Short term numerical forecasting of a shallow storm complex using bi-static and single doppler radar data
Montmerle, T. ; Caya, A. ; I. Zawadski
Année de publication
2002
A new method based on four-dimensional variational radar data <br>assimilation into a cloud-resolving model has been developed for <br>nowcasting purposes. This method allows for the retrieval of the model <br>prognostic variables that compose the initial state of the simulation. <br>The echo-free regions are filled by a 3D wind analysis from <br>single-Doppler data based on linearity of the horizontal wind components<br> in a moving reference frame, which provides a realistic mesoscale flow <br>that is in better agreement with the air circulation retrieved from <br>dual-Doppler observations within the precipitating regions. Furthermore,<br> the near-ground refractivity index of air derived from ground targets <br>is used to diagnose a high-resolution and two-dimensional distribution <br>of relative humidity in the mixed layer. Two experiments are performed: <br>one uses multiple-Doppler information coming from McGill University's <br>bistatic radar network and the second considers only single-Doppler <br>observations. This updated algorithm has been applied to a shallow <br>hailstorm and shows very encouraging skill in predicting the short-term <br>evolution of this convective system. The time evolution of the storm is <br>captured well, and a significant improvement is noticed when compared <br>with the nowcasting method based on Lagrangian persistence. When <br>compared with the results obtained with the bistatic network, results <br>when a single-Doppler radar is used show weaker capability to forecast <br>the radial velocity than the precipitation pattern but still give a <br>better forecast than the Lagrangian persistence method does.</div>
Accès à la notice sur le site du portail documentaire de Météo-France