We present the first results from the science demonstration phase for the Hi-GAL survey, the Herschel key program that will map the inner Galactic plane of the Milky Way in 5 bands. We outline our data reduction strategy and present some science highlights on the two observed 2 • × 2 • tiles approximately centered at l = 30 • and l = 59 • . The two regions are extremely rich in intense and highly structured extended emission which shows a widespread organization in filaments. Source SEDs can be built for hundreds of objects in the two fields, and physical parameters can be extracted, for a good fraction of them where the distance could be estimated. The compact sources (which we will call cores' in the following) are found for the most part to be associated with the filaments, and the relationship to the local beam-averaged column density of the filament itself shows that a core seems to appear when a threshold around A V ∼ 1 is exceeded for the regions in the l = 59 • field; a A V value between 5 and 10 is found for the l = 30 • field, likely due to the relatively higher distances of the sources. This outlines an exciting scenario where diffuse clouds first collapse into filaments, which later fragment to cores where the column density has reached a critical level. In spite of core L/M ratios being well in excess of a few for many sources, we find core surface densities between 0.03 and 0.5 g cm −2 . Our results are in good agreement with recent MHD numerical simulations of filaments forming from large-scale converging flows.

ABSTRACT. Hi-GAL, the Herschel infrared Galactic Plane Survey, is an Open Time Key Project of the Herschel Space Observatory. It will make an unbiased photometric survey of the inner Galactic plane by mapping a 2°wide strip in the longitude range |l| < 60°in five wavebands between 70 μm and 500 μm. The aim of Hi-GAL is to detect the earliest phases of the formation of molecular clouds and high-mass stars and to use the optimum combination of Herschel wavelength coverage, sensitivity, mapping strategy, and speed to deliver a homogeneous census of starforming regions and cold structures in the interstellar medium. The resulting representative samples will yield the variation of source temperature, luminosity, mass and age in a wide range of Galactic environments at all scales from massive YSOs in protoclusters to entire spiral arms, providing an evolutionary sequence for the formation of intermediate and high-mass stars. This information is essential to the formulation of a predictive global model of the role of environment and feedback in regulating the star-formation process. Such a model is vital to understanding star formation on galactic scales and in the early universe. Hi-GAL will also provide a science legacy for decades to come with incalculable potential for systematic and serendipitous science in a wide range of astronomical fields, enabling the optimum use of future major facilities such as JWST and ALMA.

We present a first study of the star-forming compact dust condensations revealed by Herschel in the two 2 • × 2 • Galactic Plane fields centered at, respectively, and observed during the science demonstration phase for the Herschel Infrared GALactic plane survey (Hi-GAL) key-project. Compact source catalogs extracted for the two fields in the five Hi-GAL bands (70, 160, 250, 350 and 500 μm) were merged based on simple criteria of positional association and spectral energy distribution (SED) consistency into a final catalog which contains only coherent SEDs with counterparts in at least three adjacent Herschel bands. These final source lists contain 528 entries for the = 30 • field, and 444 entries for the = 59 • field. The SED coverage has been augmented with ancillary data at 24 μm and 1.1 mm. SED modeling for the subset of 318 and 101 sources (in the two fields, respectively) for which the distance is known was carried out using both a structured star/disk/envelope radiative transfer model and a simple isothermal grey-body. Global parameters like mass, luminosity, temperature and dust properties have been estimated. The L bol /M env ratio spans four orders of magnitudes from values compatible with the pre-protostellar phase to embedded massive zeroage main sequence stars. Sources in the = 59 • field have on average lower L/M, possibly outlining an overall earlier evolutionary stage with respect to the sources in the = 30 • field. Many of these cores are actively forming high-mass stars, although the estimated core surface densities appear to be an order of magnitude below the 1 g cm −2 critical threshold for high-mass star formation.