Font Size: a A A

A Comparative Study Of Long-term Memory Characteristics Of Global Mean Surface Air Temperature Under Different Scenarios

Posted on:2021-10-24Degree:MasterType:Thesis
Country:ChinaCandidate:M QiuFull Text:PDF
GTID:2510306725951979Subject:Journal of Atmospheric Sciences
Abstract/Summary:PDF Full Text Request
Long-term memory(LTM)in the climate system has been well recognized and applied in different research fields,but the origins of this property are still not clear.In this work,we contribute to this issue by studying model simulations under different scenarios.The global mean surface air temperatures from the pre-industrial control runs(pi Control),historical(all forcings)simulations,natural forcing only simulations(Historical Nat),greenhouse gas forcing only simulations(Historical GHG),etc.are analyzed using the detrended fluctuation analysis.We find that the LTM already exists in the pi Control simulations,indicating the important roles of internal natural variability in producing the LTM.By comparing the results among different scenarios,the LTM from the pi Control runs is further found to be strengthened by adding natural focings such as the volcanic forcing and the solar forcing.Accordingly,the observed LTM in the climate system is suggested to be mainly controlled by both the internal natural variability and the external natural forcings.The anthropogenic forcings,however,may weaken the LTM.In the projections from RCP2.6 to RCP8.5,a weakening trend of the LTM strength is found.In view of the close relations between the climate memory and the climate predictability,a reduced predictability may be expected in a warming climate.Then,the Fractional integral statistical model(FISM)is used to extract the current dynamical excitations(?)(t)for given time series.By establishing a linear regression relationship between the(?)(t)of the CMIP5 model and the observed radiative forcing data,the physical significance of the(?)(t)is explored.It was found that the(?)(t)of the was closely related to radiative forcing both in single and all focing time series,indicating that there is indeed a close relationship between the(?)(t)and radiative forcing.And through the multiple regression of the single forcing fitting total forcing,it was found volcanic aerosol may have most significant effect on(?)(t).
Keywords/Search Tags:Long-term memory, CMIP5 model, Detrended Fluctuation Analysis, Fractional integral statistical model
PDF Full Text Request
Related items