Removing Activity-Related Radial Velocity Noise to Improve Extrasolar Planet Searches
We have made significant progress towards the proposal goals of understanding the causes and effects of magnetic activity-induced radial velocity (v_r) jitter and developing methods for correcting it. In the process, we have also made some significant discoveries in the fields of planet-induced stellar activity, planet detection methods, M dwarf convection, starspot properties, and magnetic dynamo cycles. We have obtained super high resolution (R approximately 200,000), high S / N (greater than 300) echelle study of joint line bisector and radial velocity variations using the McDonald 2-D coude. A long observing run in October 2002 in particular was quite successful (8 clear nights). We now have close to three years of data, which begins to sample a good fraction of the magnetic cycle timescales for some of our targets (e.g., kappa Ceti; P_cyc = 5.6 yrs). This will be very helpful in unraveling the complex relationships between plage and radial velocity (v-r) changes which we have uncovered. Preliminary analysis (Saar et al. 2003) of the data in hand, reveals correlations between median line bisector displacement and v_r. The correlation appears to be specific the the particular star being considered, probably since it is a function of both spectral type and rotation rate. Further analysis and interpretation will be in the context of evolving plage models and is in progress.