Response of Indian Summer Monsoon to Middle East Dust in WRF-chem: Sensitivity to Dust Absorbing Properties and Initial Conditions

This study quantifies the sensitivity of the Indian summer monsoon (ISM) system to Middle East dust aerosols (in terms of their absorbing properties) and initial conditions using the Weather Research and Forecasting model coupled with online chemistry (WRF-Chem). For the dust imaginary refractive index (IRI), seven sets of experiments are designed June–August 2008 due to high aerosol loadings, which consist of one set of control experiments without dust emission and six sets of disturbed experiments by altering the dust IRI. Each experiment set includes 16 ensemble simulations created by employing various schemes of model physical and chemical processes to account for the model uncertainties in parameterizing the planetary boundary layer, aerosol shortwave radiation, and aerosol chemical mixing rules. For the initial conditions, six additional sets of experiments are initialized with their starting dates differing by one day. Model results show that the dust-induced ISM rainfall differences (disturbed minus control experiments) vary from -0.4 to 1.0 mm day-1 with dust IRI changing from 0 to 8×10-3 at the wavelength of 600 nm. Further analyses illustrate that the ISM rainfall responses are closely related to the dust-induced temperature change in the troposphere, which is determined by the net effect of dust-induced heating and cooling in the troposphere and at the surface, respectively. With a large dust IRI, the dust heating effect dominates the troposphere temperature, which results in a low-pressure anomaly and a convergence zone over the Iranian Plateau and the Arabian Sea, which in turn transports more water vapor from the Arabian Sea to the Indian subcontinent and causes increased precipitation. Moreover, the simulated single scattering albedo becomes smaller but closer to the surface observations when a larger dust IRI is used (i.e. 8×10-3), while satellite retrievals tend to overestimate the aerosol single scattering albedo in the Middle East compared with both model simulations and surface observations. Model results also show that the uncertainties induced by model initial conditions are smaller than those by dust IRI and model parameterizations. This study highlights the importance for measuring and parameterizing dust IRI.