and I
should be periodically reevaluated, especially in the light of new informa-
tion obtained about Mars from early fly-bys, orbiters, and landers.
There are, however, other lines of work which, if properly pursued,
might at some time in the future lead to an increase of the acceptable values
of a and /. We recall that a is the probability that a single viable micro-
organism be deposited on the Martian surface. In the case of an intact soft
landing, only those organisms near the exterior of the spacecraft have a
high probability of being deposited on the Martian surface, a point recently
emphasized by Horowitz [1965]. It would appear at first sight that organ-
isms contained within sealed components, such as batteries or transistors,
have no chance of escaping from the spacecraft. If this were rigorously
true, the entire sterilization requirement would be greatly ameliorated,
because the microorganisms likely to contaminate the planet would then
be precisely those which can be killed by chemical surface sterilants. The
necessity for heat sterilization would then be diminished.
However, microorganisms may diffuse through solid materials in time
scales of the order of a decade [Sneath, 1964]. Also, there will always
be a non-zero probability of accidental hard landing on Mars, in which
case the fraction of the total load of contaminations which are distributed
over the Martian surface will depend on the fragmentation size-distribution.
If an impacting spacecraft fragments into spherical particles of characteris-
tic radius r, the fraction of contained microorganisms of 5 /i dimensions
which are sufficiently close to a surface to escape onto the planet will then
be as follows:
P(r) > l_(l_5/r)», (11)
where r is in microns [Horowitz, 1964]; e.g., if r = 10 cm, P(r) > 1.5
X 10-4. This expression gives a lower limit to P(r), because it assumes a
uniform fragmentation size. Small fragments make the greatest contribution
to planetary contamination. In fact, P(r) should be multiplied by /(r),
the frequency distribution of fragments of characteristic size r. Then, if we
knew the probability that a given landing results in a particular fragmenta-
tion distribution /(r), we could compute the probability of escape of a
given organism due to an imperfect soft landing. If we also knew the
specific probability that an organism migrates—by solid-state diffusion or
through aeolian erosion—a distance r in time /, we could compute a/v,
where v is the total number of viable organisms aboard the spacecraft. In
general, a < v.
At the present time it is difficult to make even a fair numerical estimate
of the relevant parameters. It is clear that substantial work is required on

OCR for page 470

Decontamination Standards for Martian Exploration Programs 479
the fragmentation distribution of model spacecraft, on the diffusion of
microorganisms through solids, and on the rate of aeolian erosion on Mars.
Moreover, a good estimate of the probability distribution of accidental hard
landing at various impact velocities would be useful. Until such information
is available, it is premature to assume that a/v « 1. For the present we
must design space missions as if a/v — 1.
It should also be emphasized that any spacecraft designed to sample
periodically the Martian environment will also provide at the same time
some opportunity of egress for microorganisms contained in the spacecraft.
Sampling procedures imply a route for microorganisms from the inside to
the outside of the lander.
There are departures in spacecraft design which could greatly relax the
standards on / and v. Precautionary terminal systems can be developed
for assuming non-intercept trajectories in case of imminent accidental
impact—for example, by extending the existing capabilities of midcourse-
correction motors. Such systems must be failsafe. Possibly more prom-
ising, both for intended fly-bys and orbiters and for intended soft landers,
would be the development of impact-actuated terminal-destruct sterilization
procedures. These procedures can be low in weight and cost, compared
with adequate sterilization before launch. Prelaunch sterilization of fly-bys
and orbiters is also desirable. The problems attending microorganism solid
state diffusion and aeolian erosion of soft-landed vehicles can similarly be
circumvented—here, by development of destructive heat-sterilization tech-
niques to be activated after a lander completes its experimental program.
The reliability of the timer and sterilization would have to be demonstrated
before such techniques could supplant prelaunch sterilization.
Several other decontamination requirements have been pointed out by
Brown, et al. [1962]. These include provision for an independent author-
ity to certify that the required values of a are achieved; an emphasis on
biological and chemical experiments with the first Mars fly-by, orbiter, and
landing vehicles; the acquisition and retention of large numbers of sterile
samples by the earliest missions; and an active program to continue sterility
precautions after beginning manned exploration of Mars.
Future simulation studies can contribute to our evaluation of the proba-
bility of biological contamination of Mars, and therefore to a better speci-
fication of the vehicle sterilization parameter, a. It will be important to
determine the fraction of survivors of simulated Martian environments
which also survive spacecraft sterilization procedures. It is possible that
organisms which are fortuitously pa-adapted to survive heat sterilization
are extremely maladapted to Martian regimes of freeze-thaw cycling. Cer-
tainly, there seems to be little selection pressure on Earth for survival of
the same organisms both at 135°C and —80°C. Further characterization

OCR for page 470

480 AVOIDING THE CONTAMINATION OF MARS
of the Martian environment and of the nature and modes of surviving
microorganisms should permit a better understanding of this problem.
Summary. A significant fraction of terrestrial microorganisms survive
the inclemency of simulated mean Martian conditions; a smaller fraction
may be able to grow at favored times and places on Mars. To avoid bio-
logical contamination of Mars by terrestrial microorganisms on unsterilized
landing vehicles, the number of viable microorganisms,

Decontamination Standards for Martian Exploration Programs 481
Feller, W. (1957), An Introduction to Probability Theory and Its Applications,
John Wiley & Sons, New York, pp. 155-156.
Hall, L. B., and C. Bruch (1965), Procedures Necessary for the Prevention of
Planetary Contamination. In: Life Sciences and Space Research III, M.
Florkin (Ed.), North-Holland Pub. Co., Amsterdam.
Horowitz, N. H. (1965), "Spacecraft Sterilization." Chapter 27, this volume.
Imshenetsky, A. A. (1963), Perspectives for the Development of Exobiology.
In: Life Sciences and Space Research I, R. B. Livingston et al, (Eds.),
North-Holland Pub. Co., Amsterdam, pp. 3-15.
Jaffe, L. D. (1963), Sterilization of Unmanned Planetary and Lunar Space
Vehicles—An Engineering Examination. Jet Propulsion Lab. Tech. Rept.
32-325 (Rev.); also published in Astronautics and Aerospace Engineering
1(1), 22.
Lederberg, J. (1960), Exobiology: Approaches to Life Beyond the Earth.
Science 132, 393-400.
Packer, E., S. Scher, and C. Sagan (1963), Biological Contamination of Mars,
2: Cold and Aridity as Constraints on the Survival of Terrestrial Micro-
organisms in Simulated Martian Environments. Icarus 2, 293-316.
Phillips, C. R., and R. K. Hoffman (1960), Sterilization of Interplanetary
Vehicles. Science 132, 991-995.
Posner, J. (Ed.) (1961), Proceedings of Meeting on Problems and Techniques
Associated with the Decontamination and Sterilization of Spacecraft, June
29, 1960, Washington, D. C. NASA Tech. Note D-771, Washington, D. C.
Scher, S., E. Packer, and C. Sagan (1964), Biological Contamination of Mars, 1.
Survival of Terrestrial Microorganisms in Simulated Martian Environ-
ments. Life Sciences and Space Research II, M. Florkin and A. Dollfus
(Eds.), North-Holland Pub. Co., Amsterdam, 352-356.
Sneath, P. H. A. (1964), Private communication.
Swift, J. (1961), Effects of Sterilizing Agents on Microorganisms. Jet Pro-
pulsion Lab. Astronautics Information Literature Search No. 260; and
Supplement (1963).
Young, R. S., P. H. Deal, J. Bell, and J. L. Allen (1964), Bacteria Under
Simulated Martian Conditions. Life Sciences and Space Research II, M.
Florkin and A. Dollfus (Eds.), North-Holland Pub. Co., Amsterdam,
105-111.