Difficulties to Predict the Short-term Climate over China

Precipitation is an important predictand for short-term climate prediction. It is more difficult to predict precipitation than to predict other variables since precipitation is discontinuity in spatial and temporal distribution, impacted by a lot of factors and bore distinctly regional features. Therefore, the studies of precipitation prediction are necessary and challenging.

In the studies, the NOAA daily outgoing longwave radiation (OLR), the NCEP/NCAR reanalysis daily zonal and meridional wind, the Global Precipitation Climatology Project (GPCP) daily precipitation data, the daily and monthly precipitation observations of 753 stations in China are used. The results show (1) the energy conversion is obviously enhanced in the middle and lower troposphere during the heavy rainfall both in 1997 and in 1999. In other words, the synoptic scale disturbance clearly released its energy to the low-frequency fluctuation during the heavy rainfall. It suggests the kinetic energy conversion be one of mechanisms inducing the heavy rainfall. However, the synoptic scale contribution is very difficult to predict a few months before it occurs. (2) Regression analysis with the OLR indicates the convective variations over Asian monsoon region are more closely associated with the convective activities over the western subtropical Pacific (WSP) than with those over the northern tropical Indian Ocean (NTIO). The obvious correlation with the WSP is noted over India in every year (from 1997 to 2004), but the correlation is not always exhibited over eastern China for the eight years. (3) The EOF analysis of OLR indicates the first mode (EOF1) primarily exhibits seasonal and even longer-term variations. A wavelet analysis reveals intraseasonal variation (ISV) features for OLR EOF2 in 2000, 2001, 2002, and 2004. However, the effective ISV (the significance exceeds the 95% confidence level) does not take place over Asian monsoon regions in every year and it seems to occur only when the centers of an east-west oriented dipole reach enough intensity over the tropical Indian and western Pacific Oceans. (4) The correlation analysis between the time coefficients of EOF and precipitation suggests the precipitation over India is more closely associated with the ISV, seasonal variations, and even longer-term variations than precipitation over eastern China. It suggests the regional precipitation over India is closely associated with seasonal (or even longer-term) variations and the ISV of OLR while the precipitation over eastern China may be easily impacted by some other stochastic or discrete events. Also, the results suggest that achieving accurate forecasts for floods or droughts over eastern China will be more challenging than over India.

The fast ocean-atmosphere model (FOAM) is a fully coupled ocean-atmosphere general circulation model. Using a longer simulation with the FOAM and its atmosphere equivalent, the dynamic ocean could reduce the spring and summer precipitation and postpone the time of summer monsoon onset, but enhance autumn precipitation over the Asian monsoon region. Thus, we speculate that the ocean simulation may affect the prediction about the onset and intensity of Asian monsoon. Furthermore, FOAM can be coupled with the Lund-Potsdam-Jena (LPJ) dynamic global vegetation model (DGVM). The simulation results indicate the dynamic vegetation could change the coupled relationship (pattern) between precipitation and surface temperature, namely, the similar precipitation pattern over Asian monsoon region could couple different surface temperature pattern when the variation of vegetation occurs. Finally, any ensemble forecast also cannot improve the short-term prediction. Some uncertainties always exist no matter where and when heavy rainfalls happen over China.