Information Characteristics Of Meteorological Elements And Its Application To Precipitation Forecast In Rainy Season

Less work has been done to analyze the information characteristics of the meteorological element fields systematically, and there are some initial and boundary information in the historical data which is lacking in the climate or weather dynamical model. So the information entropy and the information transport principle from information theory are employed to analyze the static and dynamical information characteristics of height and temperature fields in different time scales. Combination other methods, such as nonlinear prediction method, complex networks, dynamical-statistic objective forecast method, the predict accuracy of the Northern Hemisphere height field is studied, the air-sea bilayer correlation pattern is extracted, and the pattern is quantified to apply to the model error correction of precipitation forecast in the rainy season. A better method and the appropriate samples to calculate the entropy in meteorological element field are carried out. The detail characteristics of the static and dynamic information in the meteorological field are described. The nonlinear spatial structure of the air-sea interactions is illustrated. It is helpful to improve the precipitation forecast in the rainy season. The main results and conclusion are summarized as following:(1) The validility of the information theoryThe sample estimation method is more suitable to analyze the meteorological time series. Its advantage is that it is less sensitive to the number of samples. To calculate the time information entropy, about100samples are sufficient to fit the accuracy in the daily and monthly time scales.(2) The static characteristics of the time information entropy in the temperature and height fieldsIt is significant that the spatial heterogeneity of the time information entropy exists in the both temperature and precipitation field. The information entropy is increasing with the increase of the latitude for the two fields. In vertical, they are quasi-symmetry, centering in the300hPa level. The temperature information entropy is increasing with the increase of the altitudes when the height is above that of300hPa, while it is decreasing with the increase of the altitudes when the height is under that of300hPa. The relationship between the information entropy of the height field and the altitude is just contrary to that of the temperature field. Both the seasonal variation and inter-decadal climate variation of the two field entropy are distinct. In the temperature field, the seasonal variation is the most distinct in the high latitude regions, then the mid latitude regions, and the least distinct in the low latitude regions. The seasonal variation of the Northern Hemisphere is more distinct than that of Southern Hemisphere. The entropy of summer is the smallest, that of winter is the biggest, and that of spring and autumn are in the middle. The period of the height field entropy is about half a year. The inter-decadal climate variation of the two fields' entropy is obvious as well, and the periods of the entropy are almost the same in different heights.(3) The predict accuracy and meridional information transmission of the Northern Hemisphere geopotential heightFor the predict accuracy of the height field, it is high over the land and is low over the oceans for the important impact of the annual signal. The predict accuracy increases with the altitude increasing. Calculation indicates that the predict accuracy of interdecadal signal is higher than that of intraseasonal.As to the meridional information transmission, the information loss rate is the smallest over the Pacific Ocean and the Atlantic Ocean, and the corresponding subtropical region is called meridional "information transmission channels". The information loss rate of the intraseasonal signal is larger than that of the interdecadal signal, and the transmission direction of the two signals are contrary in horizontal and vertical direction. The four seasons have important impact on the meridional information transmission especially on the "information transmission channels", and the spring and the autumn have the most important influence.(4) The spatial information of air-sea bilayer correlation patternSpatial and temporal information entropy in the meteorological fields and their transportive characteristics were studied, we put forward using mutual information to present the nonlinear correlation between pair grid points of ocean and air, extracting and quantifying the bilayer correlation pattern of the air-sea systems with complex network method. Further the air-sea main spatial mode is identified. It composes of three key regions, and the main interactions occur between the atmospheres over the tropical islands and the sea in the Indian Ocean, and between the atmospheres over tropical islands and the sea in the Pacific Ocean. The formation of the main pattern can be explained by the "gearing between the Indian and Pacific Ocean (GIP)" model, therefore the pattern existence can be confirmed reliably. The trigger processes of the three key regions are revealed as well. It is that the anomalies of the tropical Mid-eastern Pacific Ocean SST (TMEPO-SST) appearance first, then the850hPa geopotential height over the Pacific Islands and the tropical Indian Ocean SST (TIO-SST) response that anomalies two and three months later respectively, through the Walker Circulation and the Anti-Walker Circulation. The changes of the three key regions are less consisting in the1960s, more consisting in1970s and1980s, and is complicated in1990s and2000s. Those depend on the dynamics process itself of the Indian Ocean and the Pacific Ocean and the coupled air-sea interactions between the two vertical circulations over the two oceans.(5) The application of air-sea bilayer correlation pattern to precipitation forecast in rainy seasonThe "three dimensional air-sea interaction index" is defined considering the anomaly spatial scale of the three key regions and the lead-lag relationships between them. The index is well correlated with the precipitation field and its model error field of our country, and it has positive correlation with the northern part of the precipitation field and negative correlation with the southern part of the precipitation field. So it is indicator function for precipitation prediction. By using dynamical-statistic objective forecasting method and considering the index as one of the forecasting factor to correct the model error of the precipitation field, it is found that the technique has a well corrected result for the mid-lower Yangtze River. The independent sample regressions of ACC in the region from2005-2010show0.21improvement of the analogue correction compared with the operational system correction.