Optimal error growth of South Asian monsoon forecast associated with the uncertainties in the sea surface temperature

Abstract

A recently developed method of climatic relevant singular vectors (CSVs) is applied to an atmospheric general circulation model (CAM4) to investigate the optimal error growth of the South Asian monsoon (SAM) seasonal prediction due to uncertainties in initial conditions of the sea surface temperature (SST). Emphasis is placed on the investigation of the optimal error growth of SAM seasonal forecast due to the SST uncertainties in the Indian and the equatorial Pacific Oceans. It is found that the uncertainties in the Indian Ocean can result in much larger error growth of SAM seasonal prediction than those in equatorial Pacific Ocean. Most of the CSV patterns over the Indian Ocean resembled a dipole-like structure with opposite signs spanning the northern and southern Indian Ocean. It is seen that the CSVs error growth rate changes significantly depending on the initial states whereas the CSVs patterns are insensitive to the initial conditions. The CSV patterns and error growth rates, calculated using CAM4, are also compared against those using coupled model CCSM4, indicating that the CSVs patterns from CAM4 are similar to those from CCSM4 coupled model while the error growth rate is lower in CAM4 than in CCSM4. Ensemble summer hindcasts, for the period from 2000 to 2009 over Indian Ocean, are constructed by the CSVs. For the purpose of comparison, ensemble forecasts constructed by the time lag ensemble (TLE) method are also conducted. It is found that the ensemble mean prediction by CSVs has a better skill than both the prediction by TLE and by the control run, in particular for predictions longer than 3 months, indicating the merit of CSV for SAM ensemble forecast.