A numerical study on forecasting the Henan extraordinarily heavy rainfall event in August 1975

This study is essentially an experiment on the control experiment in the August 1975 catastrophe which was the heaviest rainfall in mainland China with a maximum 24-h rainfall of 1060.3 mm, and it significantly demonstrates that the limited area model can still skillfully give reasonable results even only the conventional data are available. For such a heavy rainfall event, a grid length of 90 km is too large while 45 km seems acceptable. Under these two grid sizes, the cumulus parameterization scheme is evidently superior to the explicit scheme since it restricts instabili-ties such as CISK to limited extent, The high resolution scheme for the boundary treatment does not improve fore-casts significantly.The experiments also revealed some interesting phenomena such as the forecast rainfall being too small while af-fecting synoptic system so deep as compared with observations. Another example is the severe deformation of synoptic systems both in initial conditions and forecast fields in the presence of complicated topography. Besides, the fixed boundary condition utilized in the experiments along with current domain coverage set some limitations to the model performances.