Study on the Reconstruction of Background and Lateral Boundary Conditions for a Convective-Allowing Ensemble Forecast System

Abstract

This study demonstrates a reconstruction scheme to improve the quality of the background conditions (BCs) and lateral boundary conditions (LBCs) of the convective-allowing ensemble forecast system (CAEFS). We first analyzed the impact of BC and LBC on the CAEFS and revealed that the perturbation of BCs/ LBCs in the CAEFS mainly affects the large-scale uncertainty of short-range/ long-range forecasts, subsequently reconstruction methods on BC and LBC of the CAEFS were developed, the scheme utilized a high-quality global deterministic forecast to refine the BCs and LBCs, which were initially derived from downscaling of the coarse-resolution global ensemble forecast system (GEFS). Considering the change in the uncertainty characteristics of the BCs and LBCs after reconstruction, a perturbation amplitude adjustment scheme was applied to the reconstructed BCs and LBCs. Results reveal the following information: 1) Limited by the quality of GEFS, the downscaling scheme exhibits significant forecast errors and overspread. 2) The reconstruction scheme can improve the forecast accuracy of CAEFS. However, the problem of prediction divergence still exists due to the large amplitude of large-scale perturbation. 3) The reconstruction combined with perturbation amplitude adjustment scheme can not only improves the quality of control prediction but also effectively alleviates the problem of overspread of the CAEFS, making the system more effective in focusing on the uncertainty of small-scale forecasting and improving the reliability of convective-scale ensemble forecasting.

Keywords: convective-allowing ensemble prediction, perturbation adjustment, background conditions and lateral boundary conditions