Abstract Accurate precipitation nowcasting is one of the most challenging tasks in atmospheric sciences. The current methods of nowcasting primarily rely on inferring precipitation from radar reflectivity, which inevitably leads to uncertainties in forecasts due to the limitations of single radar data in capturing the detailed initial conditions of complex weather systems. Global Navigation Satellite Systems (GNSS) can provide accurate water vapor information of high temporal resolution. In this study, a generative network (GRENet) is designed to integrate GNSS water vapor information with radar observations to improve precipitation nowcasting. A case study on a heavy rainfall event demonstrates that GRENet can predict the range and location of the precipitation center more accurately than a baseline model employing only radar observations. This results in improved performance on critical success index and fractions skill score, indicating that detailed initial water vapor from GNSS contributes significantly to enhancing precipitation nowcasting skill.
Lu et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: