A DETAILED STUDY OF SPITZER-IRAC EMISSION IN HERBIG-HARO OBJECTS. I. MORPHOLOGY AND FLUX RATIOS OF SHOCKED EMISSION

We present a detailed analysis of Spitzer-IRAC images obtained toward six Herbig-Haro objects (HH 54/211/212, L 1157/1448, and BHR 71). Our analysis includes (1) comparisons of morphology between the four IRAC bands (3.6, 4.5, 5.8, and 8.0 mu m) and H-2 1-0 S(1) at 2.12 mu m for three out of six objects, (2) measurements of spectral energy distributions (SEDs) at selected positions, and (3) comparisons of these results with calculations of thermal H-2 emission at LTE (207 lines in four bands) and non-LIE (32-45 lines, depending on the particle for collisions). We show that the morphologies observed at 3.6 and 4.5 mu m are similar to each other and to H-2 1-0 S(1). This is well explained by thermal H-2 emission at non-LTE if the dissociation rate is significantly larger than 0.002-0.02, allowing thermal collisions to be dominated by atomic hydrogen. In contrast, the 5.8 and 8.0 mu m emission shows different morphologies from the others in some regions. This emission appears to be more enhanced at the wakes in bow shocks, or less enhanced in patchy structures in the jet. These tendencies are explained by the fact that thermal H-2 emission in the 5.8 and 8.0 mu m band is enhanced in regions at lower densities and temperatures. Throughout, the observed similarities and differences in morphology between four bands and 1-0 S(1) are well explained by thermal H-2 emission. The observed SEDs are categorized into type-A, those in which the flux monotonically increases with wavelength, and type-B, those with excess emission at 4.5 mu m. The type-A SEDs are explained by thermal H-2 emission, in particular with simple shock models with a power-law cooling function (Lambda alpha T-s). Our calculations suggest that the type-B SEDs require extra contaminating emission in the 4.5 mu m band. The CO vibrational emission is the most promising candidate, and the other contaminants discussed to date (HI, [Fe II], fluorescent H-2, and polycyclic aromatic hydrocarbon) are not likely to explain the observed SEDs.