It is proved in this paper that NWP systematic forecast errors in the zonal mean circulation are due to the difference in westerly acceleration process during the forecasting period between model and real atmospheres....It is proved in this paper that NWP systematic forecast errors in the zonal mean circulation are due to the difference in westerly acceleration process during the forecasting period between model and real atmospheres. Those forcing factors which evoke the zonal mean wind variation can be split into various linear terms according to the non-acceleration theorem in a primitive equation system,By applying this tech- nique to the diagnosis of the forecast produets of the T42L9 model in January 1992 and in July 1992, it is indicated that the model has the ability to forecast the zonal mean wind to a reasonable extent, but there are still some errors in several places,especially in the upper troposphere and lower strato- sphere in the mid-latitude region as well as near the surface.The results of analysis by employing this scheme reveal the reason responsible for the systematic forecast errors of the zonal mean wind in the model and the possible way of improving it. It is also shown that non-acceleration theorem can be used as an efficient tool to diagnose the physical processes of NWP models.展开更多
The aim of this paper is to develop and validate a procedure for constructing prediction intervals. These forecasts are produced by Box-Jenkins processes with external deterministic regressors and prediction intervals...The aim of this paper is to develop and validate a procedure for constructing prediction intervals. These forecasts are produced by Box-Jenkins processes with external deterministic regressors and prediction intervals are based on the procedure proposed by Williams-Goodman in 1971. Specifically, the distributions of forecast error at various lead-times are determined using post-sample forecast errors. Fitting a density function to each distribution provides a good alternative to simply observing the errors directly because, if the fitting is satisfactory, the quantiles of the distribution can be estimated and then the interval bounds computed for different time origins. We examine a wide variety of probability densities to search the one that best fit the empirical distributions of forecast errors. The most suitable mathematical form results to be Johnson’s system of density functions. The results obtained with several time series suggest that a Box-Jenkins process combined with the Williams-Goodman procedure based on Johnson’s distributions, provide accurate prediction intervals.展开更多
文摘It is proved in this paper that NWP systematic forecast errors in the zonal mean circulation are due to the difference in westerly acceleration process during the forecasting period between model and real atmospheres. Those forcing factors which evoke the zonal mean wind variation can be split into various linear terms according to the non-acceleration theorem in a primitive equation system,By applying this tech- nique to the diagnosis of the forecast produets of the T42L9 model in January 1992 and in July 1992, it is indicated that the model has the ability to forecast the zonal mean wind to a reasonable extent, but there are still some errors in several places,especially in the upper troposphere and lower strato- sphere in the mid-latitude region as well as near the surface.The results of analysis by employing this scheme reveal the reason responsible for the systematic forecast errors of the zonal mean wind in the model and the possible way of improving it. It is also shown that non-acceleration theorem can be used as an efficient tool to diagnose the physical processes of NWP models.
文摘The aim of this paper is to develop and validate a procedure for constructing prediction intervals. These forecasts are produced by Box-Jenkins processes with external deterministic regressors and prediction intervals are based on the procedure proposed by Williams-Goodman in 1971. Specifically, the distributions of forecast error at various lead-times are determined using post-sample forecast errors. Fitting a density function to each distribution provides a good alternative to simply observing the errors directly because, if the fitting is satisfactory, the quantiles of the distribution can be estimated and then the interval bounds computed for different time origins. We examine a wide variety of probability densities to search the one that best fit the empirical distributions of forecast errors. The most suitable mathematical form results to be Johnson’s system of density functions. The results obtained with several time series suggest that a Box-Jenkins process combined with the Williams-Goodman procedure based on Johnson’s distributions, provide accurate prediction intervals.
