The Role of Data Assimilation in Enhancing Warm Fog Predictions Over Delhi and NCR

Fog forecasting over highly urbanized and fog-prone regions like Delhi and the National Capital Region (NCR) in the Indo-Gangetic Plain (IGP) is challenging due to the complex interplay between land surface processes and atmospheric conditions. This research investigates the role of advanced data assimilation techniques in enhancing fog forecasting accuracy using high-resolution numerical weather prediction (NWP) models. The focus is on improving the prediction of fog lifecycle events, including onset, duration, and dissipation, by integrating land surface data and non-conventional atmospheric observations into NWP systems.

The study begins with the implementation of the High-Resolution Land Data Assimilation System (HRLDAS) to improve the initialization of critical land surface variables, such as soil moisture and soil temperature. These variables play a key role in surface energy exchanges and boundary layer dynamics. Sensitivity experiments show that incorporating fine-gridded land surface data into the Weather Research and Forecasting (WRF) model significantly improves near-surface temperature, humidity, and wind forecasts. This results in a substantial reduction in soil moisture bias and a more accurate representation of land-atmosphere interactions, leading to enhanced predictions of fog onset and dissipation in the NCR region. Building on this, the research employs cyclic assimilation of microwave radiometer (MWR) observations to improve the vertical profiles of atmospheric temperature and humidity. The combined assimilation of MWR profiles and HRLDAS-generated land surface fields enhances the boundary layer structure, which is critical for fog formation and dissipation. Results demonstrate that the assimilation of non-conventional data sources improves the spatial and temporal accuracy of fog forecasts, reducing errors in predicting fog intensity and duration. These findings emphasize the importance of high-resolution observational data and advanced assimilation techniques for capturing the microphysical and thermodynamic processes governing fog development.

This study demonstrates the importance of integrating land data assimilation and high-resolution atmospheric observations into NWP systems to enhance fog forecasting. The techniques developed address challenges posed by complex land-atmosphere interactions and can be applied to other fog-prone regions. By improving the understanding of fog dynamics, this work enables more accurate and timely forecasts, benefiting critical sectors like aviation, transportation, and public safety in regions such as the IGP.