The Application Of Markov Chain Model In The Research Of Persistent Feature Of Daily Precipitation

Markov chain model is often applied in the weather forecast and weather changes in meteorology. The essay applies the Markov chain model to research the persistent feature of daily precipitation, which uses the daily rainfall data of50years(1960-2009) of160stations in the entire country. It studies the persistence and transformation of the status evolution of weather and climate through the calculation and analysis of statistical characteristic values. The main results are:I Research into continuous precipitation based on Markov transition probability limit distributionThe transition probability limit distribution of weather condition in different climate condition actually represents persistence and transition of the climate, at the same time indicates its predictable period. The research reflects inherent difference of persistence and transition of daily weather condition caused by difference affecting weather system in different areas. The research significance is that it can be used as a theoretical basis for zone-classification of climate.The results of statistical analysis show that the national average duration of summer is shortest and increases from north to south, from west to east. The average rainfall duration in spring lasts5.1days,5days in summer,6.5days in autumn and6.2days in winter. It means that the rainfall duration in summer half year is shorter than in winter half year. The reason might be that the weather system is more complicated in spring and summer and medium and small scale weather system is more.It proves once again from another side that the evolution process of the daily precipitation weather conditions in various regions displays the natural transition sustainability of certain state of weather and climate, namely the climate condition of natural weather cycle, which provides the climate background for the short-term weather forecast.Ⅱ The weather and climate state evolution and continuous feature based on the Markov chain modelThe daily precipitation records around the country were made up of alternate precipitation days (wet day) and non-precipitation days (dry day),which was called wet and dry-length sequence. In the sequence, the length of dry day or wet day and the transition was crucial to weather forecast or long-term forecast. The essay, which was based on the concept of run-length and transition, studied the statistical characteristics of dry (wet) run-length through the Markov chain run-length and transition theory. The continuous feature of weather and climate state evolution was concluded by calculation of the statistics, such as the average length and variance of the dry (wet) day runs, the transition frequency of weather state, the average length and return time of weather and climate cycles. The result reflected from another point that the evolution had a persistence of weather natural transition, which is climate condition of natural weather cycle, which provided the climate background for short-term weather forecast.