The models tend to perform better in cases like this where there's a very robust and coherent MJO pulse in the eastern hemisphere, this large-scale circulation response is consistent w/ what you'd expect as an MJO event encroaches on the Pacific. View attachment 3391
Ok, essentially what this diagram is showing is a time-longitude plot of 200 hpa velocity potential anomalies. Time is on the y axis and longitude (in degrees east) is on the x axis. Hovmoller diagrams like this are useful for tracking wave propagation in real-time, with Rossby waves in the tropics moving from west to east (diagonally from top right to bottom left) on the diagram and phenomena such as Kelvin Waves and MJO events moving from east to west (diagonally from top left to bottom right.) Since you know the distance between each longitudinal point on the earth for a particular latitude, diagrams like these are also useful for monitoring the speed of these waves which can be useful in deciphering whether you're dealing with a faster moving convectively coupled kelvin wave or MJO wave for example. The variable plotted here is known as velocity potential, and its provided the upper troposphere (~200 hPa). Velocity potential essentially measures the irrotational (purely divergent) part of the flow which serves as a proxy for deep cumulonimbus towers in the tropics. If we measured divergence we wouldn't necessarily be measuring the same thing or purely upper level divergence from these areas of deep convection because some of the observed flow response associated w/ upper level divergence receives some contribution from the coriolis effect (even though the coriolis effect is small near the equator & in the tropics).