A sensitivity study of decadal climate prediction to aerosol variability using ECHAM6-HAM (GCM)

Abstract

The influence of aerosols on climate is an important but still highly uncertain aspect in climate research. Using the atmospheric general circulation model ECHAM6-HAM the objective of this study is to quantify the aerosol effect over the decadal time scale in comparison to variability induced by varying sea surface temperature (SST) concentration taken from the atmospheric model inter-comparison project (AMIP) data base and the inevitable internal climate noise. This specific modeling setup allows for a quantitative estimation and separation of the stationary aerosol induced variability (AIV), transient SST induced variability (SIV) and the internal variability due to the large scale atmospheric instabilities and non-linearities using the methods of one-way and two-way analysis of variance (ANOVA). The ensemble data of aerosol optical depth (AOD) is analyzed with the help of one-way ANOVA. Furthermore, the two-way ANOVA is implemented on the data of the atmosphere radiative energy balance (AREB), temperature and large scale circulation like velocity potential (VP) and stream function (SF). The correlation coefficients method is used to determine a connection for the AIV and SIV with the 2m temperature.<br /> The four ensemble model runs in this study are: no-aerosol, aerosol, HAM-full and HAM-dir. The model ECHAM6 without the HAM model estimated the no-aerosol and aerosol ensemble, where no-aerosol has no transient aerosol influence and aerosol utilized the aerosol climatological data. The HAM-full and HAM-dir ensemble both are simulated by utilizing the aerosol comparisons between observations and models (AEROCOM), both of them are based different aerosol scheme they differ in the integration techniques used in HAM model. The HAM-full ensemble integrated the direct and indirect scheme, whereas the HAM-dir integrated only the direct effect of aerosols. A comparison is made between ECHAM6 and ECHAM6-HAM ensemble data of AOD. The dust burden of African aerosol particles and the industrial plume of the west China is well captured by the ECHAM6. Using the interactive aerosol scheme in HAM-dir ensemble, an additional anthropogenic aerosol plume towards the north of India is simulated in ECHAM6-HAM. The difference between these model realizations is due to the different aerosol micro-physics processes used in the HAM-model.<br /> The satellite products like surface radiation budget (SRB), global precipitation climatology project (GPCP) are used for the validation of the model data. The model emission data is compared against the ERA-Interim. The model realizations for the global mean of planetary albedo (PA) is in good agreement with the SRB except for the HAM-model ensemble. However, the errors and uncertainties in global mean PA are propagating further into the global mean of TOA radiative energy balance (TREB) and surface radiative energy balance (SREB). Despite of this, the global mean of AREB for all model realizations is estimated with a reasonable accuracy. The imbalances in the global mean of PA, TREB and SREB is further investigated with the global mean of fresh water using GPCP and ERA-Interim data. However, the imbalances in radiative energy balance did not show any link with the latent heat flux but they are associated with the sensible heat flux. The zonal mean of radiative energy balance results explained how the errors or uncertainties are transferred from global to zonal mean scale.<br /> The statistical technique ANOVA is used for the estimation of the local AIV, SIV and climatic noise variability. Reasonable SST signals are captured by all model runs for the AREB, temperature and large scale circulation model data and the results of those signals are justified when analyzed by the correlation coefficients method. Therefore, it is concluded on the basis of these results that the real time forecast is only possible for the SIV and not for the AIV. It is recommended that the decadal climate forecast of SIV over the Pacific is possible by ECHAM6-HAM but with the same experiment setup as it has been done in this study.