Research On Theories And Methods Of Complexity Analysis For Crop Water Requirement Scale Features

Water is the source and basis for all beings. Immediate guarantee for crop water requirement plays an important role in the agricultural production. Irrigation plays a more important part in our country's agricultural production and food security than in other countries. It costs 50% of our country's irrigated cultivated land to produce over 70%of the total grain output, over 80% of the total cotton output, over 90% of the total vegetables output. Effective guarantee for irrigation water requirement serves as the basis and presupposition for the improvement of our country's grain production level. But our country's agriculture irrigation is faced with the serious problem of water shortage, because of the limited water resource, the inefficiency of irrigation water use, the change of water consuming structure and so on. Agriculture water shortage becomes a major restricting factor for agricultural production, especially for those arid and semi-arid areas.The study of crop water requirement has always focused on the minor scale, great progress has been made in the study and application of the experience formula, the water quantity balance method, micrometeorology and so on. But in practice, we need to grasp the trend of the regional crop water requirement, in the study of which we usually draw a linear distribution chart to get the amount of regional water requirement. But the calculation result has much deviation from the reality due to the unavoidable influence of meteorology factors and the spatial-temporal heterogeneity of crop water requirement. As far as the scale is concerned, there's mainly two reasons. On the one hand, the system of regional crop water requirement, involving meteorology, hydrology, soil and etc, usually exists on a comparatively larger spatial-temporal scale. But most relevant variables can only be measured on the spot within a short time and a limited range, so that the data collected in this way can't meet the demand of calculation. On the other hand, due to meteorological, soil, planting conditions and so on, crop water requirement is obviously characterized by its spatial-temporal heterogeneity on different spatial-temporal scales. As the result of this heterogeneity, the theories and models of crop water requirement are highly scale-dependent, that is, the model on one scale can't be applied to a higher or a lower level of scale. Based upon the above, this thesis, mainly concentrates on the study of crop water requirement scale features by means of theoretical research, mathematical model, experimental data analysis and test. The major achievements are as follows:(1) Surface layer micro-meteorological factors are key influential elements for crop water requirement. Based upon the complete systematic analysis of how meteorologicalfactors influence crop water requirement, the thesis researches into the applicability of different calculating methods for crop water requirement Considering water requirement features of southern crops, the thesis concludes that it is advisable to calculate Reference Crop Evapotranspiration (RCE) on the field scale by means of the standardized Penman-Monteith formula in the southern area.(2) Considering the spatial-temporal stochastic features of RCE, the thesis, for the first time, uses the fractal theory to study the spatial-temporal fractal features of RCE, and the no-scale interval of RCE, which provides a theoretical basis for applying the achievements from the data areas to the lack-data areas.(3) Within a certain region, RCE, in the unified structure, is characterized by its information system. Thus, the thesis, for the first time, based upon the information entropy theory, proposes a new method to analyze the spatial scale features of RCE, and studies the spatial distribution features of RCE, which to a certain degree advances the study of RCE spatial-temporal scale features.(4) RCE time series is characterized by its periodicity, due to crops' physiological mechanism and the unsteady, non-linear stochastic variation of meteorological factors. Considering to the time-frequency localization of the wavelet analysis theory, this thesis, for the first time, studies the multi-time scale feature of RCE time series process, and reveals the evolutionary law of RCE time series process, which provides a theoretical basis for the RCE spatial-temporal scale features research.(5) RCE time series, characterized by its nonlinearity, may vary with the change of time, under the influence of stochastic variation of meteorological factors such as temperature, sunlight, wind speed, vapor pressure and so on. Therefore, this thesis, based upon the wavelet analysis and the neural network theory, establishes a RCE forecasting model with the wavelet and the BP network combined together, which serves as a new method for non-linear systematic time series forecasting.(6) Planting structure is another important element which influences the crop water requirement. Considering the present situation of the planting structure of our country, and according to the comparative advantage theory, this thesis, for the first time, proposes a market-oriented, efficiency-centered, output-oriented comparative advantage approach to adjusting the planting structure, which provides a theoretical basis for optimizing the region planting structure and the region resource allocations..(7) Within a certain range of area, the irrigation canal system of irrigated fields is a fractal unity similar to a natural drainage area system. It is an important water conveying channel to ensure the crop water requirement. According to the fractal theory, this thesis,for the first time, studies the spatial fractal features of the irrigation canal system. By means of correlation-principle, this thesis proves the strong correlation existing between the fractal dimension of the irrigation canal system and its irrigation module, which provides a new approach to estimating irrigation canal arrangement.The study of crop water requirement features on the spatial-temporal scale involves meteorology, soil, hydrology, planting structure and so on. The theoretical and methodological research requires further development. This thesis makes some significant researches and proposes some innovative views which are applied to the field and region scale of the Yangtze valley, and yield some practical use-value. Owing to the restriction of time and data, some problems of the thesis requires further research.