Dynamics of the Indian monsoon onset

The Indian monsoon is a globally significant meteorological event, bringing widespread precipitation annually between May and September. The monsoon first onsets at the beginning of June in the southeastern state of Kerala, then propagates to the northwest, against the mean mid-level wind field. Low-level moist inflow from over the Arabian sea moistens the lower troposphere over southeast India, enabling shallow convection and then supporting deep convection. Recent theory has argued that the spatial distribution of monsoon onset is controlled by the action of convection, which erodes the dry layer at mid-levels from below. This in turn allows the monsoon onset to progress to the northwest. Accurate forecasting of the onset and progression of the monsoon is important for Indian farmers, which constitute a large portion of the population and economy. In par- ticular, the prediction of the spatial pattern, intensity and timing of precipitation is key. It is difficult to represent the physical processes and dynamical interactions associated with the Indian monsoon in numerical weather and climate prediction models, as these processes are imperfectly parameterised. Additionally, the complexity of the system, in- volving a number of balancing processes, is difficult to represent computationally. Current weather and climate prediction models have large biases for monsoon rainfall, and the root causes of these biases are not known. Idealised modelling studies can increase understand- ing regarding the roles of different processes and allow testing of their effects on the Indian monsoon onset. To investigate the propagation mechanisms, an idealised model that reproduces the onset and propagation of the Indian monsoon is developed. It is a two-layer model of moisture dynamics, based on conservation laws, for a vertical plane representing a tran- sect from the India-Pakistan border in the northwest to southeast India). In the model, the balance between low-level moist inflow, mid-level dry advection and the rate of con- vection, controls the onset of the monsoon. For a prescribed low and mid-level wind field, the coupled ordinary differential equations describing the evolution of water vapour con- tent can be studied both analytically and numerically, enabling monsoon onset fronts to be identified and an onset front speed to be calculated. The dependence of these front speeds on the assumed (parameterised) representations of evaporation, precipitation and convection is investigated. It is found that a realistic onset speed can be obtained from a highly idealised setup, for a particular range of convective-mixing timescales. The Weather Research and Forecasting (WRF) model is used to simulate the 2016 sea- vii viii Abstract son, validating performance against reanalysis and observational data. The WRF model is examined in the framework of the two-layer idealised model, focusing on the evolution of moisture content over lower and upper atmospheric layers, increase in low-level moisture flux at onset and the decrease in the mid-level northwesterly wind. The parallels between the WRF model and idealised model lend support to the theory of monsoon onset. A moisture budget analysis is also conducted for the WRF model, enabling a vertical con- vective flux to be a diagnosed and through its correlation with total column moisture, a convective timescale is derived. In the idealised model, a range of 0.5–7 days is initially assumed, which is verified by the WRF model results of 1–2 days. The methodology used to derive a convective timescale in the WRF model can be applied to other models, build- ing a more complete picture of the range of possible convective timescales associated with the Indian monsoon.