Identifying the predictable and unpredictable patterns of spring-to-autumn precipitation over eastern China

The patterns of interannual variability that arise from the slow (potentially predictable) and fast or intraseasonal (unpredictable) components of seasonal mean precipitation over eastern China are examined, based on observations from a network of 106 stations for the period 1951-2004. The analysis is done by using a variance decomposition method that allows identification of the sources of the predictability and the prediction uncertainty, from March-April-May (MAM) to September-October-November (SON). The average potential predictability (ratio of slow-to-total variance) of eastern China precipitation is generally moderate, with the highest value of 0.18 in June-July-August (JJA) and lowest value of 0.12 in April-May-June (AMJ). The leading predictable precipitation mode is significantly related to one-season-lead SST anomalies in the area of the Kuroshio Current during AMJ-to-JJA, the Indian-western Pacific SST in July-August-September (JAS), and the eastern tropical Pacific SST in MAM and SON. The prolonged linear trends, which are seen in the principal component time series associated with the second or third predictable precipitation modes in MJJ-to-ASO, also serve as a source of predictability for seasonal precipitation over eastern China. The predictive characteristics of the atmospheric circulation-precipitation relationship indicate that the western Pacific subtropical high plays a key role in eastern China precipitation. In addition, teleconnection patterns that are significantly related to the predictable precipitation component are also identified. The leading/second unpredictable precipitation modes from MAM to SON all show a monopole/dipole structure, which are accompanied by wavy circulation patterns that are related to intraseasonal events.