Relationships Between Area Changes Of West Pacific Subtropical High And SSTA Of Tropical Indian And Pacific Oceans, And Their Impacts On Precipitation Anomalies Over China

Using the NCEP-NCAR reanalysis, HADISST, in situ 160 station data of precipitation, and the circulation index released by National Climate Center of China, we have investigated the variations of WPSH and its association with sea surface temperature anomalies (SSTA) in tropical Indian and Pacific Oceans, interdacadal variations in precipitation during the last 50 years as well. Our results ara summarized as follows:(1) SSTA of both the tropical Indian and Pacific Oceans have significant impacts on area changes of the WPSH. Changes in yearly mean area index of WPSH show that the WPSH responds with a lag of one year to the Indian Ocean SSTA whereas it responds simultaneously to the SSTA of tropical Pacific. In the lower troposphere (850hPa) , the positive(negative) SSTA in central part of the tropical Indian Ocean induces the divergent flows that emanate from the maritime continent northward to west Pacific where an anticyclonic (cyclonic) circulation is generated as a result of Gill type response of the atmosphere. The WPSH is hence enlarged (shrunk) in west Pacific region. On the other hand, the positive (negative) SSTA in the middle tropical Pacific tends to weaken the WPSH because the convergence in the central tropical Pacific induces an anomalous cyclonic (anticyclonic) circulation over the west Pacific. Therefore, the WPSH is expected to be larger (smaller) when the central part of the Indian Ocean is in its warmer (colder) phase while the central equatorial Pacific in its colder (warmer) phase. On the other hand, if a larger (smaller) WPSH is observed in years when the positive (negative) SSTA appear in both the central equatorial Indian Ocean and the central equatorial Pacific, it is expected that this larger (smaller) WPSH is caused by the Indian Ocean other than Pacific. The area of WPSH in boreal summer can be predicted by a regression model based on the SSTA of the central equatorial Indian Ocean in the past spring and the SSTA in central equatorial Pacific in last summer. The predicted area of WPSH in years from 1997 to 2007 using the regression model as built up with the observed data from 1958 to1996 is surprisingly highly correlated (0.91) with the time series of the observed area index of WPSH in the passed decade.(2) In the background of the global warming, there exist apparently the increasing trends in area of WPSH. The area of WPSH abruptly changed in late 1960s and 1990s. These changes possibly induce the precipitation in China to change as a result of the circulation changes. It is amazing that the increasing trends are observed in precipitation respectively in periods of 1951-1970, 1971-1990, and 1990-2007, showing that the region where more precipitation is received migrates northward with time going on although the precipitation tends to be less in northern China during the past 50 years. After each of the two abrupt changes as observed respectively in late 1960s and late 1990s, the WPSH is found to be abnormally bigger, stronger and more westward, while East Asia Summer Monsoon is weakened, leading to more rainfall in regions south of Yangtze River valley.(3) In the years during 1951-2007, both the SON mean SSTA of central equator Pacific Ocean (CEP) and DJF mean SSTA of Indian Ocean (CEI) has significant impacts on area changes of the WPSH in DJF. These impacts from both the oceans last from winter to the next autumn, strengthening the WPSH in middle troposphere. While positive SSTA is in both autumn CEP and winter CEI (called P-P model), there often occur El Nino events. The area of WPSH is therefore bigger than normal, leading to more rainfall in China. On the other hand, the N-N model often associates with La Nina events, induce WPSH to shrink, resulting less rainfall in south China instead. The predicted area of WPSH (DJF) in years from 1998 to 2007 using the regression model as built up with the observational data from 1951 to 1997 is surprisingly highly correlated (0.9) with the time series of the observed area index of WPSH in the passed decade, which is very useful for us to predict the area changes of WPSH.