The MACHO Project: Microlensing Results from 5.7 Years of LMC Observations

TL;DR

This study analyzes 5.7 years of photometry for 11.9 million LMC stars, yielding 13–17 microlensing events with timescales ${\widehat{t}}$ between 34 and 230 days. Using an improved efficiency analysis and background handling, the authors derive an optical depth ${\tau}_{\text{2}}^{400}$ of about $1.1$–$1.3\times10^{-7}$ and conclude that a 100% MACHO Milky Way halo is ruled out, instead favoring a Galactic halo composed of roughly 20% MACHOs with typical masses around $(0.15$–$0.9)\ M_\odot$, depending on the halo model. The results imply a total MACHO mass within 50 kpc of roughly $9\times10^{10}\,M_\odot$, and demonstrate that the observed microlensing signal is not easily explained by known stellar populations alone, pointing to an extended halo lens population as the most consistent interpretation. The analysis also addresses potential backgrounds from bumpers and supernovae, and emphasizes the role of Milky Way/LMC halo modeling in interpreting microlensing as a probe of dark matter.

Abstract

We report on our search for microlensing towards the Large Magellanic Cloud (LMC). Analysis of 5.7 years of photometry on 11.9 million stars in the LMC reveals 13 - 17 microlensing events. This is significantly more than the $\sim$ 2 to 4 events expected from lensing by known stellar populations. The timescales ($\that$) of the events range from 34 to 230 days. We estimate the microlensing optical depth towards the LMC from events with $2 < \that < 400$ days to be 1.2 ^{+0.4}_ {-0.3} \ten{-7}$, with an additional 20% to 30% of systematic error. The spatial distribution of events is mildly inconsistent with LMC/LMC disk self-lensing, but is consistent with an extended lens distribution such as a Milky Way or LMC halo. Interpreted in the context of a Galactic dark matter halo, consisting partially of compact objects, a maximum likelihood analysis gives a MACHO halo fraction of 20% for a typical halo model with a 95% confidence interval of 8% to 50%. A 100% MACHO halo is ruled out at the 95% C.L. for all except our most extreme halo model. Interpreted as a Galactic halo population, the most likely MACHO mass is between $ 0.15 \msun$ and $ 0.9 \msun$, depending on the halo model, and the total mass in MACHOs out to 50 kpc is found to be 9+4-3 10^{10} msun, independent of the halo model. These results are marginally consistent with our previous results, but are lower by about a factor of two. Besides a larger data set, this work also includes an improved efficiency determination, improved likelihood analysis, and more thorough testing of systematic errors, especially with respect to the treatment of potential backgrounds to microlensing, such as supernovae in galaxies behind the LMC. [Abridged]

Figures (14)

• Figure 1: An R-band image of the LMC, $8.2$ degrees on a side (G. Bothun, private communication), showing the locations of the 30 MACHO fields used here. Also shown are the locations of the 17 microlensing candidates discussed in the text.
• Figure 2: Illustration of the cuts used to select microlensing candidates for criteria B. The x-axis is $\Delta\chi^2/(\chi^2_{\rm ml}/N_{\rm dof})$, where $\Delta\chi^2 \equiv \chi^2_{\rm const} - \chi^2_{\rm ml}$ is the improvement in $\chi^2$ between a constant brightness fit and a microlensing fit. The y-axis is the fitted maximum magnification. The 29 lightcurves are shown as solid dots and are labeled. The remaining symbols are explained in the figure and in detail in § \ref{['sec-events']}. The solid lines show the final cuts for criteria B. The dotted line shows the same for criteria A. See text for details.
• Figure 3: Illustration of the cuts used to select microlensing candidates for criteria B in the color--magnitude diagram. The 29 lightcurves are shown as solid dots and are labeled. The 'bumper' cut is outlined with solid lines and labeled with '$A_{\rm max} > 1.75$' for criteria B. The dotted lines shows the same for criteria A. The symbols are explained in the figure and in detail in § \ref{['sec-bumpers']}. The magnitudes displayed here use the rough global calibrations § \ref{['sec-obs']}.
• Figure 4: The lightcurves for the 29 candidates (25 stars) in § \ref{['sec-events']}. For each object, the upper and lower panels show red and blue passbands. Flux is in linear units with $1\sigma$ estimated errors, normalized to the fitted unlensed brightness. Full lightcurves are shown with 2 day binning, insets of the event regions are unbinned. The full lightcurves can be found on the World Wide Web at http://wwwmacho.anu.edu.au/ The thick line is the fit to unblended microlensing (Table \ref{['tab-blend']}), except for probable supernovae where both the blended fit (solid line) and type Ia fit (dashed line) are shown.
• Figure 5: Microlensing detection efficiency (normalized to $u_{\rm min} < 1$) for the year 5.7 MACHO data, as a function of event timescale ${\widehat{t}}\,$. The solid line shows the 'photometric' efficiency computed for cut A, and the dotted line for cut B as described in § \ref{['sec-eff']}. For comparison the corresponding curves from year-1 (A96) and year-2 (A97) are also shown.