Glacial lake outburst floods(GLOFs)are serious disasters in glacial areas.At present,due to global warming,glaciers are generally diminishing;thus increasing the glacial lake area and the outburst risk.Therefore,the research on early warning methods for GLOFs is important to disaster prevention and relief in glacial areas.This paper proposes an early warning method using the temperature and rainfall as indices.We selected the daily growth rate of prior period positive cumulative temperature and the prior thirty-day cumulative rainfall as the representative indices of temperature and rainfall.Based on the statistics of the 21outburst events before 2010and the data from the 21nearby meteorological stations,we built an outburst early warning threshold and a flow chart of the GLOF early warning system.The system was verified by the GLOF event in the Ranzeaco glacial lake on 2013-07-05.