This research focuses on understanding different atmospheric patterns like blocking, zonal, and transition regimes and analyzing their predictability. We used an idealized land–atmosphere coupled model to simulate Earth's atmosphere. Then we identified these blocking, zonal, and transition regimes using Gaussian mixture clustering and studied their predictability using Lyapunov exponents.