Decade-long surface meteorological measurements from a Japan Meteorological Agency buoy at 29degreesN, 135degreesE are analyzed to elucidate the surface air-sea flux forcing in the western North Pacific Ocean. Besides the well-defined annual cycles, the observed heat and momentum fluxes are dominated by signals related to synoptic-scale weather disturbances. The synoptic-scale heat flux signals have a dominant time scale of 3-14 days, whereas the wind stress signals have a scale of 2-8 days. A comparison between the heat fluxes estimated using the buoy measurements and those from the NCEP reanalysis reveals that the daily NCEP product overestimates both the incoming solar radiation at sea surface and the turbulent heat flux amplitude associated with the individual weather events. The rms amplitude of the synoptic-scale net heat flux of the NCEP product is found to be positively biased by 23%. Despite this amplitude bias, the NCEP product captures the timing and relative strength of the synoptic-scale net heat flux forcing very well. A favorable comparison is also found between the daily surface wind stress forcing from the buoy and that from the NCEP product on the synoptic time scales. Using a bulk surface mixed layer model, we find that the synoptic-scale forcing can significantly change the SSTs in spring-summer seasons. The synoptic-scale heat flux-induced SST anomalies have a typical amplitude of +/-1degreesC, whereas the wind-induced SST anomalies depend on the accumulation of large-amplitude wind events. Excessive accumulation, which occurred, for example, in 1997, can result in unseasonally cold summertime SST anomalies. From both the observations and the model, the frequency spectra for the synoptic-scale SST signals show a clear omega(-2) dependency. While this dependency is consistent with the "white'' surface heat flux forcing in the frequency band of 1/100-1/16 days, short-term mixed layer depth changes induced by the synoptic-scale atmospheric forcing are argued to be important in determining the SST spectral shape in the higher-frequency band.