Herbig-Haro Jet Movies from HST
These movies compliment the paper "Fluid Dynamics of Stellar Jets in Real Time: Third
Epoch HST Images of HH 1, HH 34, and HH 47", 2011, ApJ 736, 29. Please reference
this paper if you use the movies for any purpose. The movies are
constructed from three epochs, where proper motions are measured for each feature,
and the features are propagated forward or backward in time and averaged to construct
an image at an intermediate time. This procedure works remarkably well and is largely
independent of whatever initial proper motions are input. Some inaccuracies may be present
right near the sources, when the knots first emerge. At present, the HH 111 movies only
span two epochs.
Colors are red = [S II] (6716+6731A lines), and green = H-alpha. Yellow
shows places where emission occurs in both filter, and continuum sources are yellow for
this reason. In most cases, H-alpha shows where the shock fronts are located and the
[S II] emission follows in a cooling zone.
Note that HST may release its own versions of these, but whatever HST
releases will derive from the movies shown here, possibly with a different color
scheme. The figure captions in the
original paper (3.3Mb)
explain what is going on, and you can also access large-scale images of each object there.
Please refer to the paper, which outlines the science goals in the Introduction,
and summarizes the main categories of results we observe in the Discussion and
Conclusion sections, for more information.
ESA/HST have put out a very nice
public video of the research.
HH 1 and HH 2
HH 1; Slow movie (Color, 0.5Mb)
HH 1; Fast movie (Color, 0.5Mb)
HH 1 Jet; Slow movie (Color, 0.2Mb)
HH 1 Jet; Fast movie (Color, 0.2Mb)
HH 2; Slow movie (Color, 0.4Mb)
HH 2; Fast movie (Color, 0.4Mb)
HH 34 Jet (Color, 0.3Mb)
HH 34 Jet; Slow movie(Color and B&W, 0.7Mb)
HH 34 Jet; Fast movie(Color and B&W, 0.7Mb)
HH 34 Jet; Zoomed version; see note above on interpolation (B&W, 1.9Mb)
HH 34 Bow; Slow movie (Color, 3.8Mb)
HH 34 Bow; Fast movie (Color, 3.8Mb)
HH 34 Bow; H-alpha (B&W, 1.8Mb)
HH 47 Large-Scale (Color, 5.3Mb)
HH 47 Jet; 3-epochs, then interpolated (Color, 7.5Mb)
HH 47 Bow (Color, 9.1Mb)
HH 111 Jet (Color, 0.9Mb)
HH 111 Knot V (Color, 0.4Mb)
Movies of unsteady magnetic jets
from Hartigan et al., 2007, ApJ 661, 910-918.
These movies trace conditions along the axis of the jet in the 2.5D
simulation. The average velocity is 200 km/s, and the
simulation alters that randomly within the 100-300 km/s range to generate pulses.
The starting Alfven speed is 35 km/s, so perturbations less than this do not initially
In a typical MHD disk wind,
B ~ n^0.5, n ~ r^-2, so B ~ r^-1, and the Alfven speed V_A ~ B/n^0.5 ~ constant with r.
This is a problem for stellar jets because we know from proper motion studies and
from line excitation measurements that velocity perturbations of
~ 30 km/s produce shocks along the jet beam at distances of 1000+ AU,
so V_A must be even smaller there, but then this is a weak field that
won't be capable of driving a 300 km/s jet unless V_A declines with distance.
Hence, we need a way to remove field,
perhaps by ambipolar diffusion or reconnection close to the source. Our simulation
assumes this has already occurred, so the initial Alfven speed is 35 km/s.
The main goal of the paper is to explore how a pulsed flow affects the dynamics.
In the postshocked regions of a pulsed flow, B ~ n, which bunches
field into dense clumps and lowers it in the preshock gas enough that the Alven
speed declines and shocks become possible with small amplitude perturbations.
The field within the postshock regions becomes strong enough that these clumps
of different velocites occasionally `bounce' off each other.
This movie shows how the initial velocity perturbations steepen to form shocks, and in
each working surface one can then follow the development of the forward and reverse shocks.
Occasional strong rarefactions decouple various parts of the flow from one-another
as faster portions run away from slower ones.
Same as the velocity movie, but with a spatial scale where it is easier to see how
the perturbations are produced in the simulation, and one can also see the forward
and reverse shocks in each working surface form a little easier.
The density movies show sharp peaks where the shocks form, and then the
knots gradually expand and become less dense as they propagate down the beam.
These peaks are superposed upon the declining density typical for a constant opening
Same as above but with a log scale for the density. Observe how the lowest valleys
(rarefactions) propagate outward. These represent free expansion, as the rarefactions
effectively decouple the portions of the flow ahead of it from those behind it.
Same as above but with logs on both axes. The solid line shows free expansion, n ~ r^-2.
B vs. n:
A very interesting, albeit complex-looking movie. Here we are plotting B vs. n. The large
+ denotes where the simulation begins (beginning magnetization). The slope of the blue
line shows B ~ n, which happens whenever material goes into a shock or rarefaction wave.
The green line should be followed for a typical steady disk wind
(B ~ n^0.5, or V_alfven = constant). Looking at the ensemble
of points at the end of the simulation, the overall slope is intermediate between these
two limits. Hence, a pulsed flow gathers field into the dense parts, and therefore
reduces the Alfven speed between clumps, allowing for shocks to occur there which would
simply be magnetic waves if it were not for the cooling in the shocked gas.
The main point here is that even if preshock fields are weak and Alfven velocities are low
between knots at large distances from the source, they will both be much higher closer to
the source, implying the existence of a magnetic-dominated zone close to the star where
knots are ejected.