Figure. Directed flow (v1) versus pseudorapidity for Au + Au collisions at intermediate centrality. Also shown are data from the NA49 experiment at lower energy.

If we focus on a subset of particles emerging from a collision with about the same beam-axis velocity (or rapidity), then directed flow represents a preferential emission towards a single direction in the transverse plane (aligned with the plane of the colliding nuclei), whereas the much larger elliptic flow represents a preferential emission in two directions 180 degrees apart in the transverse plane. Models predict that directed flow at RHIC should be almost zero over a range of rapidities centered on midrapidity. Nonetheless, whatever small departures from flat behavior might be observed over this range can reveal important clues regarding stopping of the incoming matter and space-momentum correlations, and also have implications in the search for Quark Gluon Plasma.

This is the first measurement of directed flow (v₁) at RHIC. We observe the predicted flat region centered at zero on the horizontal axis (midrapidity), whereas at rapidities further from the center, the signal rises relatively steeply. We note that v₁ at rapidities within a couple of units from beam rapidity appears unchanged as a function of y - y_beam if the beam energy is lowered by a factor of more than 10 — hence the motivation for mapping NA49 (CERN) data from the green points (observed rapidities) to the blue points (shifted left and right by the difference in beam rapidity between RHIC and CERN). The 2004 RHIC run has accumulated much improved statistics, and forthcoming analysis of these data is expected to address the remaining interesting physics issues associated with the nearly flat v₁ region centered on midrapidity.