Research On Drought Characteristics Analysis And Forecast Approach Of Chaoyang Area

In recent years, extreme weather events increase significantly and floods and droughts occur more frequently with die change of global climate,these disasters lead to great economic loss every year and threaten humans,normal social and economic activities seriously. The global water crisis makes utilizing and exploiting water resources be a significant project to national economy development and humans,survival environment improvement. While China is a country with a shortage of water resources, drought becomes the main restrictive factor to economic development and agricultural production,some critical factors such as drought characteristics,occurrence regularities should be studied deeply and urgently. Chaoyang lies in the west of Liaoning province with less and concentrated precipitation, belongs to semi-arid, semi-humid and easy to aridity region, drought and poor water resources is the basic environmental characteristics of Chaoyang,frequent droughts bring serious influence th agricultural production and people’s life,that are the main cause to agricultural backwardness and main constraint to economic development. Therefore, the analysis and scientinc forecast about drought occurrence characteristics and regularities of Chaoyang possesses very important reality and necessity, it will provide decision foundation for managers,thus droughts-thsisting measures can be formulated timely and the losses from the droughts can be decreased to minimum. In this paper, the main research contend are as follows:(1) Drought characteristics analysis methods of Chaoyang Annual and quarterly drought characteristics of each year are analyzed with the methods of palmer drought severity index and precipitation anomaly percentage according to 60 years of daily precipitation, temperature and other monitoring data from 1952 to 2011 in 6 different areas such as Jianping,Beipiao, Chaoyang, Yebaishou,Lingyuan and Kazuo of Chaoyang in this paper,the additional error control method introducing Malikov criterion improves the drought index description precision.The COntrast experiment of two methods in flood and drought characteristics analysis shows that tile precipitation anomaly percentage method can directly reflect the precipitation from the average level,but the drought sensitivity is low.Palmer Drought Severky Index can describe the drought intensity in Chaoyang area more accurately, and the method is more suitable for analyzing characteristics in Chaoyang where is dry and more actual evaporation,because it has considered the precipitation,temperature,surface transpiration and other factors.(2)Key Techniques of Drought Characteristics AnalysisWavelet function selection and decomposition layers principle in the wavelet analysis of drought characteristics palmer index are discussed deeply based on wavelet transform, drought wavelet decomposition and characteristics analysis solution are established:selecting wavelet function based on following principles, characteristics drought index signal correlates maximum with wavelet function, synthetically considers wavelet vanishing movements, selects mother wavelet with suitable support length. Selecting wavelet decomposition layers based on following principles, avoids frequency division with frequency/cycle in the drought characteristics analysis, reduces energy leakage, provides theory basis for reasonable application of wavelet decomposition in drought characteristics analysis. The experiment about 59 years of drought characteristics analysis in Chaoyang using the combination of wavelet decomposition and palmer index shows that the drought occurs with the rules of 5,10 and 20 years repeated cycle, and the drought aggravates gradually.(3) Drought forecast optimization method with grey modelModel selection, sequence length determination, parameter calculation and other judgment methods about precipitation forecast using gray model are researched, the research using quasi smoothness of original sequence and quasi index regularity shows that GM(1,1) model is not suitable for direct forecast of precipitation, precipitation is 2 layers decomposed with wavelet method,1 low frequency components and 2 high frequency components are obtained, different frequency components are used to model forecast:low frequency components are used to gray model, appropriate model group is selected to improve the simulation accuracy, the optimum sequence length is 4, simulation error is 0.0071.High frequency components are used to the sequence of contour lines and contour time with waveform forecast method, future time value of each contour line is predicted separately, thus waveform forecast of high frequency components is obtained, average simulation error is 0.159 and 0.209. At last, the forecast values of low frequency components and high frequency components are used to construct the forecast values of precipitation, synthetic simulation error is 0.051.(4) Droughts forecast method based on artificial neural networkNeural network application method in the precipitation forecast is researched and BP network and RBF network training forecast methods are discussed in this paper, the results show that RBF network is faster than BP network, but lower in precision, experimental training time and forecast error are 0.390s/1.406s、0.3329/0.1887,so high-precision BP network forecast method is selected in this experiment. The combinative method of BP neural network and gray model is used further to simulative forecast to 10 years of precipitation data from 2002 to 2011 in Chaoyang, at first, gray model is established to 10 years of precipitation data, then BP neural network model to residual error of gray simulative, thus simulative value of residual error is generated and simulative value of precipitation is reconstructed, the average relative error is 0.0799 after testing, the forecast error reduces 0.1311 than using gray model simply, the results show that the combinative method of BP neural network and gray model is suitable for precipitation forecast of Chaoyang.The theories and methods of drought characteristics analysis and droughts forecast used in the experiments are scientific and feasible, the results also conform to actual situation. They are valuable reference basis for droughts forecast and response of Chaoyang, and scientific decision-making basis for managers to drought resistance and disaster mitigation. They are also very important reference for droughts research and forecast of other regions.