Collective effects of stellar winds and unidentified gamma-ray sources Diego F. Torres

1. THE MULTI-MESSENGER APPROACH TO UNIDENTIFIED GAMMA-RAY SOURCES Collective effects of stellar winds and unidentified gamma-ray sources Diego F. Torres Institut de Ciencies de l’Espai, Barcelona, Spain http://europa.ieec.fcr.es/english/recerca/map dtorres@ieec.uab.es

2. 2 things discussed in the conference • Observational discovery of TeV sources without lower energy counterparts (not even GeV), e.g., Funk • Possibility of –at least two- explanations (Aharonian) • Fastest diffusion of higher energy hadrons and follow-up interaction in molecular clouds a bit separated from the accelerator • Effect of stellar winds • (single stellar winds Torres et al. 2004, ApJ Letters) • (collective stellar winds, this talk)

3. Unidentified TeV sources: The HEGRA source in Cygnus Aharonian et al. 05 1999 – 2001 Observations 112 hr June - September 2002 Observations 137 hr

4. Making a long story short: no counterpart is identified Butt et al. 05

5. Butt, et al. 2003 Enhancement of stellar density at the HEGRA TeV source Butt et al. 05 110 cataloged OB stars in Cyg OB2 shown as a surface density plot (stars per 4 arcmin2). Note that many stars in Cyg OB2 remain uncataloged – the total number of OB stars alone is expected to be ~2600 (Knodlseder 2002). Although the extinction pattern towards Cyg OB2 may control the observed surface density of OB stars, our analysis assumes that the observed distribution of OB stars tracks the actual distribution. The thick contours show the location probability (successively, 50%, 68%, 95%, and 99%) of the non-variable EGRET source 3EG 2033+4118. The red circle outlines the extent of the TeV source.

6. Unidentified TeV sources: HESS J1303 HESS J1303-631: first serendipitous discovery during monitoring of the PSR B1259. [A&A 439 (2005) 1013] Still unidentified. Extended source. In correspondence the OB association Cen OB1 that contain al least 20 O stars and 1 WR star. Also no known counterpart at lower freq yet. Aharonian et al. 05

7. Hadronic -rays from stellar associations: needed ingredients • High-energy accelerated particles (up to hundreds TeV)  Collective effects of strong stellar winds & SNe at the association core  supperbubble  large scale shock CR acceleration (see, e.g., Montmerle 1979; Casse & Paul 1980; Bykov & Fleishman 1992a,b; Bykov 2001) • Dense enough target medium (for p-p interactions  0  )  Proposed target: collective winds of OB stars near the acceleration region

8. Winds of massive OB stars: basic observational facts • O and B stars lose a significant fraction of their mass in stellar winds with terminal velocities of order 103 km s-1. • With mass loss rates as high as M ~ 510-7 – 110-5 M◦ yr-1 the density at the base of the wind can reach 10-12 g cm-3 (e.g., Lamers & Cassinelli 1999, Ch.2). • Such winds are also permeated by significant magnetic fields, that can reach mG or even higher at tenths of pc from the surface (Surface magnetic field  B  ~ 10-100 G, Donati et al. 2001, 2002) .

9. The gas within a collective wind Wind-wind interaction that takes place in a dense stellar cluster or sub-cluster. The stars are assumed to be uniformly distributed within the outer radius of the cluster. The material ejected from the stars goes through stellar wind shocks (drawn as circles) and then participates in an outward flow [with mean velocity V(R) and density n(R)] which eventually leaves the cluster and interacts with the surrounding interstellar matter.

10. The gas within a collective wind From Domingo-Santamaria & Torres, Hadronic Processes in Winds A&A, 448, 613, 2006 Istropic distribution of stars Each with a wind velocity and mass loss rate Mass and momentum conservation Within the cluster Algebra + initial and boundary conditions Ouside the cluster Numerically solvable set of equations

11. The gas within a collective wind From Domingo-Santamaria & Torres, Hadronic Processes in Winds A&A, 448, 613, 2006 Wind Velocity Wind Density Wind Velocity Wind Density

12. If >1  Convection dominates If<1  Diffusion dominates Penetrating the wind: diffusion vs. convection Approach: comparison between timescales • Diffusion- Convection =1 defines Epmin for protons to penetrate the wind E.g. only protons with energies will enter into the wind whereas lower energy protons will be convected away  wind modulation effect MeV-GeV -ray emission substantially reduced

13. Numerical modeling of gamma-ray production, diffusion, losses Torres, ApJ, 617, 966 (2004)

14. Knock-on emissivity positron emissivity Losses Effect of modulation From Domingo-Santamaria & Torres, Hadronic Processes in Winds A&A, 448, 613, 2006

15. Effect of modulation II From Domingo-Santamaria & Torres, Hadronic Processes in Winds A&A, 448, 613, 2006 Differential Flux Integral Flux Opacity

16. Effect of modulation III X=not detectable √=detectable

17. Concluding remarks • Collective effects of winds, related perhaps with unidentified TeV sources? We have studied collective wind configurations produced by a number of massive stars, and obtained densities and expansion velocities of the stellar wind gas that is to be target for hadronic interactions in several examples.We have computed secondary particle production, electrons and positrons from charged pion decay, electrons from knock-on interactions, and solve the appropriate diffusion-loss equation with ionization, synchrotron, bremsstrahlung, inverse Compton, and expansion losses to obtain expected γ-ray emission from these regions, including in an approximate way the effect of cosmic ray modulation. Examples where different stellar configurations can produce sources for GLAST satellite, and theMAGIC/HESS/VERITAS telescopes in non-uniform ways, i.e., with or without the corresponding counterparts were shown. • Combined observations with GLAST and MAGIC/HESS will tell. • Most models produce no source at GeV energies, even for GLAST • Good candidates: Cygnus OB2 neighbourhood, Westerlund 1, HESS J1303-631