Zu¨rich University ZU-TH 1/96 MICROLENSING IMPLICATIONS FOR HALO 6 DARK MATTER 1 9 9 1 Philippe Jetzer 2 and Eduard Mass´o 3 n a 2 Paul Scherrer Institut, Laboratory for Astrophysics, J CH-5232 Villigen PSI, 3 2 and Institute of Theoretical Physics, University of Zu¨rich, 1 v Winterthurerstrasse 190, CH-8057 Zu¨rich, Switzerland 2 3Departament de F´ısica and IFAE, Universitat Aut`onoma de Barcelona, 2 1 E-08193 Bellaterra, Spain. 1 0 6 9 Abstract / h p The most accurate way to get information on the mass of the - MACHOs (Massive Astrophysical Compact Halo Objects) is to o use the method of mass moments. For the microlensing events r t detected so far by the EROS and the MACHO collaborations s a in the Large Magellanic Cloud the average mass turns out to : v be 0.08M⊙. Assuming a spherical standard halo model we find i that MACHOs contribute about 20% to the halo dark matter. X The eleven events recorded by OGLE, mainly during its first two r years of operation, in the galactic bulge lead to an average mass a of 0.29M , whereas forty events detected by MACHO during its ⊙ first year give 0.16M , thus suggesting that the lens objects are ⊙ faint disk stars. 1 Talk presented by Ph. Jetzer at the second workshop on “The dark side of the Universe: experimentaleffortsandtheoreticalframeworks”(Rome,13-14November1995). 1. Introduction It has been pointed out by Paczyn´ski [1] that microlensing allows the detection of MACHOs in the mass range [2] 10−7 < M/M < 10−1. Starting ⊙ from September 1993 the French collaboration EROS [3] and the American– Australian collaboration MACHO [4] announced the detection of at least six microlensing events discovered by monitoring over several years millions of stars in the Large Magellanic Cloud (LMC). Moreover, the Polish-American collaboration OGLE [5] and the MACHO team [6] found altogether more than ∼ 100 microlensing events by monitoring stars located in the galactic bulge. The inferred optical depth for the bulge turns out to be higher than previously thought. An important issue is the determination of the mass of the MACHOs that acted as gravitational lenses as well as the fraction of halo dark mat- ter in form of MACHOs. The most appropriate way to compute the average mass and other important information is to use the method of mass moments developed by De Ru´julaet al. [7], which will bebriefly presented in section 3. 2. Most probable mass for a single event First, we compute the probability P that a microlensing event of duration T and maximum amplification A be produced by a MACHO of mass max µ (in units of M ). Let d be the distance of the MACHO from the line ⊙ of sight between the observer and a star in the LMC, t=0 the instant of closest approach and v the MACHO velocity in the transverse plane. The T magnification A as a function of time is calculated using simple geometry and is given by u2 +2 d2 +v2t2 A(t) = , where u2 = T . (1) u(u2 +4)1/2 R2 E R is the Einstein radius which is R2 = 4GMDx(1−x) = r2µx(1−x) with E E c2 E M = µM the MACHO mass and D (xD) the distance from the observer ⊙ to the source (to the MACHO). D = 55 kpc is the distance to the LMC and r = 3.17×109 km. We use here the definition: T = R /v . E E T We adopt the model of an isothermal spherical halo in which the normal- ized MACHO number distribution as a function of v is T 2 f(v )dv = v e−vT2/vH2 dv , (2) T T v2 T T H 1 with v ≈ 210 km/s the velocity dispersion implied by the rotation curve of H our galaxy. The MACHO number density distribution per unit mass dn/dµ is given by dn dn a2 +R2 dn = H(x) 0 = GC 0, (3) dµ dµ a2 +R2 +D2x2 −2DR xcosα dµ GC GC with dn /dµ the local MACHO mass distribution. We have assumed that 0 dn/dµ factorizes in functions of µ and x [7]. We take a = 5.6 kpc as the galactic “core” radius (our final results do not depend much on the poorly known value of a), R = 8.5 kpc our distance from the centre of the galaxy GC and α = 820 the angle between the line of sight and the direction of the galactic centre. For an experiment monitoring N stars during a total ob- ⋆ servation time t the number of expected microlensing events is given by obs [7, 8] dn N = dN = N t 2Dr v f(v )(µx(1−x))1/2H(x) 0dµdu dv dx ev ev ⋆ obs E T T min T dµ Z Z (4) where the integration variable u is related to A : A = A[u = u ]. min max max min For a more complete discussion in particular on the integration range see [7]. From eq.(4) with some variable transformation (see [9]) we can define, up to a normalization constant, the probability P that a microlensing event of duration T and maximum amplification A be produced by a MACHO of max mass µ, that we see first of all is independent of A [9] max µ2 1 r2µx(1−x) P(µ,T) ∝ dx(x(1−x))2H(x)exp − E . (5) T4 0 v2 T2 ! Z H We also see that P(µ,T) = P(µ/T2). The measured values for T are listed in the Tables 1 and 2, where µ is the most probable value. We find that MP the maximum corresponds to µr2/v2 T2 = 13.0 [9, 10]. The 50% confidence E H interval embraces for the mass µ approximately the range 1/3µ up to MP 3µ . Similarly one can compute P(µ,T) also for the bulge events (see MP [10]). Table 1: Values of µ (in M ) for the six microlensing events detected MP ⊙ in the LMC (A = American-Australian collaboration events (i = 1,..,4); F i 1 and F French collaboration events). For the LMC: v = 210 km s−1 and 2 H r = 3.17×109 km. E 2 A A A A F F 1 2 3 4 1 2 T (days) 16.9 9 14 21.5 27 30 τ(≡ vHT) 0.097 0.052 0.08 0.123 0.154 0.172 rE µ 0.12 0.03 0.08 0.20 0.31 0.38 MP Table 2: Values of µ (in M ) as obtained by the corresponding MP ⊙ P(µ,T) for eleven microlensing events detected by OGLE in the galactic bulge [10]. (v = 30 km s−1 and r = 1.25×109 km.) H E 1 2 3 4 5 6 7 8 9 10 11 T 25.9 45 10.7 14 12.4 8.4 49.5 18.7 61.6 12 20.9 τ 0.054 0.093 0.022 0.029 0.026 0.017 0.103 0.039 0.128 0.025 0.043 µ 0.61 1.85 0.105 0.18 0.14 0.065 2.24 0.32 3.48 0.13 0.40 MP 3. Mass moment method A more systematic way to extract information on the masses is to use the method of mass moments as presented in De Ru´jula et al. [7]. The mass moments < µm > are defined as dn < µm >= dµ ǫ (µ) 0µm . (6) n dµ Z < µm > is related to < τn >= τn, with τ ≡ (v /r )T, as constructed events H E from the observations and which can also be computed as follows P < τn >= dN ǫ (µ) τn = Vu Γ(2−m)H(m) < µm > , (7) ev n TH Z with m ≡ (n+1)/2 and c N t V ≡ 2N t D r v = 2.4×103 pc3 ⋆ obs , (8) ⋆ obs E H 106 stars/year ∞ v 1−n Γ(2−m) ≡ T f(v )dv , (9) T T Z0 (cid:18)vH(cid:19) 1 H(m) ≡ (x(1−x))mH(x)dx . (10) 0 Z c 3 The efficiency ǫ (µ) is determined as follows (see [7]) n dN⋆ (µ¯) ǫ(T) τn ǫ (µ) ≡ ev , (11) n dN⋆ (µ¯) τn R ev R where dN⋆ (µ¯) is defined as dN in eq.(4) with the MACHO mass distri- ev ev bution concentrated at a fixed mass µ¯: dn /dµ = n δ(µ − µ¯)/µ. In Fig.1 0 0 we show the experimental detection efficiency ǫ(T) of the MACHO exper- iment when looking to the LMC [11]. In Fig.2 we plot the corresponding ǫ (µ) as calculated from eq.(11). This function indicates how efficient is the 0 experiment to detect a MACHO with a given mass M = µM . ⊙ A mass moment < µm > is thus related to < τn > as given from the measured values of T in a microlensing experiment by < τn > < µm >= . (12) Vu Γ(2−m)Hˆ(m) TH The mean localdensity ofMACHOs (number per cubic parsec) is< µ0 >. The average local mass density in MACHOs is < µ1 > solar masses per cubic parsec. The mean MACHO mass, which we get from the six events detected so far toward the LMC, is [10] < µ1 > = 0.08 M . (13) < µ0 > ⊙ (To obtain this result we used the values of τ as reported in Table 1, whereas Γ(1)H(1) = 0.0362 and Γ(2)H(0) = 0.280 as plotted in figure 6 of ref. [7]). The mean MACHO mass, which one gets fromthe eleven events of OGLE in thce galactic bulge is ∼ 0c.29M [10]. From the 40 events discovered 2 ⊙ during the first year of operation by the MACHO team [6] we get an average value of 0.16M . The lower value inferred from the MACHO data is due ⊙ to the fact that the efficiency for the short duration events (∼ some days) is substantially higher for the MACHO experiment than for the OGLE one. The above average values for the mass suggests that the lens are faint disk stars. The resulting mass depends obviously to some extent on the parameters used to describe the halo (or the galactic centre respectively). In order to 2 We considered only the events used by the MACHO team to infer the optical depth without the double lens event. 4 check this dependence we varied the parameters within their allowed range and found that the average mass changes at most by ± 30%, which shows that the result is rather robust. Another important quantity is the fraction f of the local dark mass den- sity (the latter one given by ρ ) detected in the form of MACHOs, which 0 is given by f ≡ M /ρ ∼ 126 pc3 < µ1 >. Using the values given by the ⊙ 0 MACHO collaboration for their first year data [11] (in particular u = 0.83 TH corresponding to A > 1.5 and an effective exposure N t of ∼ 2×106 star- ⋆ obs years for the observed range of the event duration T between 10 - 20 days) we find f ∼ 0.2, which compares quite well with the corresponding value (f = 0.19+0.16) obtained by the MACHO group in a different way. −0.10 Once several moments < µm > are known one can get information on the mass distribution dn /dµ. However, since at present only few events toward 0 the LMC are at disposal the different moments (especially the higher ones) can only be determined approximately. Instead, we can make the ansatz dn /dµ = aµ−α. Knowing, for instance, < µ1 > and < µ0 > (as well as ǫ (µ) 0 1 and ǫ (µ) from eq.(11)) we can determine a and α. The solution for a and −1 α is acceptable only if we get the same values using other moments, such as e.g. < µ1.5 >. Remarkably, we find that a ≃ 6.5 × 10−4 and α ≃ 2 is a consistent solution. Moreover, from the relation ∼0.1 dn 0 µdµ = fρ (14) 0 dµ ZMmin with the above values for a, α and f ≃ 0.2 it follows that M ∼ 10−2M . min ⊙ Obvioulsy these results have to be considered as preliminary and as an illus- trationof how one canget useful informationwith themass moment method. Once more data are available it will also be possible to determine other im- portant quantities such as the statistical error in eq. (13) . Nevertheless, the results obtained so far are already of interest and it is clear that in a few years it will be possible to draw more firm conclusions. 5 References [1] B. Paczyn´ski, Astrophys. J. 304 (1986) 1 [2] A. De Ru´jula, Ph. Jetzer and E. Mass´o, Astron. and Astrophys. 254 (1992) 99 [3] E. Aubourg et al., Nature 365 (1993) 623; R. Ansari et al., submitted to Astron. and Astrophys. (1995) [4] C. Alcock et al., Nature 365 (1993) 621; Astrophys. J. 445, (1995) 133; D. Bennett et al., astro-ph 9510104 [5] A. Udalski et al., Acta Astron. 43 (1993) 289; 44 (1994) 165 and 227 [6] C. Alcock et al., astro-ph 9512146 (1994) 426 [7] A. De Ru´jula, Ph. Jetzer and E. Mass´o, Mont. Not. R. Astr. Soc. 250 (1991) 348 [8] K. Griest, Astrophys. J. 366, 412 (1991) [9] Ph. Jetzer and E. Mass´o, Phys. Lett. B 323 (1994) 347 [10] Ph. Jetzer, Astrophys. J. 432 (1994) L43 [11] C. Alcock et al., Phys. Rev. Lett. 74 (1995) 2867; Astrophys. J. in press. Figure Captions 1. ǫ(T) as given by the MACHO collaboration. 2. ǫ (µ) as one gets with eq.(11). 0 6 This figure "fig1-1.png" is available in "png"(cid:10) format from: http://arXiv.org/ps/astro-ph/9601122v1 This figure "fig1-2.png" is available in "png"(cid:10) format from: http://arXiv.org/ps/astro-ph/9601122v1