The relative importance of variables in initial conditions for dynamical weather prediction¹

1 Some of these results were presented at the WMO/rUGG Symposium on Numerical Weather Prediction, Tokyo, 1968, under the title, “The relative importanoe of variables in initial conditions for numerical predictions”, by J. Smgorbky and K. Miyakoda.

Abstract

Tests were made to evaluate the relative importance of large errors in some of the variables formally required to specify initial conditions for the time integration of the hydro-thermodynamic equations for the atmospheric motion. The prime purpose was to analyze the adjustment process immediately following the initial disturbance. The variables selected were the surface pressure, the water vapor, the boundary layer temperature, and the boundary layer wind. It is likely that these quantities are not individually nor possibly even collectively essential to the definition of initial conditions because of dynamic coupling. The experiments showed that the predictions appear almost the same after about six hours irrespective of the initial values of these quantities. But after several days, the subsequent error behaves similar to the growth noted in predictability experiments.

Notes

1 Some of these results were presented at the WMO/rUGG Symposium on Numerical Weather Prediction, Tokyo, 1968, under the title, “The relative importanoe of variables in initial conditions for numerical predictions”, by J. Smgorbky and K. Miyakoda.