Precipitation nowcasting based on multifractal advection and deformation

The X-band radars deliver precipitation estimates with high accuracy and space resolution (up to 100 m in space and 1 min in time). With their increasing deployment around large cities, there is an appealing need for short-term nowcasting of rainfall at high resolutions for urban applications.

Nowcasting means forecasting with lead times of up to six hours. Classical precipitation nowcasting methods include methods of image processing to identify precipitation cells and extrapolate their motion. Due to the strong nonlinearity of the precipitation processes, such methods face a number of limitations, e.g., cell identification lacks physics and can be quite ad-hoc or even fail because of their fast deformation. On the contrary, this presentation aims to demonstrate how the fast multiscale deformation of the rainfall cells could be used to improve precipitation nowcasting, with the help of new radar data and products.

Spectral and multifractal analyses of radar data enable a comparison of the structure and the morphology of both the precipitation and vector fields through space time scales. This provides a unique framework to nowcast both fields over scales relevant to urban decision-making. Overall, this presentation contributes to the development of new, reliable, operational tools to use in their full extent the high-resolution X-band data.