- Email: [email protected]
The ionization balance of the atmosphere
The ionization VICTOR
balance of
the atmosphere
F. HESS and ROWER P. YANCOUR Fordham tiniversity, (Rec&ed
Xew York, N.T.
30 Janua7y 1950)
ABSTRACT Two methods were dereloped for determining the residual ionization in brass chambers and the absolute ionization of the hard and the soft component of cosmic radiation in the atmosphere at sea level, without secondaries from the walls of the chambers. The portable GISH-SHERMAX ionization meter in connection with four geometrically similar c$indrical vessels of widely different size was used. One method is based on the use of all four chambers, the other requires only one lerge chamber in which the pressure is varied between about O-1 and 1O atm. These measurements were carried out on the campus of Fordham University
in a garden.
It was found that, on bhe average, the hard component [filtered through 10 cm of iron) produces an ionization of 1.51 I while the total intensity (hardfsoft component) amounts t.o 1.88 1. The effect, of the gamma rags from the radioac~ire substances floating in the atmosphere was determined b,y a new method, described in this paper. This “ air radiation *’ amounts to about 0.13 I. varying ronslderably from day tG day. Another set of experiments was carried Gut to find t,he ionization produced by the alpha rays from radon. thoron, and their subsequent products. Samples of air from outdoors were filled into a large ionization eliamber and the varia.tion of ionization u-as observed for several days. In this way it was possible to determine the ionizing effect of the radon and thuron product.s separately. It was found that radon and its subsequent products give O-93I, while thoron and its products give O-83 1. The latter. t,herefore conrribut,e 47 “; of the t.Gt,alionization bj- alpha rays. The radon content itsetf was 46 x 1Vz8 curie per’ cm3 (average of November and December). The ionization by gamma rays from the ground, at Fordham. amounted to 3.0 1. The contribution by beta rays from radium, thorium and their decay products in the ground and in the atmosphere was estimated as 04 I. 1 m above gmund, tonsist,ed therefore of the
The total ionization, on the point of obserration, following four components : B&a rays Oamma rays Cosmic ra+s
(I.40 I 3.161 l.YOI
!
, ’
Total
: i.21 I
Using tile linear law of recombination of small ions (E. SCHKEIDLER, J. J. SOL~K and others) which holds for air polluted with a great number of nuclei. one ca.n compute the mean number of small ions per cm3 if one knows the average lifetime of such ions. Taking this lifetime as lO*(i set (from a recent determination at the same location), one would get less than 100 small ions per cm3 present in the stationary state. This is to be expected in tile vicinity of a big city, where the number Gf nuclei of condensation ordinarily exceeds 4OOOOper cnP,
Our knowledge of the ionization balance of the atmosphere is still rather incomplete. We know that over the continents the radioactive substances in the ground and in the air contribute most of the ionization observed while cosmic radiation is practically Here the variations of ionization are the only ionizing agency over the oceans. They are mainly due to variations of the barometric pressure relatively small. There are also other small and of the temperature in the lower atmosphere. fluctuations of cosmic ray intensity with time and with geomagnetic latitude which were studied so extensively within the last two decades. h’evertheless? data on the absolute intensity of cosmic radiation. even at sea level and at medium latitudes,
13
VICTOR F. HESS and ROGER P. Vmcow
are rather conflicting: for unshielded ionization chambers at geomagnetic latitudes beyond the “ knee “, the reported values range from 1.8 to 2.8 I (“1 ” denotes the number of pairs of ions produced in air per cm3, per set). It is to be noted that in all these measurements various authors did not take into account the increase of ionization by secondaries from the walls of the vessel. However, it is now generally believed that the additional ion production by secondaries in a thinwalled brass chamber is quite small, perhaps only a few per cent of the total effect. Quite recently the senior author in collaboration with R. P. VAXCOTJRdevised two methods [lJ of determining the absolute intensity of cosmic rays in the atmosphere without secondaries and also the residual io~zation in ionization chambers_ These methods, originally tried out in the laboratory, were used last summer in outdoor experiments which will be described in this paper (Sections I, II and IV). In addition, we determined the ionization produced by the gamma rays from the radioactive products in the atmosphere (Section III), the radon content of air and the relative contribution of thorium products in the atmosphere to the total ionizatioxrnear the ground (Section V). In Section VI, the ionization balance is discussed. All experiments were performed on the campus of Fordham University which ia situated in the Bronx, near the Botanical Gardens, New York City. They should give a good average picture of the io~zation balance in the lower atmosphere. I.
INTENSITY OF COSMX RADIATION AND THE RESIDUAL IONIZATION IX BUSS CHAMBERS
The ionization of the atmosphere at sea level, over land, is produced by the alpha, beta and gamma rays from the radioactive products in the soil and in the atmosphere, and, to a lesser estent, by cosmic radiation. If we measure this ionization in a hermetically sealed vessel (wall 2.5 mm thick) filled with a gas free of radon, the total ionization is due to: 1) occasional alpha rays from radioactive impurities in the metal of the ohamber (residual ionization, Q*); 2) gamma rays from the radioactive substances in the soil (a~) ; 3) gamma rays from the radioactive decay products of radon, thoron and actinon in the atmosphere &) ; and 4) cosmic rays (PC). q=qO+qE+qA+qC
If the ionization chamber is placed in an iron house with walls thick enough to absorb practically all the gamma rays (10 cm of iron), the total ionization is due to (lo and qC. In this case, we observe the effect of cosmic rays filtered through 10 cm of iron. If the apparatus is set up outdoors and the top of the iron house is lefi open, the ionization is due to unfiltered cosmic radiation and to the small gamma ray component from the radioactive products in the atmosphere qA which in most cases will not amount to much more than 0.1 I. The residual ionization (cfo) due to alpha rays from the inner surface of the wall of the chamber can be kept at much less than 1 I if the chamber walls are very carefully cleaned and the volume of the chamber is not too small. Since this residual ionization is proportional to the surface, while the other ionizing agencies are volume effects, it is advantageous to use chambers of fairly large size in which the ratio of surface area (A) to the volume (W) is rather small. The total ionization, as observed in the iron house (closed on top), is q=qo +
qc=Akna +qc 14
(1)
The ionization balance of the atmosphere
where n, denotes the number of alpha particles emitted from the wall per cm2 per set, A the surface area of the chamber wall, and k the average number of ion pairs produced by each alpha particle. The ionization produced by cosmic rays could also be considered partly as a surface effect since the secondaries produced in the walls of the chamber will increase (at constant thickness of the wall) with the area of the wall exposed to cosmic rays. Experiments to be discussed later will show that in thin-walled ionization chambers (as used by us) this effect of secondaries is very small. Therefore, we can treat the cosmic ray ionization as a volume effect, especially since the ranges of these secondaries are considerably longer than the diameter and height of our chambers. We have used ionization chambers of cylindrical shape, made of yellow brass (wall thickness 2.5 mm) in conjunction with the ionization meter, devised by 0. GISH, K. L. SHERMAN and one of the authors, which is described elsewhere [Z]. The chambers were geometricaliy similar (height equal to twice t’he diameter of the chamber) with volumes 4368, 13.17, 4.89 All four cylinders were and 1,645 1, and radii 15, 10.16, 7.30 and 5.08 cm respectively. airtight, and could be evacuated or Bled with any gas through a Hoke needle valve. We used air from steel cylinders (stored for at least one month). The amber-insulated inner electrode of each chamber is directly connected to the fibre system of a Lindemann electrometer which is mounted in a housing fitted exactly to the top of each of the chambers, which are used in a vertical position. The guard ring and the case of the electrometer are connected with a battery supplying 135 volts, and the centre point of the quadrant battery is kept at the same potential. A needle contactor is used for connecting or disconnecting the &bre system from the battery with which ‘the guard ring is permanently connected. The wall of each chamber is kept at ground potential. When the contactor is lifted the needle begins to float, and the rate of drift is a measure of the ionization in the vessel. We preferred to use the “ null-method “: a t-volt auxiliary battery with a potentiometer and a precision voltmeter (range 3 volts) allowed us to add any voltage from 0 to -3 volts to the 135.volt sweep voltage, thus inducing an opposite charge on the floating system just enough to keep the electrometer needle always at or close to the zero point. When using this nullmethod, all corrections for insulation leakage of the electrometer are avoided. The reading of the compensating voltage atethe end of each measurement (from 5 to 60 min in duration) multiplied by a reduction factor, which is proportional to the induction coefficient between the guard ring system and the inner electrode for each chamber, gives the ionization (q) expressed i ion pairs per cmJ per sec. The residual ionization of ihe three smaller chambers was determined in 1947 by using them within the iron house in a subway tunnel under 210 ft of solid rock where the cosmic-ray intensity is practically zero and local radiation was eliminated by the 10 cm iron wall around the apparatus [3]. These direct measurements served as a control for determinations of the resi ma1 ionization with the two simpler methods discussed below. We shall now describe one method used for the determination of the residual ionization and the intensit.y of cosmic rays in our brass chambers, and then proceed to the determination Jf the cosmic-ray intensity in t,he free at,mosphere. This method is based on the successive use of all four chambers in the iron house at the location where a determination of the cosmicray intensny is desired. After the tests in the laboratory which were described in our previous publication, all these experiments were performed outdoors : the iron house was set up in the garden of the Seismic Station of Fordham University on even ground. The numerical values for the dimensions of the four chambers are as follows: 15
VICTOR F. HESSant1 ROGERP. Va~ocrr
Utumber
II~~IL~I~( W) Inner (A= cm3
The total
Zl’urj’ace lot+)
-4/W
(r)
WI2
Cl)1
I II
43680 13173
7070 3243
15.0 10.16
O-1618 0.2462
III IV
4888 1645
16i4 Sll
7.30 5.08
O-3425 0.4930
ionization
current
in each chamber
I=eAkn Dividing we get
Radius
by eA and inserting
I
__
This is a linear relationship
between
to equation
(1) is (in e.a.u.)
(2)
.+eWqC
A = lCk*x and W=
1 Oer%
according
4r% (radius-2
=knc(+04
the total ionization
rqc
of the height
of each cylinder),
(3)
observed
and the radii of the chambers.
The experimental test of this relationship was carried out for two cases; with the iron house set up in the garden of the Seismic Station at Fordham, we put each of the four chambers successively in the iron house and observed the ionization in each chamber a) when t’he iron house was completely closed (10 cm iron on all sides); b) when the top of the iron house was removed. In this case, the soft component of the cosmic radiation comes into play and, in addition, a small part (qA) of the observed ionization is contributed by the gamma rays of the radioactive substances in the lower atmosphere (up to about 4 km from the ground). In this particular case the term qC in equation (3) should be replaced by (q, + qA), while kn, remains We, therefore, can expect to get two straight lines with different slopes; their unchanged. intercept should coincide at the axis of the ordinates.
As can be seen from Figure 1, our experimental with this expectation.
points are in perfect agreement
It may be stated that our value for the residual ionization, as found here, was even lower than the one reported in our previous publication, when the chambera were carefully cleaned, repeatedly evacuated. and refilled with air from our t,nnk. ThB extrapolated value of IllOde for the radius r = 0 represents kna. Taking for k (the average number of pairs of ions produced by one alpha p&icle) 105, we obtain for Finxti l-25 and the number of particles from the wall (per cm2) 1.25 x 10-j per set or 4.5 x 1O-2 per hr which is of t,he same order of magnitude as the lowest alpha-ray emission reported so far by J. A. BEARDEN, for steel (O-10 >er hr). We are now in a position to compute the residual ionization (q,,) simply hy multiplying knr= 1.26 with the surface area -1 and dividing by the volume VV for each chamber. Figure 2 gives the curve (arc of a parabola) obtained when these computed values of q. are plotted against the radii of the chambers. The values q. for the four chambers obtained were 0.20 I for No. I, O-31 I for No. II, 0.43 I for No. III. and 0.62 I for Ko. IV. Three of these were checked directly by observing the residual ionizat’ion of three chambers by an underground experiment when they were placed within the iron house in a subway tunnel under 16
The ionization balance of the atmosphere 70 m of
5 1oq;
The comparisons gave a satisfactory agreement within solid rock. between experimental and computed values. Subtracting the values q,,
1
8 72 761x1~ P-r Fig. l-Extrapolation of residual ionization from measurementsof four chambers. 8
Y
Fig. 2-Residual ionimtion vs. radius of the ionizationchambers.
given above from the actually observed values of total ionization house, the following cosmic ray intensities (qc) are obtained.
in the iron
Table 2 Cosmic Ray
Intensities
Chamber Xo.
I II III IV
(Sea Level) in Brass Chambers (Wall 2-5 mm thick)
(a) Iron House C1ose.d (IO cm Fe)
1.52 1.55 1.54 1.55
(b) Iron House. Top Removed
I I
1.94 I 1.96 I
I I
1.94 I 1.99 I
The values in column (a) correspond to the hard component, in column (b) to the hard plus the soft component (including a small effect qA of the gamma rays of the radioactive substances in the air). II.
DETERMWATIOK OF THE TREE VALUE OF THE IONIZATION BY COSMIC RAYS IN THE FREE BTMO~PHERE: WITHOUT SECONDARIES
In order to obtain the ionization without secondaries from the walls of the chamber, we plotted the observed values of q = Ah + qcf (total we proceeded as follows: The extrapolated value of q for ionization observed in the iron house) vs. A/W. A/W=0 (infinite volume) would indicate the ionization produced in a chamber of infinite volume filled with radon-free air at atmospheric pressure, at the point We obtained a linear relationship of q vs. A/W for all four chambers of observation. in both cases 1) iron house closed, 10 cm Fe around the chambers and 2) with the Case (1) gives an extrapolated value of 1.51 I; t,op of Dhe iron house removed. Case (2), l-931 (see Figure 3). These values were confirmed by direct computations from the equations of the two straight lines. 17
B
VICTOB F. HESS=d
ROGERP. VANCOUR
These figures actually represent the true ion production in air without any secondaries emitted from the walls, since in a chs.mber of infinite volume any contribution of residual ionization (alpha rays from the walls) as well as from electrons knocked out of the walls by cosmic rays would be negligible. The two figures are only a few per cent lower than the ones given in Table 2 for the brass chambers. The latter values obviously do not represent the true ionization in the atmosphere at the point of observation (50 degrees N geomagnetic latitude). However, the difference is very small, especially for the largest chamber No. I (43 1 volume) ; this indicates that the secondaries from the brass walls (2.5 mm thick) increase the ionization by only s, few per cent. The difference between the two values l-93 I and 151 I gives, approximately, the ionization produced by the soft component of cosmic radiation. This value (O-42 I) amounts to about 20% of the total effect of cosmic rays at sea level, and this agrees with previous estimates of various authors. However, it is importent to remember that the still somewhat uncertain and variable effect of the gamma rays from radioactive substances floating in the lower atmosphere is included in the figure l-93 I. This effect was determined (see the following section) as a7 017 ay 55 a6 about O-15 1 (average). Subtracting it from l-93 I A/w-. we get 1.78 I as the mean total ionization by cosmic Fig. 3-ExtmpoLtion of ionim- rays (soft and hard component) rtt sea, level, in the tion bY coamic m~s in the free free atmosphere. A small correction of this figure is hXl0fpheP3. necessary on account of the fact that the ionization chambers in the iron house are not recording the actual total intensity of the soft component when the top of the iron house is taken off. Rays coming in from the lower regions of the sky are partly absorbed when penetrating the 10 cm iron shield in an oblique direction. We computed from the geometry of the whole arrangement the absorbed fraction of the soft component, assuming the distribution of the unidirectional intensity of cosmic rays according to STEINKE’S metimements [4]. The results indicate that the absorbed fraction of the soft component ctmounts to about O-1 1. The total ionization of the soft and hard component, therefore, is l-78 + O-10= 1.88 I at sea level in the free atmosphere (50 degrees geomagnetic latitude N) while the hard component alone amounts to l-51 I. III.
APPROXIMATE DETERMINATION OF THE “Am
(Ionizcctionby Gamma
Rays from
the Radioactive
Substances
RADIATION” in the Atmmpphere)
The “ air radiation ” (qd ) was estimated by V. F. HESS [5] in 1934 as about O-1 I. A better estimate was derived directly from our experiments ; since we know the sverage absorption coefficients of the gamma rays from RaC and ThC we csn estimate the percentage loss in an absorbing p&e of known thickness. We took two iron plates (together l-2 cm thick) and observed repeatedly the ionization in the iron house when its top was off and when the two plates were placed on the top. Calling the two values q1 and qz we have the relation
(1 -eqpd”d)
q1 -h=Ph 18
The ionization
balance of the atmosphere
where p is the average absorption coefficient [6] of hard gamma rays in iron (0.356 cm-l). The absorption of the “ soft ” as well as of the hard component of The term (1 -e-Id) corresponds to the ftaction cosmic rays is hereby neglected. of gamma rays absorbed in 1.2 cm Fe. We performed seven long series of observations alternatively with “ top off” and “ 1.2 cm Fe on top ” with chambers I and II in the iron house between July 26 and September 7, 1949, and obtained the following results : Table 3 Difference of Tonitation Observed $?I;$
Average
From
these figures we determined 0*0691 =qA
0.050 I 0.223 I 0.044 I 0.035 I 0.012 I 0.048 I 0.0691 I _cO*O26I (probable error f
O-0178 I)
qA, using the equation
( 1-e-o*427) and get, qA =is
= 0.198 I
The large fluctuations of the single values given above (although partly statistical) This can be expected since we know that the indicate that the effect is variable. The value radon content of the atmosphere is varying greatly, from day to day. mentioned above gives an upper limit of the average value qA since it was tacitly assumed that the whole difference (ql - q2) is due to the absorption of gamma rays from the radioactive substances in the atmosphere in 1.2 cm iron. However, we can derive an estimate also of a lower limit of q4 by computing the possible effect of the absorption of the soft component ,of cosmic radiation in 1.2 cm Fe; taking the mass absorption coefficient of this soft component as 8.6 x 10m3 cm2/g (E. REGENER) and the water equivalent of 1.2 cm Fe as 9.4 cm the exponential e-o.oo*5 x g*4 gives O-923. Since the soft component could, according t’o our measurements, amount to not more than the total difference of 0.42 I observed when the top of the iron house (10 cm Fe) was off and on, we see that 0.42 x 0*923= O-388 I would be the amount getting through l-2 cm Fe end only If we assume that all of the difference of the ionization 0.03 I would be absorbed. tentatively observed with the “ top on ” (10 cm Fe) and “ top off “-arrangement was due to the soft component of cosmic radiation, we can now derive a lower limit for the amount of air radiation (qa): Since we observed a difference q1 - q2= 0.069 I a O-018 I when the iron house was completely open on the top and when it was covered with 1.2 cm Fe we can now say that at the most only 0.03 I of this difference can be ascribed to absorption Subtracting this from O-069 I we of the soft component of cosmic radiation. get O-039 I as the lower limit for (ql - q2) to be ascribed to absorption of gamma Inserting rays of the radioactive substances in the atmosphere in 1.2 cm iron. this value in our equation (4) we get O-039 = qA (1 - e-o*427) = 0.348 q1 and from this ql=0.112 I 19
VKCTOR F. HESSand ROC+ER P. VANCOUR
Thus we are able to bracket the true value of the air-radiation qA as lying somewhere between 0.198 and 0.112 I and taking as an average for qA= 0.161 we see that this is about six times the mean average error of our observed values for (ql - qe). Thus, it is evident that the contribution of gamma rays from the radioactive substances in the air is real and cannot be neglected in the ionization balance. Its variability is clearly reflected in our measurements (Table 3). These variations naturally will very often mask a&us1 fluctuations of cosmic radiation, if the ionization chamber is not shielded on the top against these “atmospheric ” gamma rays. XV.
ANOTHER METHOD FOR DETE~~~~
TEE IONIZATION BY COSMXCRAYS AND THE RESIDUAL IONIZATION OF IONIZATION CHAMBERS
This method was described in a previous publication [l] in detail. Since it was u&i a&n la& summer in our experiments outdoors, a short explanation of our procedure will be necessary, It is well known that the ionizationin a closed vessel incre~es almost linearly with pressure, but the curve flattens out st b&her Dmssures except when very mrre 8sgon is used F71. Since -we ~snted to wo& with sir in the chamber, we preferred to use low pressures up to atmospheric pressure ss C. T. R. W-ON did in his earlv ermeriments. We filled our lamest chamber (43.7 1) with “air &om & tsnk (stored for 2 least one month) at stmospheric pressure and observed the ion&&ion at that pressure and at various lower pressures when the chamber wss placed in the iron house. It wsa to be expected that the ion&&ion vs pressure curve would be linear &Blong se the pressure p (expressed in atrn~~e~) wss not lower then that at whiuh the range of alpha particles is equal to
the diameter of the chamber (30 cm). At still lower pressure, the curve was expected to deviate downward toward zero.
Since the current .I- Aet&a + Weqcp and the first term, is constant as long as the range does not exceed 30 cm we obtain
0.80
Pfe.ssufe
The slope of the linear part of the q vs. p curve, therefore, corresponds to the actual total ionization of the cosmic radiation in the largest brass chamber, while the extrapolation of the linear part of the curve for p= 0 would give the approximate value of the residual mmHg ionization (q,,) ; it can be shown that this would be
Fig. P-Ionization by cosmic rays vs. pressure.CurveI: hard component; Curve II: soft -Ihad component.
Qo-P-
dq dp’
P
where q denotes the total ionization observed at the pressure r). It is, of course, more accurate to obtain dq/dp from the graph directly. Figure 4 shows two sample curves obtained with air in the chamber. Curve I corresponds to the case when the iron house was closed (10 cm Fe surrounding the chamber on all sides). In this case, only the hard component of the cosmic rays was effective. Curve II (top of iron house removed) gives the values for soft and hard components together. The iron bars in this case reached up just to the upper rim of the ionization chamber. Therefore, the soft 20
The ionization
balance of the atmosphere
component (plus a small amount of gamma rays from the atmosphere) a solid angle which was less than 2~: sterradian.
came from
The value of dq/dp can be corrected for this case (half space) by graphical methods. From the slope of the two curves. we obtain for Case I, qc = l-45 I (Meson The extracomponent) and for Case II, qc = l-84 I (soft and Meson component). polated value (0.20 I) for the residual ionization (qo) agrees quite well with the The correction mentioned above and disvalue obtained with the other method. cussed at, the end of Section II amounts t,o 0.1 I and therefore the value for the ionizat,ion due to the soft and hard component is 1.84 + 0.10 = l-94 I which is in good agreement with the value derived in Sect,ion II. T’.
THE IOKIZATION PRODFCED BY RADON. THORON,AND THEIR DECAY PRODUCTS IN THE ATMOSPHERE A very simple method of det,ermining the ionization produced by the alpha rays of these
products was tried successfully during the fall of 1949 when the outdoor experiments were completed: we assembled the iron house in the basement laboratory of the Physics Building at Fordha,m and det.ermined first the ionization in the chamber II (13 1) when it was filled with aged, inactive air from a steel cylinder while standing in the iron house. Then the chamber was evacuated and refilled with atmospheric air in the garden of the seismic station where all our previous gamma ray and cosmic ray experiments were performed. The atmospheric air samples were taken about 3 ft above the ground. After filling, the chamber was put. back into the iron house and the total ionization measured immediately. In a number of cases, the readings were continued for several hours and further readings taken on the following day and even two or three days afterward, in order t,o get, a decay curve.
It is clear that t’he difference between t,he total ionization observed in the iron house with aged (inactive) air and the atmospheric air in the chamber corresponds to the ionization by t,he alpha rays of radon, thoron and their decay products (pa), Thoron itself will: of course, not be measured since all of it would decay within the ten minutes or more which ela’psed between the moment when the sample was taken and when t,he first measurement could be made. The effects observed. therefore: are mainly due to Radon and its decav products (Rn + RaA The half life of ThB (10.6 hrs)“is short, as compared +RaC) and (ThB f 5°K’). with the one of Rn (3-82 days) and so we can expect that within 2-3 days most of the ThB has decayed and only the Rn, in equilibrium with its decay products. survives. The very low background of our chambers plus. the cosmic ray ionization (q. + qc) in t,he basement laboratory amounting together to only about 1.6 I allowed us to detect quite easily and wit,h sufficient accuracy t,he effect (aa) of the alpha rays from the radon, ThB, etc. int’roduced into the chamber when the ,atmospheric air sample was admitted in the chamber. This effect amounted in most’ cases to between 1 and 2 I. By taking readings immediately after the sample was tilled into the chamber and one or two days later, it was even possible to determine the percentage of ionization due to thorium products. This was accomplished in the following way: When atmospheric air is introduced in the chamber it produces an ionization qa which, added to the ionization produced by cosmic rays and the residual (wall) radiation gives a total effect of Q1=Qo+9c+46!
(5)
Measuring (q. + qc) in the chamber before the experiment, we immediatelv obtain qkis comqa about ten minutes after the atmospheric air sample is introduced. 21
VICTOBF. Fhs and ROGERP. VANCOUR
posed of the ionizstion produced by (Rnf RaA + . . .) and (ThB+ThC+ . . .), both assumed to be in equilibrium. We denote the decay constant of radon with hr (2.097 x 10-S se+) and of thorium B 5ts i2 (l-82 x 10m5set-1) and assume that a fraction z of QO~ is due to (Rn + decay products) while the fraction (1 -x) of qa is due to the decay products of ThB. Then, we have, at a known time t later z$$ -*It + (1 - 2) q,e- Aat = q;I - A,t _ e -ii,t ) = q’,/q, - e- Aat
or
x(e
PorlPor-e-hd -__ x=&t_, -A,t
and
From this, we computed the relative contribution of the decay products of radium and thorium in the atmospheric air sample in each case, when pa and ql, (at least 10 hrs later) were determined. We carried out a number of experiments determining qa and the frrtction z of radium products in November and December 1949. The results are listed below in Tables 4 and 5. A slight correction has to be applied to the measured values of qa on account of the fact that alpha rays emitted from points near the chamber walls are not fully utilized. An empirical correction was worked out by W. DUANE and A. LABORDE [8] for radon and its decay products in a cylindrical chamber ; the corrected value is obtained by dividing the measured value by (l-o.572 4) where A is the surface and W the volume of the ionization chamber. Accepting this correction as approximately valid also for the .decay products of ThB.and taking the value of A/W as O-249 for our chamber (No. II), we have
l-0.572
;=
0.8575
The values qa listed in the table are already corrected in this way. The third column (equivalent in cu.ries/cm3) gives the number of curies of radon plus decay products which would produce the same ionization ss the one actually measured in each case ; this would include the thorium products, expressed in equivalent amounts of mdon. The computation of the radon equivalent in curies per cm3 from the observed ionization Qa was based on the values given by ST. MEYER and E. SCEWEIDLER [9] : 1 curie (radon) per cm3 gives a saturation current of 2.75 x lo6 e.s.u. and 1 curie of radon with its decay products 2.75 x 106x 1.55+1’70+2*20=-9.67x 106 e .s .u .). An amount of radon of lo-l6 curie/ 1.55
cm3, therefore, corresponds to 2.01 I. Date (1949) Nov. 3 Nov. 4 Nov. 7 Nov. 10 Nov.. 14 Nov. 15 Nov. 16 Nov. 21 Dec. 5
q,
(cow.)
Table 4 Radon equivalent in curie/cm3
1.61I 1.28I 2.21I 1*57I 1.02I 1.58I 1.43I 1.56I 2.67I
79.8x IO--‘@ 63.5x lo-l8 109.6x lo-l8 78.1x lo-*8 5o?Jx lo-18 78.5x lo-‘* 70.9 x 1O-‘8 77.3 x lo-‘8 132*6x IO-‘* 22
The ionization
balance
of the atmosphere
Table 4 (continuedj. 40.0 CM32I 64.0 1.29 I 1.06 I 52‘6 2.37 1 117% Averdge = 78.2
Dec. 6 Dec. 9 Dee. 14
Dec. 19
x x x x x
10-l” lo-l8 lo-‘8 lo-‘8 lo-‘*
The radon content of air on the campus of Fordham University was measured in 1941 and 1942 by V. F. HESS [lo], who used a diffe~ntial ionization chamber method ; the chambers were calibrated in curies by means of a radium standard solution and the samples were taken 12 m above ground. Our recent values of qx axe of the same order of magnitude as the average (97 x 10-i* c~ie~cm3~ of HESS’ value in 1941-1942. The following table gives the percentage x of (radon + decay products) in the t’otal ionization qILmeasured in all of those cases where qa was conveniently large so that the effect could be measured at a subsequent time with sufficient accuracy. Table 5 Date (1949) Nov. Nov. Nov. Kov. Dec. Dec. Dec.
(Radon
Fraction x + Decay Products) 0’
7-8 10-11 16-17 21-23 9-12 14-16 19-21 Average
is.4 60.3 65.6 43.0 71-4 63.1 46.5 = 59.0
These experiments urill be continued by FL P. YAXCO~. The percentage x for radon products is to be corrected for the presence of tioron We have already pointed out that our in the lowest part of the atmosphere. experimental procedure does not include the ionization produced by thoron itself, An estimate of this effect is however possible if we assume again that thoron is in equilibrium with (ThA t ThB + ThC) within the first few metres above the ground. Since the number of ions ‘produced by each alpha particle of TM is 1.92 x 106, of ThG is 1.71 x lo5 and of The’ is 2.54 x 105 while it, is 1=23x lo5 for thoron itself and TM?’ amounts to 6574 while ThC only to 35?b of the branched products, the average ratio of the ionization produced by Th,A + successive products to (Thoron + ThA + successive products) is 4.17 [1.92+(0.36x 1.71)-t (0.65 x 2+54)] x --____ lo5 -____-__Cl.23 + 1.92 + (0.35 x 1.71) + (0.65 x 2.54)] x lo6 = m”
o’77
If, therefore, we obtain an average value of 2=590/b for the radon family and (l -x)=41% for the (ThA+ ThC) products. the latter value would actually-if we want to include thoron-have to be raised. The content of ,radon plus decay products found (Table 5) is 78.2 x 0.59 x lO”s= The equivalent of thorium products (including t.horon) 46.2 x lo-l8 curie/cm3. vould
therefore be (??“-46’3) 0.77
10V1*=41*5 x lO_ls equivalent
curie/cm3.
Also we see
that the ionization produced by radon and its successive products at our point, of observation is 0,462 x IO-l6 curie/cm3 which corresponds to 0.93 I while thoron and its products would contribute 0.415 x lo-l6 curie/cm3 which corresponds to 0.83 1. The conclusion is that thoron and its decay prodwts contribute almost aa 23
VICTOR F. HESS and ROQER P. VAXCOUR smuch as radon and its successive products to the total alpha ray ionization in the atmosphere near the ground. This is in good agreement with the findings of other authors (see MEYER and SCHWEIDLER [9]) based on entirely different methods. :Etwould be of interest to check these conclusions in other localities. VI. THE IONIVTION BALANCE It is now possible to get a comparison between calculated and observed values of the total ionization at the point of observation. Let us consider first the ionization by gamma rays and cosmic rays : we observed with our ionization apparatus when it was not shielded at all and placed on the ground in the garden of the Seismic Station at Fordham University e,total ionization (qr) of 5.40 I (with chamber No. II). Subtracting its residual ionization (0.31 I) we have a total ionization of 5.09 I. Observations over water and on the campus of Fordham in the summers of 1947 and 1948 gave a value of about 3-OI for the gamma radiation from the ground (qr). Adding the value qc = 1.96 I (cosmic rays, chamber II) which includes the ionization qA by the g@mma rays from the atmosphere we get & total qr qT=qE+qc+qA
= 4.96 .i
which agrees very well with the experimental value. For atmospheric-electric considerations, we have now to add the ionization produced by the alpha rays from the radioactive substances in the atmosphere (qa) and by the beta rays of these products in the ground and in the atmosphere. In Section V we found that on the average (November-December 1949) the ionization produced by the alpha rays of radon and its decay products was 0.93 I while*thoron and its subsequent products gave 0.83 I. Thus the total ionization by alpha rays was l-761. The contribution by beta rays from radium, thorium, and their decay products in the ground was estimated by the senior author [5] as about O-2 to O-5I at a distance of 1 m above ground. (This is the distance at which most atmospheric-electric measurements are carried out). According to the .same author the ionization produced by beta rays of the radioactive products in the atmosphere (radon, thoron, and their decay products) will at most be 0.04 I. Summing up we have, therefore, the following picture of the total ionization in the atmosphere, 1 m above ground: rays Beta r8J’S
Alpha
1.76 I
Gamma
0.40I
rays
Cosmic rays
3.15 I 1.96 I
This total ionization (7.2 I) has, naturally, only local significance. In localities where the radon and thoron content of the atmosphere is higher, the ionization due to alpha rays may go up to more than 5 I: For an average radon content of 130x lo-l8 curie/cm3 the ionization due to Rn, RaA and RaC would amount to 2.75 I [9] and the products of thoron would add another 2-3 I. The number of small ions (n) present at our point of observation can be computed if we know the average life e of these ions. Since at this locality SCR~EIDLER’S linear recombination 1a.w [ll] was found to hold and the average life time (e) of small ions was determined recently by M. DONNELLY[12] as 10.6 set the equation, qe = n 24
The ionization balance of the atmosphere therefore would lead to an average value of n.=70 to 80 ions per cm3 which is of the expected order of magnitude in the suburbs of a big city where the average number of AITKEK nuclei is around 40000 per cm3. The senior author wishes to thank the dmnericn?a Philosophical Society in Philadelphia for a grant supporting this work and also Dr. 3%.A. TCVE, director of the Department of Terrestrial Magnetism (Carnegie Institution) for arranging construction work in the shop of his department.
REFERENCES [l] HESS, 1.. F. and TANCOUR, R. P.; Phys. Rev. 1949 76 1205. [2] HESS, V. F.; Trans. Amer. Geophys. Union 1946 27 670. [3] HESS, V. F. and ROLL, J. D.; Phys. Rev. 1948 73 592; ibid. 194873 916. 141 STEINKE,E.; Z. Phys. 1928 48 675. [5]HESS, V. F.; Die Ionisierungsbilanz der Atmosphlire, Leipzig: Akad. Cedagsges. 1934. [6] KOHLRAUSCR,K. K. F.; Radioaktivitirt, Handbuch der Experimentalph+k, l-01. 15. Leipzig: Akad. Verlagsges. 1928. [7] COMPTON,A. H. and HOPFIELD,J. J.; Rev..Sci. Instr. 1933 4 491. [8] DUAXE, W. and LABORDB,A.; Compt. rend. 1910 150 1421. [9] MEYER, ST. and SCHWEIDLER,E. ; Radioaktiritiit p. 587, Leipzig: Teubner 1927. [lo] HESS, V. F.; Terr. Meg. 1943 48 203. [ll] CHALNERS,J. ALAX; -4tmospheric Electricity, p. 48, Oxford: 1949. [12] DONRELLY, MARY; Thesis, New York: Fordham Univ. 1949.
balance of
the atmosphere
F. HESS and ROWER P. YANCOUR Fordham tiniversity, (Rec&ed
Xew York, N.T.
30 Janua7y 1950)
ABSTRACT Two methods were dereloped for determining the residual ionization in brass chambers and the absolute ionization of the hard and the soft component of cosmic radiation in the atmosphere at sea level, without secondaries from the walls of the chambers. The portable GISH-SHERMAX ionization meter in connection with four geometrically similar c$indrical vessels of widely different size was used. One method is based on the use of all four chambers, the other requires only one lerge chamber in which the pressure is varied between about O-1 and 1O atm. These measurements were carried out on the campus of Fordham University
in a garden.
It was found that, on bhe average, the hard component [filtered through 10 cm of iron) produces an ionization of 1.51 I while the total intensity (hardfsoft component) amounts t.o 1.88 1. The effect, of the gamma rags from the radioac~ire substances floating in the atmosphere was determined b,y a new method, described in this paper. This “ air radiation *’ amounts to about 0.13 I. varying ronslderably from day tG day. Another set of experiments was carried Gut to find t,he ionization produced by the alpha rays from radon. thoron, and their subsequent products. Samples of air from outdoors were filled into a large ionization eliamber and the varia.tion of ionization u-as observed for several days. In this way it was possible to determine the ionizing effect of the radon and thuron product.s separately. It was found that radon and its subsequent products give O-93I, while thoron and its products give O-83 1. The latter. t,herefore conrribut,e 47 “; of the t.Gt,alionization bj- alpha rays. The radon content itsetf was 46 x 1Vz8 curie per’ cm3 (average of November and December). The ionization by gamma rays from the ground, at Fordham. amounted to 3.0 1. The contribution by beta rays from radium, thorium and their decay products in the ground and in the atmosphere was estimated as 04 I. 1 m above gmund, tonsist,ed therefore of the
The total ionization, on the point of obserration, following four components : B&a rays Oamma rays Cosmic ra+s
(I.40 I 3.161 l.YOI
!
, ’
Total
: i.21 I
Using tile linear law of recombination of small ions (E. SCHKEIDLER, J. J. SOL~K and others) which holds for air polluted with a great number of nuclei. one ca.n compute the mean number of small ions per cm3 if one knows the average lifetime of such ions. Taking this lifetime as lO*(i set (from a recent determination at the same location), one would get less than 100 small ions per cm3 present in the stationary state. This is to be expected in tile vicinity of a big city, where the number Gf nuclei of condensation ordinarily exceeds 4OOOOper cnP,
Our knowledge of the ionization balance of the atmosphere is still rather incomplete. We know that over the continents the radioactive substances in the ground and in the air contribute most of the ionization observed while cosmic radiation is practically Here the variations of ionization are the only ionizing agency over the oceans. They are mainly due to variations of the barometric pressure relatively small. There are also other small and of the temperature in the lower atmosphere. fluctuations of cosmic ray intensity with time and with geomagnetic latitude which were studied so extensively within the last two decades. h’evertheless? data on the absolute intensity of cosmic radiation. even at sea level and at medium latitudes,
13
VICTOR F. HESS and ROGER P. Vmcow
are rather conflicting: for unshielded ionization chambers at geomagnetic latitudes beyond the “ knee “, the reported values range from 1.8 to 2.8 I (“1 ” denotes the number of pairs of ions produced in air per cm3, per set). It is to be noted that in all these measurements various authors did not take into account the increase of ionization by secondaries from the walls of the vessel. However, it is now generally believed that the additional ion production by secondaries in a thinwalled brass chamber is quite small, perhaps only a few per cent of the total effect. Quite recently the senior author in collaboration with R. P. VAXCOTJRdevised two methods [lJ of determining the absolute intensity of cosmic rays in the atmosphere without secondaries and also the residual io~zation in ionization chambers_ These methods, originally tried out in the laboratory, were used last summer in outdoor experiments which will be described in this paper (Sections I, II and IV). In addition, we determined the ionization produced by the gamma rays from the radioactive products in the atmosphere (Section III), the radon content of air and the relative contribution of thorium products in the atmosphere to the total ionizatioxrnear the ground (Section V). In Section VI, the ionization balance is discussed. All experiments were performed on the campus of Fordham University which ia situated in the Bronx, near the Botanical Gardens, New York City. They should give a good average picture of the io~zation balance in the lower atmosphere. I.
INTENSITY OF COSMX RADIATION AND THE RESIDUAL IONIZATION IX BUSS CHAMBERS
The ionization of the atmosphere at sea level, over land, is produced by the alpha, beta and gamma rays from the radioactive products in the soil and in the atmosphere, and, to a lesser estent, by cosmic radiation. If we measure this ionization in a hermetically sealed vessel (wall 2.5 mm thick) filled with a gas free of radon, the total ionization is due to: 1) occasional alpha rays from radioactive impurities in the metal of the ohamber (residual ionization, Q*); 2) gamma rays from the radioactive substances in the soil (a~) ; 3) gamma rays from the radioactive decay products of radon, thoron and actinon in the atmosphere &) ; and 4) cosmic rays (PC). q=qO+qE+qA+qC
If the ionization chamber is placed in an iron house with walls thick enough to absorb practically all the gamma rays (10 cm of iron), the total ionization is due to (lo and qC. In this case, we observe the effect of cosmic rays filtered through 10 cm of iron. If the apparatus is set up outdoors and the top of the iron house is lefi open, the ionization is due to unfiltered cosmic radiation and to the small gamma ray component from the radioactive products in the atmosphere qA which in most cases will not amount to much more than 0.1 I. The residual ionization (cfo) due to alpha rays from the inner surface of the wall of the chamber can be kept at much less than 1 I if the chamber walls are very carefully cleaned and the volume of the chamber is not too small. Since this residual ionization is proportional to the surface, while the other ionizing agencies are volume effects, it is advantageous to use chambers of fairly large size in which the ratio of surface area (A) to the volume (W) is rather small. The total ionization, as observed in the iron house (closed on top), is q=qo +
qc=Akna +qc 14
(1)
The ionization balance of the atmosphere
where n, denotes the number of alpha particles emitted from the wall per cm2 per set, A the surface area of the chamber wall, and k the average number of ion pairs produced by each alpha particle. The ionization produced by cosmic rays could also be considered partly as a surface effect since the secondaries produced in the walls of the chamber will increase (at constant thickness of the wall) with the area of the wall exposed to cosmic rays. Experiments to be discussed later will show that in thin-walled ionization chambers (as used by us) this effect of secondaries is very small. Therefore, we can treat the cosmic ray ionization as a volume effect, especially since the ranges of these secondaries are considerably longer than the diameter and height of our chambers. We have used ionization chambers of cylindrical shape, made of yellow brass (wall thickness 2.5 mm) in conjunction with the ionization meter, devised by 0. GISH, K. L. SHERMAN and one of the authors, which is described elsewhere [Z]. The chambers were geometricaliy similar (height equal to twice t’he diameter of the chamber) with volumes 4368, 13.17, 4.89 All four cylinders were and 1,645 1, and radii 15, 10.16, 7.30 and 5.08 cm respectively. airtight, and could be evacuated or Bled with any gas through a Hoke needle valve. We used air from steel cylinders (stored for at least one month). The amber-insulated inner electrode of each chamber is directly connected to the fibre system of a Lindemann electrometer which is mounted in a housing fitted exactly to the top of each of the chambers, which are used in a vertical position. The guard ring and the case of the electrometer are connected with a battery supplying 135 volts, and the centre point of the quadrant battery is kept at the same potential. A needle contactor is used for connecting or disconnecting the &bre system from the battery with which ‘the guard ring is permanently connected. The wall of each chamber is kept at ground potential. When the contactor is lifted the needle begins to float, and the rate of drift is a measure of the ionization in the vessel. We preferred to use the “ null-method “: a t-volt auxiliary battery with a potentiometer and a precision voltmeter (range 3 volts) allowed us to add any voltage from 0 to -3 volts to the 135.volt sweep voltage, thus inducing an opposite charge on the floating system just enough to keep the electrometer needle always at or close to the zero point. When using this nullmethod, all corrections for insulation leakage of the electrometer are avoided. The reading of the compensating voltage atethe end of each measurement (from 5 to 60 min in duration) multiplied by a reduction factor, which is proportional to the induction coefficient between the guard ring system and the inner electrode for each chamber, gives the ionization (q) expressed i ion pairs per cmJ per sec. The residual ionization of ihe three smaller chambers was determined in 1947 by using them within the iron house in a subway tunnel under 210 ft of solid rock where the cosmic-ray intensity is practically zero and local radiation was eliminated by the 10 cm iron wall around the apparatus [3]. These direct measurements served as a control for determinations of the resi ma1 ionization with the two simpler methods discussed below. We shall now describe one method used for the determination of the residual ionization and the intensit.y of cosmic rays in our brass chambers, and then proceed to the determination Jf the cosmic-ray intensity in t,he free at,mosphere. This method is based on the successive use of all four chambers in the iron house at the location where a determination of the cosmicray intensny is desired. After the tests in the laboratory which were described in our previous publication, all these experiments were performed outdoors : the iron house was set up in the garden of the Seismic Station of Fordham University on even ground. The numerical values for the dimensions of the four chambers are as follows: 15
VICTOR F. HESSant1 ROGERP. Va~ocrr
Utumber
II~~IL~I~( W) Inner (A= cm3
The total
Zl’urj’ace lot+)
-4/W
(r)
WI2
Cl)1
I II
43680 13173
7070 3243
15.0 10.16
O-1618 0.2462
III IV
4888 1645
16i4 Sll
7.30 5.08
O-3425 0.4930
ionization
current
in each chamber
I=eAkn Dividing we get
Radius
by eA and inserting
I
__
This is a linear relationship
between
to equation
(1) is (in e.a.u.)
(2)
.+eWqC
A = lCk*x and W=
1 Oer%
according
4r% (radius-2
=knc(+04
the total ionization
rqc
of the height
of each cylinder),
(3)
observed
and the radii of the chambers.
The experimental test of this relationship was carried out for two cases; with the iron house set up in the garden of the Seismic Station at Fordham, we put each of the four chambers successively in the iron house and observed the ionization in each chamber a) when t’he iron house was completely closed (10 cm iron on all sides); b) when the top of the iron house was removed. In this case, the soft component of the cosmic radiation comes into play and, in addition, a small part (qA) of the observed ionization is contributed by the gamma rays of the radioactive substances in the lower atmosphere (up to about 4 km from the ground). In this particular case the term qC in equation (3) should be replaced by (q, + qA), while kn, remains We, therefore, can expect to get two straight lines with different slopes; their unchanged. intercept should coincide at the axis of the ordinates.
As can be seen from Figure 1, our experimental with this expectation.
points are in perfect agreement
It may be stated that our value for the residual ionization, as found here, was even lower than the one reported in our previous publication, when the chambera were carefully cleaned, repeatedly evacuated. and refilled with air from our t,nnk. ThB extrapolated value of IllOde for the radius r = 0 represents kna. Taking for k (the average number of pairs of ions produced by one alpha p&icle) 105, we obtain for Finxti l-25 and the number of particles from the wall (per cm2) 1.25 x 10-j per set or 4.5 x 1O-2 per hr which is of t,he same order of magnitude as the lowest alpha-ray emission reported so far by J. A. BEARDEN, for steel (O-10 >er hr). We are now in a position to compute the residual ionization (q,,) simply hy multiplying knr= 1.26 with the surface area -1 and dividing by the volume VV for each chamber. Figure 2 gives the curve (arc of a parabola) obtained when these computed values of q. are plotted against the radii of the chambers. The values q. for the four chambers obtained were 0.20 I for No. I, O-31 I for No. II, 0.43 I for No. III. and 0.62 I for Ko. IV. Three of these were checked directly by observing the residual ionizat’ion of three chambers by an underground experiment when they were placed within the iron house in a subway tunnel under 16
The ionization balance of the atmosphere 70 m of
5 1oq;
The comparisons gave a satisfactory agreement within solid rock. between experimental and computed values. Subtracting the values q,,
1
8 72 761x1~ P-r Fig. l-Extrapolation of residual ionization from measurementsof four chambers. 8
Y
Fig. 2-Residual ionimtion vs. radius of the ionizationchambers.
given above from the actually observed values of total ionization house, the following cosmic ray intensities (qc) are obtained.
in the iron
Table 2 Cosmic Ray
Intensities
Chamber Xo.
I II III IV
(Sea Level) in Brass Chambers (Wall 2-5 mm thick)
(a) Iron House C1ose.d (IO cm Fe)
1.52 1.55 1.54 1.55
(b) Iron House. Top Removed
I I
1.94 I 1.96 I
I I
1.94 I 1.99 I
The values in column (a) correspond to the hard component, in column (b) to the hard plus the soft component (including a small effect qA of the gamma rays of the radioactive substances in the air). II.
DETERMWATIOK OF THE TREE VALUE OF THE IONIZATION BY COSMIC RAYS IN THE FREE BTMO~PHERE: WITHOUT SECONDARIES
In order to obtain the ionization without secondaries from the walls of the chamber, we plotted the observed values of q = Ah + qcf (total we proceeded as follows: The extrapolated value of q for ionization observed in the iron house) vs. A/W. A/W=0 (infinite volume) would indicate the ionization produced in a chamber of infinite volume filled with radon-free air at atmospheric pressure, at the point We obtained a linear relationship of q vs. A/W for all four chambers of observation. in both cases 1) iron house closed, 10 cm Fe around the chambers and 2) with the Case (1) gives an extrapolated value of 1.51 I; t,op of Dhe iron house removed. Case (2), l-931 (see Figure 3). These values were confirmed by direct computations from the equations of the two straight lines. 17
B
VICTOB F. HESS=d
ROGERP. VANCOUR
These figures actually represent the true ion production in air without any secondaries emitted from the walls, since in a chs.mber of infinite volume any contribution of residual ionization (alpha rays from the walls) as well as from electrons knocked out of the walls by cosmic rays would be negligible. The two figures are only a few per cent lower than the ones given in Table 2 for the brass chambers. The latter values obviously do not represent the true ionization in the atmosphere at the point of observation (50 degrees N geomagnetic latitude). However, the difference is very small, especially for the largest chamber No. I (43 1 volume) ; this indicates that the secondaries from the brass walls (2.5 mm thick) increase the ionization by only s, few per cent. The difference between the two values l-93 I and 151 I gives, approximately, the ionization produced by the soft component of cosmic radiation. This value (O-42 I) amounts to about 20% of the total effect of cosmic rays at sea level, and this agrees with previous estimates of various authors. However, it is importent to remember that the still somewhat uncertain and variable effect of the gamma rays from radioactive substances floating in the lower atmosphere is included in the figure l-93 I. This effect was determined (see the following section) as a7 017 ay 55 a6 about O-15 1 (average). Subtracting it from l-93 I A/w-. we get 1.78 I as the mean total ionization by cosmic Fig. 3-ExtmpoLtion of ionim- rays (soft and hard component) rtt sea, level, in the tion bY coamic m~s in the free free atmosphere. A small correction of this figure is hXl0fpheP3. necessary on account of the fact that the ionization chambers in the iron house are not recording the actual total intensity of the soft component when the top of the iron house is taken off. Rays coming in from the lower regions of the sky are partly absorbed when penetrating the 10 cm iron shield in an oblique direction. We computed from the geometry of the whole arrangement the absorbed fraction of the soft component, assuming the distribution of the unidirectional intensity of cosmic rays according to STEINKE’S metimements [4]. The results indicate that the absorbed fraction of the soft component ctmounts to about O-1 1. The total ionization of the soft and hard component, therefore, is l-78 + O-10= 1.88 I at sea level in the free atmosphere (50 degrees geomagnetic latitude N) while the hard component alone amounts to l-51 I. III.
APPROXIMATE DETERMINATION OF THE “Am
(Ionizcctionby Gamma
Rays from
the Radioactive
Substances
RADIATION” in the Atmmpphere)
The “ air radiation ” (qd ) was estimated by V. F. HESS [5] in 1934 as about O-1 I. A better estimate was derived directly from our experiments ; since we know the sverage absorption coefficients of the gamma rays from RaC and ThC we csn estimate the percentage loss in an absorbing p&e of known thickness. We took two iron plates (together l-2 cm thick) and observed repeatedly the ionization in the iron house when its top was off and when the two plates were placed on the top. Calling the two values q1 and qz we have the relation
(1 -eqpd”d)
q1 -h=Ph 18
The ionization
balance of the atmosphere
where p is the average absorption coefficient [6] of hard gamma rays in iron (0.356 cm-l). The absorption of the “ soft ” as well as of the hard component of The term (1 -e-Id) corresponds to the ftaction cosmic rays is hereby neglected. of gamma rays absorbed in 1.2 cm Fe. We performed seven long series of observations alternatively with “ top off” and “ 1.2 cm Fe on top ” with chambers I and II in the iron house between July 26 and September 7, 1949, and obtained the following results : Table 3 Difference of Tonitation Observed $?I;$
Average
From
these figures we determined 0*0691 =qA
0.050 I 0.223 I 0.044 I 0.035 I 0.012 I 0.048 I 0.0691 I _cO*O26I (probable error f
O-0178 I)
qA, using the equation
( 1-e-o*427) and get, qA =is
= 0.198 I
The large fluctuations of the single values given above (although partly statistical) This can be expected since we know that the indicate that the effect is variable. The value radon content of the atmosphere is varying greatly, from day to day. mentioned above gives an upper limit of the average value qA since it was tacitly assumed that the whole difference (ql - q2) is due to the absorption of gamma rays from the radioactive substances in the atmosphere in 1.2 cm iron. However, we can derive an estimate also of a lower limit of q4 by computing the possible effect of the absorption of the soft component ,of cosmic radiation in 1.2 cm Fe; taking the mass absorption coefficient of this soft component as 8.6 x 10m3 cm2/g (E. REGENER) and the water equivalent of 1.2 cm Fe as 9.4 cm the exponential e-o.oo*5 x g*4 gives O-923. Since the soft component could, according t’o our measurements, amount to not more than the total difference of 0.42 I observed when the top of the iron house (10 cm Fe) was off and on, we see that 0.42 x 0*923= O-388 I would be the amount getting through l-2 cm Fe end only If we assume that all of the difference of the ionization 0.03 I would be absorbed. tentatively observed with the “ top on ” (10 cm Fe) and “ top off “-arrangement was due to the soft component of cosmic radiation, we can now derive a lower limit for the amount of air radiation (qa): Since we observed a difference q1 - q2= 0.069 I a O-018 I when the iron house was completely open on the top and when it was covered with 1.2 cm Fe we can now say that at the most only 0.03 I of this difference can be ascribed to absorption Subtracting this from O-069 I we of the soft component of cosmic radiation. get O-039 I as the lower limit for (ql - q2) to be ascribed to absorption of gamma Inserting rays of the radioactive substances in the atmosphere in 1.2 cm iron. this value in our equation (4) we get O-039 = qA (1 - e-o*427) = 0.348 q1 and from this ql=0.112 I 19
VKCTOR F. HESSand ROC+ER P. VANCOUR
Thus we are able to bracket the true value of the air-radiation qA as lying somewhere between 0.198 and 0.112 I and taking as an average for qA= 0.161 we see that this is about six times the mean average error of our observed values for (ql - qe). Thus, it is evident that the contribution of gamma rays from the radioactive substances in the air is real and cannot be neglected in the ionization balance. Its variability is clearly reflected in our measurements (Table 3). These variations naturally will very often mask a&us1 fluctuations of cosmic radiation, if the ionization chamber is not shielded on the top against these “atmospheric ” gamma rays. XV.
ANOTHER METHOD FOR DETE~~~~
TEE IONIZATION BY COSMXCRAYS AND THE RESIDUAL IONIZATION OF IONIZATION CHAMBERS
This method was described in a previous publication [l] in detail. Since it was u&i a&n la& summer in our experiments outdoors, a short explanation of our procedure will be necessary, It is well known that the ionizationin a closed vessel incre~es almost linearly with pressure, but the curve flattens out st b&her Dmssures except when very mrre 8sgon is used F71. Since -we ~snted to wo& with sir in the chamber, we preferred to use low pressures up to atmospheric pressure ss C. T. R. W-ON did in his earlv ermeriments. We filled our lamest chamber (43.7 1) with “air &om & tsnk (stored for 2 least one month) at stmospheric pressure and observed the ion&&ion at that pressure and at various lower pressures when the chamber wss placed in the iron house. It wsa to be expected that the ion&&ion vs pressure curve would be linear &Blong se the pressure p (expressed in atrn~~e~) wss not lower then that at whiuh the range of alpha particles is equal to
the diameter of the chamber (30 cm). At still lower pressure, the curve was expected to deviate downward toward zero.
Since the current .I- Aet&a + Weqcp and the first term, is constant as long as the range does not exceed 30 cm we obtain
0.80
Pfe.ssufe
The slope of the linear part of the q vs. p curve, therefore, corresponds to the actual total ionization of the cosmic radiation in the largest brass chamber, while the extrapolation of the linear part of the curve for p= 0 would give the approximate value of the residual mmHg ionization (q,,) ; it can be shown that this would be
Fig. P-Ionization by cosmic rays vs. pressure.CurveI: hard component; Curve II: soft -Ihad component.
Qo-P-
dq dp’
P
where q denotes the total ionization observed at the pressure r). It is, of course, more accurate to obtain dq/dp from the graph directly. Figure 4 shows two sample curves obtained with air in the chamber. Curve I corresponds to the case when the iron house was closed (10 cm Fe surrounding the chamber on all sides). In this case, only the hard component of the cosmic rays was effective. Curve II (top of iron house removed) gives the values for soft and hard components together. The iron bars in this case reached up just to the upper rim of the ionization chamber. Therefore, the soft 20
The ionization
balance of the atmosphere
component (plus a small amount of gamma rays from the atmosphere) a solid angle which was less than 2~: sterradian.
came from
The value of dq/dp can be corrected for this case (half space) by graphical methods. From the slope of the two curves. we obtain for Case I, qc = l-45 I (Meson The extracomponent) and for Case II, qc = l-84 I (soft and Meson component). polated value (0.20 I) for the residual ionization (qo) agrees quite well with the The correction mentioned above and disvalue obtained with the other method. cussed at, the end of Section II amounts t,o 0.1 I and therefore the value for the ionizat,ion due to the soft and hard component is 1.84 + 0.10 = l-94 I which is in good agreement with the value derived in Sect,ion II. T’.
THE IOKIZATION PRODFCED BY RADON. THORON,AND THEIR DECAY PRODUCTS IN THE ATMOSPHERE A very simple method of det,ermining the ionization produced by the alpha rays of these
products was tried successfully during the fall of 1949 when the outdoor experiments were completed: we assembled the iron house in the basement laboratory of the Physics Building at Fordha,m and det.ermined first the ionization in the chamber II (13 1) when it was filled with aged, inactive air from a steel cylinder while standing in the iron house. Then the chamber was evacuated and refilled with atmospheric air in the garden of the seismic station where all our previous gamma ray and cosmic ray experiments were performed. The atmospheric air samples were taken about 3 ft above the ground. After filling, the chamber was put. back into the iron house and the total ionization measured immediately. In a number of cases, the readings were continued for several hours and further readings taken on the following day and even two or three days afterward, in order t,o get, a decay curve.
It is clear that t’he difference between t,he total ionization observed in the iron house with aged (inactive) air and the atmospheric air in the chamber corresponds to the ionization by t,he alpha rays of radon, thoron and their decay products (pa), Thoron itself will: of course, not be measured since all of it would decay within the ten minutes or more which ela’psed between the moment when the sample was taken and when t,he first measurement could be made. The effects observed. therefore: are mainly due to Radon and its decav products (Rn + RaA The half life of ThB (10.6 hrs)“is short, as compared +RaC) and (ThB f 5°K’). with the one of Rn (3-82 days) and so we can expect that within 2-3 days most of the ThB has decayed and only the Rn, in equilibrium with its decay products. survives. The very low background of our chambers plus. the cosmic ray ionization (q. + qc) in t,he basement laboratory amounting together to only about 1.6 I allowed us to detect quite easily and wit,h sufficient accuracy t,he effect (aa) of the alpha rays from the radon, ThB, etc. int’roduced into the chamber when the ,atmospheric air sample was admitted in the chamber. This effect amounted in most’ cases to between 1 and 2 I. By taking readings immediately after the sample was tilled into the chamber and one or two days later, it was even possible to determine the percentage of ionization due to thorium products. This was accomplished in the following way: When atmospheric air is introduced in the chamber it produces an ionization qa which, added to the ionization produced by cosmic rays and the residual (wall) radiation gives a total effect of Q1=Qo+9c+46!
(5)
Measuring (q. + qc) in the chamber before the experiment, we immediatelv obtain qkis comqa about ten minutes after the atmospheric air sample is introduced. 21
VICTOBF. Fhs and ROGERP. VANCOUR
posed of the ionizstion produced by (Rnf RaA + . . .) and (ThB+ThC+ . . .), both assumed to be in equilibrium. We denote the decay constant of radon with hr (2.097 x 10-S se+) and of thorium B 5ts i2 (l-82 x 10m5set-1) and assume that a fraction z of QO~ is due to (Rn + decay products) while the fraction (1 -x) of qa is due to the decay products of ThB. Then, we have, at a known time t later z$$ -*It + (1 - 2) q,e- Aat = q;I - A,t _ e -ii,t ) = q’,/q, - e- Aat
or
x(e
PorlPor-e-hd -__ x=&t_, -A,t
and
From this, we computed the relative contribution of the decay products of radium and thorium in the atmospheric air sample in each case, when pa and ql, (at least 10 hrs later) were determined. We carried out a number of experiments determining qa and the frrtction z of radium products in November and December 1949. The results are listed below in Tables 4 and 5. A slight correction has to be applied to the measured values of qa on account of the fact that alpha rays emitted from points near the chamber walls are not fully utilized. An empirical correction was worked out by W. DUANE and A. LABORDE [8] for radon and its decay products in a cylindrical chamber ; the corrected value is obtained by dividing the measured value by (l-o.572 4) where A is the surface and W the volume of the ionization chamber. Accepting this correction as approximately valid also for the .decay products of ThB.and taking the value of A/W as O-249 for our chamber (No. II), we have
l-0.572
;=
0.8575
The values qa listed in the table are already corrected in this way. The third column (equivalent in cu.ries/cm3) gives the number of curies of radon plus decay products which would produce the same ionization ss the one actually measured in each case ; this would include the thorium products, expressed in equivalent amounts of mdon. The computation of the radon equivalent in curies per cm3 from the observed ionization Qa was based on the values given by ST. MEYER and E. SCEWEIDLER [9] : 1 curie (radon) per cm3 gives a saturation current of 2.75 x lo6 e.s.u. and 1 curie of radon with its decay products 2.75 x 106x 1.55+1’70+2*20=-9.67x 106 e .s .u .). An amount of radon of lo-l6 curie/ 1.55
cm3, therefore, corresponds to 2.01 I. Date (1949) Nov. 3 Nov. 4 Nov. 7 Nov. 10 Nov.. 14 Nov. 15 Nov. 16 Nov. 21 Dec. 5
q,
(cow.)
Table 4 Radon equivalent in curie/cm3
1.61I 1.28I 2.21I 1*57I 1.02I 1.58I 1.43I 1.56I 2.67I
79.8x IO--‘@ 63.5x lo-l8 109.6x lo-l8 78.1x lo-*8 5o?Jx lo-18 78.5x lo-‘* 70.9 x 1O-‘8 77.3 x lo-‘8 132*6x IO-‘* 22
The ionization
balance
of the atmosphere
Table 4 (continuedj. 40.0 CM32I 64.0 1.29 I 1.06 I 52‘6 2.37 1 117% Averdge = 78.2
Dec. 6 Dec. 9 Dee. 14
Dec. 19
x x x x x
10-l” lo-l8 lo-‘8 lo-‘8 lo-‘*
The radon content of air on the campus of Fordham University was measured in 1941 and 1942 by V. F. HESS [lo], who used a diffe~ntial ionization chamber method ; the chambers were calibrated in curies by means of a radium standard solution and the samples were taken 12 m above ground. Our recent values of qx axe of the same order of magnitude as the average (97 x 10-i* c~ie~cm3~ of HESS’ value in 1941-1942. The following table gives the percentage x of (radon + decay products) in the t’otal ionization qILmeasured in all of those cases where qa was conveniently large so that the effect could be measured at a subsequent time with sufficient accuracy. Table 5 Date (1949) Nov. Nov. Nov. Kov. Dec. Dec. Dec.
(Radon
Fraction x + Decay Products) 0’
7-8 10-11 16-17 21-23 9-12 14-16 19-21 Average
is.4 60.3 65.6 43.0 71-4 63.1 46.5 = 59.0
These experiments urill be continued by FL P. YAXCO~. The percentage x for radon products is to be corrected for the presence of tioron We have already pointed out that our in the lowest part of the atmosphere. experimental procedure does not include the ionization produced by thoron itself, An estimate of this effect is however possible if we assume again that thoron is in equilibrium with (ThA t ThB + ThC) within the first few metres above the ground. Since the number of ions ‘produced by each alpha particle of TM is 1.92 x 106, of ThG is 1.71 x lo5 and of The’ is 2.54 x 105 while it, is 1=23x lo5 for thoron itself and TM?’ amounts to 6574 while ThC only to 35?b of the branched products, the average ratio of the ionization produced by Th,A + successive products to (Thoron + ThA + successive products) is 4.17 [1.92+(0.36x 1.71)-t (0.65 x 2+54)] x --____ lo5 -____-__Cl.23 + 1.92 + (0.35 x 1.71) + (0.65 x 2.54)] x lo6 = m”
o’77
If, therefore, we obtain an average value of 2=590/b for the radon family and (l -x)=41% for the (ThA+ ThC) products. the latter value would actually-if we want to include thoron-have to be raised. The content of ,radon plus decay products found (Table 5) is 78.2 x 0.59 x lO”s= The equivalent of thorium products (including t.horon) 46.2 x lo-l8 curie/cm3. vould
therefore be (??“-46’3) 0.77
10V1*=41*5 x lO_ls equivalent
curie/cm3.
Also we see
that the ionization produced by radon and its successive products at our point, of observation is 0,462 x IO-l6 curie/cm3 which corresponds to 0.93 I while thoron and its products would contribute 0.415 x lo-l6 curie/cm3 which corresponds to 0.83 1. The conclusion is that thoron and its decay prodwts contribute almost aa 23
VICTOR F. HESS and ROQER P. VAXCOUR smuch as radon and its successive products to the total alpha ray ionization in the atmosphere near the ground. This is in good agreement with the findings of other authors (see MEYER and SCHWEIDLER [9]) based on entirely different methods. :Etwould be of interest to check these conclusions in other localities. VI. THE IONIVTION BALANCE It is now possible to get a comparison between calculated and observed values of the total ionization at the point of observation. Let us consider first the ionization by gamma rays and cosmic rays : we observed with our ionization apparatus when it was not shielded at all and placed on the ground in the garden of the Seismic Station at Fordham University e,total ionization (qr) of 5.40 I (with chamber No. II). Subtracting its residual ionization (0.31 I) we have a total ionization of 5.09 I. Observations over water and on the campus of Fordham in the summers of 1947 and 1948 gave a value of about 3-OI for the gamma radiation from the ground (qr). Adding the value qc = 1.96 I (cosmic rays, chamber II) which includes the ionization qA by the g@mma rays from the atmosphere we get & total qr qT=qE+qc+qA
= 4.96 .i
which agrees very well with the experimental value. For atmospheric-electric considerations, we have now to add the ionization produced by the alpha rays from the radioactive substances in the atmosphere (qa) and by the beta rays of these products in the ground and in the atmosphere. In Section V we found that on the average (November-December 1949) the ionization produced by the alpha rays of radon and its decay products was 0.93 I while*thoron and its subsequent products gave 0.83 I. Thus the total ionization by alpha rays was l-761. The contribution by beta rays from radium, thorium, and their decay products in the ground was estimated by the senior author [5] as about O-2 to O-5I at a distance of 1 m above ground. (This is the distance at which most atmospheric-electric measurements are carried out). According to the .same author the ionization produced by beta rays of the radioactive products in the atmosphere (radon, thoron, and their decay products) will at most be 0.04 I. Summing up we have, therefore, the following picture of the total ionization in the atmosphere, 1 m above ground: rays Beta r8J’S
Alpha
1.76 I
Gamma
0.40I
rays
Cosmic rays
3.15 I 1.96 I
This total ionization (7.2 I) has, naturally, only local significance. In localities where the radon and thoron content of the atmosphere is higher, the ionization due to alpha rays may go up to more than 5 I: For an average radon content of 130x lo-l8 curie/cm3 the ionization due to Rn, RaA and RaC would amount to 2.75 I [9] and the products of thoron would add another 2-3 I. The number of small ions (n) present at our point of observation can be computed if we know the average life e of these ions. Since at this locality SCR~EIDLER’S linear recombination 1a.w [ll] was found to hold and the average life time (e) of small ions was determined recently by M. DONNELLY[12] as 10.6 set the equation, qe = n 24
The ionization balance of the atmosphere therefore would lead to an average value of n.=70 to 80 ions per cm3 which is of the expected order of magnitude in the suburbs of a big city where the average number of AITKEK nuclei is around 40000 per cm3. The senior author wishes to thank the dmnericn?a Philosophical Society in Philadelphia for a grant supporting this work and also Dr. 3%.A. TCVE, director of the Department of Terrestrial Magnetism (Carnegie Institution) for arranging construction work in the shop of his department.
REFERENCES [l] HESS, 1.. F. and TANCOUR, R. P.; Phys. Rev. 1949 76 1205. [2] HESS, V. F.; Trans. Amer. Geophys. Union 1946 27 670. [3] HESS, V. F. and ROLL, J. D.; Phys. Rev. 1948 73 592; ibid. 194873 916. 141 STEINKE,E.; Z. Phys. 1928 48 675. [5]HESS, V. F.; Die Ionisierungsbilanz der Atmosphlire, Leipzig: Akad. Cedagsges. 1934. [6] KOHLRAUSCR,K. K. F.; Radioaktivitirt, Handbuch der Experimentalph+k, l-01. 15. Leipzig: Akad. Verlagsges. 1928. [7] COMPTON,A. H. and HOPFIELD,J. J.; Rev..Sci. Instr. 1933 4 491. [8] DUAXE, W. and LABORDB,A.; Compt. rend. 1910 150 1421. [9] MEYER, ST. and SCHWEIDLER,E. ; Radioaktiritiit p. 587, Leipzig: Teubner 1927. [lo] HESS, V. F.; Terr. Meg. 1943 48 203. [ll] CHALNERS,J. ALAX; -4tmospheric Electricity, p. 48, Oxford: 1949. [12] DONRELLY, MARY; Thesis, New York: Fordham Univ. 1949.