The Atmospheric Modeling and Weather Prediction (AMF) group focuses on incorporating our improved understanding of the relevant atmospheric processes in numerical models and assimilating non-conventional observations like radars and satellites to improve the initial conditions. In order to address the forecast uncertainty arising from the inability of the physical parameterization suites to capture the physical process realistically and from the inaccurate estimation of the initial state, the group is focusing on developing an optimum configuration of physical and initial condition ensembles. Complimentary efforts using simplified and toy models are also undertaken to understand the individual physical process. Further, this improved understanding of individual physical parameterization will be tested in the comprehensive and complex modelling frame work. This way, the improved understanding of the complex physical process especially the cloud process, radiation and boundary layer process, and representation of their complex interactions in the model will lead to the improvement of the forecast accuracy. With the addition of new observations of the atmosphere and ocean, the group is also focusing on the optimum use of these heterogeneous observations to improve the best estimate of the initial state of the atmosphere and ocean through advanced data assimilation techniques. Observational and theoretical research plays a crucial role in the model development process. To ensure the models incorporate the most up-to-date information from these field observations, our group attempted to bridge the gap between observational studies and model development by ensuring an adequate flow of information among the partners.

The ensemble forecasting research provides a range of value-added products to the public and various stakeholders. Work is also in progress to develop a robust and reliable high-resolution hybrid forecasting system across short to extended-range time scales. Group undertakes a real-time evaluation of model performance to further tune the model to improve its accuracy. The group is developing various model output statistics, bias correction, and calibration techniques to reduce systematic biases and random errors in the model for the real-time operational forecasts. The group is actively involved in generating community-level weather monitoring systems and developing various customized tailor-made forecast products for different stakeholders.

Team