Probing dynamics of dark energy with supernova, galaxy clustering and the three-year Wilkinson Microwave Anisotropy Probe (WMAP) observations

Using the Markov chain Monte Carlo (MCMC) method, we perform a global analysis constraining the dynamics of dark energy in light of the supernova (Riess "gold" samples), galaxy clustering (SDSS 3D power spectra and SDSS Lyman-alpha forest) and the latest three-year Wilkinson Microwave Anisotropy Probe (WMAP) observations. We have allowed the dark energy equation of state to get across -1 and pay particular attention to the effects when incorrectly neglecting dark energy perturbations. We find the parameter space of dynamical dark energy is now well-constrained with w(z = 0.3) = -1.046(-0.086-0.196)(+0.096+0.226) (1, 2 sigma). Neglecting dark energy perturbations will make the parameter space of dark energy non-trivially smaller. The dynamical dark energy model where the equation of state crosses -1 is mildly favored and the standard Lambda CDM model is still a good fit to the current data.