资源环境科技发展态势分析平台

The spatial structure and propagation of tropical intraseasonal convection anomalies diagnosed with the outgoing longwave radiation-based Madden-Julian Oscillation index are examined in the boreal summer and winter seasons. It is shown that the outgoing longwave radiation-based Madden-Julian Oscillation index represents both northward and eastward propagation in summer and eastward propagation in winter in a manner consistent with Madden-Julian Oscillation and Boreal Summer Intraseasonal Oscillation propagation as diagnosed in many previous studies. The outgoing longwave radiation-based Madden-Julian Oscillation index and three other widely used indices for tracking the Boreal Summer Intraseasonal Oscillation are then compared in their lag correlation structure over selected reference areas, cross-correlation coefficients of the two principal component time series, and time-dependent phase angle composites. The outgoing longwave radiation anomalies from these different indices propagate differently according to these diagnostics. One of them exhibits little propagation at all, even though one would expect good propagation based on composites of the phases separately. This illustrates the general point that while composites of individual phases, presented in a sequence, are generally taken to imply that the phases tend to occur in that sequence in time, that need not be the case. It is suggested that propagation characteristics are relevant to the application of the indices and that smoother propagation is desirable.

Plain Language Summary The tropical intraseasonal oscillation is a large-scale pattern of cloud and circulation in the equatorial regions recurring every 30 to 60 days. The pattern propagates both northward and eastward in the northern summer season, and brings a large amount of rain and disturbed weather along its path. To track its movement, scientists have developed simple indices using winds and the energy radiating from the Earth as seen from satellites. Here we examine how well these indices accurately track the oscillation by looking into a large number of historical events, and show that our diagnostics may further help improve these indices.