Evaluation of Quasi-biweekly Oscillation Prediction in the Asian Summer Monsoon Regions by BCC S2S Model

The quasi-biweekly oscillation (QBWO) plays an important role in global weather and climate change. It's a very important source of sub-seasonal to seasonal (S2S) predictability. Using the sub-seasonal to seasonal forecast model of Beijing Climate Center (BCC S2S), the boreal summer QBWO is simulated, the forecast skill is discussed, and the model bias is analyzed. QBWO can be obtained from the third and fourth modes of multivariate empirical orthogonal function (MV-EOF) analysis on daily anomalies of outgoing longwave radiation (OLR) and zonal wind at 850 hPa (U850) in the Asian monsoon region. According to reanalysis data, QBWO shows a northeast-southwest-tilted convection-circulation structure, propagating north/northwestward from the equatorial western Pacific and Indian Ocean. The forecast skill of BCC S2S on QBWO decreases as the forecast lead time increases, and biases become very significant in the period, propagation characteristics and strength when the lead time comes to 9 days. BCC S2S reveals a higher forecast skill of QBWO structure and propagation over the western North Pacific, while it significantly underestimated convection signal of QBWO over the tropical Indian Ocean. The convection-circulation wave structure of QBWO in 9-day lead time prediction over the Indian Ocean is loose and appears over the Arabian Sea (instead of over the tropical eastern Indian Ocean and Bay of Bengal where the reanalyzed QBWO is active). It suggests that the unrealistic Indian Ocean QBWO is related to biases of model mean state. The simulated low-level moisture and convection during boreal summer are enhanced over the western Pacific and the Arabian Sea. However, the model underestimates the abundant moisture and vigorous convection over the eastern Indian Ocean and Bay of Bengal. BCC S2S captures the structure and propagation of QBWO over the western North Pacific, but slightly underestimates the strength of QBWO wave train. This underestimation of QBWO convection might be attributable to the relatively weaker vorticity to the northwest of QBWO convection. By diagnosing the vorticity equation, it's found that although the model well simulates positive contributions of geostrophic vorticity advection and convergence effects to the northwest of convection, these contributions are still underestimated. Moreover, the simulated relative vorticity advection shows an opposite effect to reanalysis data in the 9-day lead time prediction, weakening the favorable environment of QBWO development associated with positive vorticity to the northwestern part of convection.