资源环境科技发展态势分析平台

Motivated by the fact that the spatial pattern of the observed precipitation anomalies during 2015/16 winter (a year of strong El Nino) over the west coast of the US and that of the El Nino composite precipitation pattern had considerable differences, the variability in the winter precipitation during strong El Nino events is assessed. The analysis is based on a set of hindcasts (1982-2011) and real-time forecasts (2012-2015) from NCEP Climate Forecast System version 2 (CFSv2), and the following aspects for seasonal mean precipitation variability were examined: (1) the mean signal during strong El Nino based on the composite analysis, and further, the variability from the composite on an event-to-event basis; (2) probability of occurrence for precipitation anomalies to be opposite to the signal (inferred as the composite mean); (3) the probability to have precipitation anomaly in different categories varying from wet to dry; and (4) variations in the characteristics of precipitation from OND, NDJ, to DJF (early to late boreal winter). The results show that the model forecasted seasonal mean precipitation composite for strong El Nino was similar to the linear regression signal with the Nino 3.4 index in observations, with negative anomalies over the Pacific Northwest and positive anomalies over California. However, although in response to an El Nino event, the California precipitation PDF was shifted towards positive values relative to the climatological PDF, the overlap between climatological PDF and the PDF for El Nino events was considerable. This is because of the large variability in seasonal mean outcomes of precipitation from one forecast to another, and therefore, chances to have precipitation anomalies with their sign opposite to the composite El Nino signal remain appreciable. In this paradigm, although the seasonal mean precipitation during 2015/16 winter over the west coast of the US differed from the mean signal for a strong El Nino event, the observed anomalies were well within the envelope of possible outcomes. This has significant implications for seasonal predictability and prediction skill, and further, poses challenges for decision makers in the uptake of seasonal forecast information.