STAR FORMATION IN MASSIVE CLUSTERS VIA THE WILKINSON MICROWAVE ANISOTROPY PROBE AND THE SPITZER GLIMPSE SURVEY

We use the Wilkinson Microwave Anisotropy Probe (WMAP) maximum entropy method foreground emission map combined with previously determined distances to giant H II regions to measure the free-free flux at Earth and the free-free luminosity of the Galaxy. We find a total flux f(nu) = 54, 211 Jy and a flux from 88 sources of f(nu) = 36, 043 Jy. The bulk of the sources are at least marginally resolved, with mean radii similar to 60 pc, electron density n(e) similar to 9 cm(-3), and filling factor phi(HII) approximate to 0.005 (over the Galactic gas disk). The total dust-corrected ionizing photon luminosity is Q = 3.2 x 10(53) +/- 5.1 x 10(52) photons s(-1), in good agreement with previous estimates. We use GLIMPSE and Midcourse Space Experiment (MSX) 8 mu m images to show that the bulk of the free-free luminosity is associated with bubbles having radii r similar to 5-100 pc, with a mean of similar to 20 pc. These bubbles are leaky, so that ionizing photons emitted inside the bubble escape and excite free-free emission beyond the bubble walls, producing WMAP sources that are larger than the 8 mu m bubbles. We suggest that the WMAP sources are the counterparts of the extended low density H II regions described by Mezger. The 18 most luminous WMAP sources emit half the total Galactic ionizing flux. These 18 sources have 4 x 10(51) s(-1) less than or similar to Q less than or similar to 1.6 x 10(52) s(-1), corresponding to 6 x 10(4) M-circle dot less than or similar to M-* less than or similar to 2 x 10(5) M-circle dot; half to two thirds of this will be in the central massive star cluster. We convert the measurement of Q to a Galactic star formation rate (SFR) M-* = 1.3 +/- 0.2 M-circle dot yr(-1), where the errors reflect only the error in free-free luminosity. We point out, however, that our inferred M-* is highly dependent on the exponent Gamma approximate to 1.35 of the high-mass end of the stellar initial mass function. For 1.21 < Gamma < 1.5, we find 0.9M(circle dot) yr(-1) < M-* < 2.2 M-circle dot yr(-1). We also determine a SFR of 0.14 M-circle dot yr(-1) for the Large Magellanic Cloud and 0.015 M-circle dot yr(-1) for the Small Magellanic Cloud.