Study On The Change Law And Describe Method Of River Runoff Under The Nature-Manpower Duality Mode

Because of the influence of climate change and human activities, change has been occurred of hydrologic cycle character, that Changing from single to nature-manpower duality mode of water cycle. By way of the main tache, river runoff is the important foundation of the comprehensive development and utilization of water resources, of scientific management and optimal dispatching, affected by human activities and the global changes. In this paper, combining with the national nature science fund, illustrating as the mid-upstream of the Yellow river and the upstream of the Hanjiang River, adopting modern system scientific approach, we study on river runoff variation law in duality mode, and probe into its described method.In the end, obtain plenty of theory and practice conclusions, and offer foundation for water resources reasonable development and utilization. So provided with important theory and practice value. Main results as follows:1. Study on the influence of climate change and human activity on water cycle trait. Analyzed the influence of climate change and human activity on water cycle character, water resources and its management, offered concept of nature-manpower duality water cycle and illustrated its meaning.2. Probed into characteristic, form and medium-and long-range forecast approach of hydrologic time series analysis, creatively handled hydrologic time series analysis with wavelet theory, discussed variation trait of riverrunoff series under the different time measurement, initiated the new approach of hydrologic time series analysis.3. Analyzed runoff variation law and cause of mid and upstream of the Yellow river and the upstream of the Hanjiang River adopting many methods, obtained naturally annual runoff series of mid and upstream of the Yellow river, measured annual runoff series of upstream of the Hanjiang River, drawn a conclusion of its non-obvious variation in the whole serial sample.4. Analyzed on affected factors of runoff forming and variation law, such climate condition as precipitation, mainly controlled river runoff variation. But variations of drainage basin underlying surface and local climate factor brought by human activity affected naturally annual runoff forming and variation law.5. Medium- and long-range forecast of river runoff. Included uncertainty factors in the annual runoff long-range forecast, such as complexity and randomization of system dynamics itself and recognizant non-integrality, make it difficulty. Handled annual flow series with RBF NN model and mix-regression model based on wavelet analysis, and threshold auto-regression model based on genetic simulation anneal algorithm to medium-and long-range forecast. Based on the theory of non-linearity combination forecast, established combination forecast model based on BP NN, obtained better practicability about flow forecast outcome, provided a probably new approach for runoff medium- and long-range forecast.6. Firstly introduced the theory of non-linearity combination forecast into runoff medium- and long-range forecast, make up unilateralism of the single model of forecast, concentrated plenty of available information, make the best of the merit of the different model, accordingly improved adaptation ability of forecast model for environment change. Make it clearthat combination forecast has lower mean square error than any forecast model.7.Coincidenced forecast results of radial network model mix regression model and threshold auto-regression model based on genetic simulation anneal algorithm by non-linearity combination forecast model. Adopted known information and laws, discovered furthest forecast range by the whole uncertainty analysis. Both kept basic statistic character of original observation series and considered evolvement rule of runoff series.