Absorbing aerosols facilitate transition of Indian monsoon breaks to active spells

Abstract

While some long breaks of monsoon intraseasonal oscillations (MISOs) are followed by active spells (BFA), some others are not (BNFA). The circulation during BFA (BNFA) cases helps (prevents) accumulation of absorbing aerosols over central India (CI) resulting in almost three times larger Aerosol Index (AI) over CI, during BFA cases compared to BNFA cases. A seminal role played by the absorbing aerosols in the transition from break to active spells is unraveled through modification of the north-south temperature gradient at lower levels. The meridional gradient of temperature at low level (?T) between aerosol-rich CI and pristine equatorial Indian Ocean is large (>6°C) and sustains for long time (>10 days) during BFA leading to significant moisture convergence to CI. The stability effect arising from surface cooling by the aerosols is overcome by the enhanced moisture convergence creating a moist static unstable atmosphere conducive for the large-scale organized convection over the CI region leading to the resurgence of active spells. The moisture convergence induced by ΔT was also able to overcome possible aerosol indirect effect (Twomey effect) and initiate deep convection and transition to active condition. During BNFA cases, however the maximum ΔT, which was weaker than the BFA cases by more than 1.5°C, could not sustain required moisture convergence and failed to lead to a sustained active spell. Using data from MODIS (MODerate resolution Imaging Spectroradiometer) onboard Terra and several other input parameters from various satellites for the period 2000-2009, the aerosol induced radiative forcing representative of two regions-the CI to the north and the pristine ocean to the south-were estimated and support the differences in observed ΔT during the two cases. Our results highlight the need for proper inclusion of absorbing aerosols in dynamical models for simulation of the observed variability of MISOs and their extended range prediction.