This paper introduces the development background and research contents of database system of weather forecast product of Shanxi Province,and probes into the implementation of system functions and some key technologies.

By reviews on the state of the art in aeronautical meteorological forecasting products,the recent advancements of aeronautical meteorological services in USA,Canada and Europe are summarized.

A weather forecast system of lancang river valley based on time sequence data mining;

Uran Stagnant Floods Weather Forecasting Business Service System;

The specialized weather forecast system in Datong;

In This article the progress of the operational system for meteorological prediction in China were reviewed In particular, the major R&D achievements in NWP operational system are summarized Some key experiences for modernization of meteorological services in China are discussed Finally, the outlook of future meteorological prediction systems in the 21st century are addressed briefly

Applying the theory and methods of machine leaning in meteorological prediction,research of rainfall amount classification is conducted based on Bayesian deducing and learning theory,an algorithm named learn-and-classify--rainfall is presented for predicting rainfall amount with Na?ve Bayesian classifier to improve prediction accuracy.

In thorough research on agricultural weather yield forecast system and BP neural network, a weather forecast for agricultural yield based on BP neural network was put forward to solve the shortage of the estimate-accurately problem.

According to the actual publication of weather products,combining with the universal Web Services objectencapsulation technique,a solution is put forward and achieved,which publish weather products with Web Service technique to support multi-platform and seamless integration.