【作者】 彭胜民；

【导师】 王福林；

【作者基本信息】 黑龙江八一农垦大学 ， 农业机械化工程， 2010， 博士

【摘要】 水土资源不但是人类社会生存最基本的物质基础,而且是生产与发展的基本条件,更是决定地区间经济差异和国家实力大小的重要因素。水土资源问题长期以来都是世界各国最为关注的焦点问题之一,尤其是1970年以后,由于人类日益增长的水土资源需求和其本身的稀缺性、有限性之间的矛盾越来越大,水土资源开发利用问题更加明显地成为全球性的热点问题之一。本论文立足于黑龙江省齐齐哈尔市,结合国家“十一五”科技支撑计划项目,采用现代数据处理方法和手段,对齐齐哈尔市水土资源发展趋势、水土资源多时间尺度特性、水土资源承载力、不同尺度下粮食增产潜力分析与预测及种植结构优化调整等问题进行了深入细致的研究,得出以下主要结论:(1)分别采用BP神经网络模型、改进的GM(1,1)模型、偏最小二乘回归模型和小波随机耦合建立了齐齐哈尔市年降水量、年地下水平均埋深、年径流量、耕地面积数学预测模型,研究结果表明:所建立的模型均能很好地对齐齐哈尔市的水土资源进行预测。其中:降水量预测结果表明在最近几年会持续较多,不会有严重旱灾发生;地下水预测结果表明如果仍然按照当前开采模式继续下去,未来10年内地下水埋深降幅为6.45m左右;耕地面积预测结果表明未来十年内耕地面积波动性较大,呈现增减交替波动趋势。(2)主要采用小波理论,分析了齐齐哈尔市年降水量、主汛期降水量、年径流量、月径流量、年耕地面积的多时间尺度变化特征。研究结果表明:年降水量的第一、第二、第三主周期分别为29年、6年和16年,主汛期降水量的第一、第二、第三主周期分别为6年、11年和29年,主汛期降水量的变化在很大程度上控制着年降水量的变化;年径流量序列的第一、第二、第三主周期分别为32年、5年(或7年)和13年,月径流量序列的第一、第二、第三主周期分别为384个月(32年)、84个月(7年)的39个月(3.3年),月径流量的变化决定着年径流量的变化;耕地面积序列的第一、第二、第三主周期分别为40年(甚至更大)、17年和23年,这几个主周期共同控制着齐齐哈尔市耕地面积的动态变化规律。(3)通过采用DPSIR模型(Driving Force-Pressure-State-Impact-Response Framework,驱动力-压力-状态-影响-响应框架)构建了齐齐哈尔市水土资源承载力评价指标体系,在指标体系信度检验的基础上,采用基于实数编码加速遗传算法优化的投影寻踪评价模型对齐齐哈尔市市区及所属各市县的水土资源承载力进行评价,结果表明:市区>甘南县>龙江县>富裕县>克东县>泰来县>依安县>讷河市>克山县>拜泉县。(4)各粮食增产影响因子对黑龙江省粮食总产量影响程度排序为:粮食单产>粮食播种面积>农田成灾面积>化肥施用量>农机总动力>有效灌溉面积>农药使用量,而在齐齐哈尔市粮食增产中起着关键作用的影响因子依次为有效灌溉面积、农药使用量、农机总动力。上述研究成果可以为不同尺度下黑龙江省和齐齐哈尔市制定粮食发展战略提供参考,同时也说明,随着研究尺度的改变,各粮食增产影响因子对区域粮食总产量影响程度也是在不断变化的。(5)从复杂适应系统理论的基本思想着手,对复杂适应系统及理论的特点、主体及运行机制进行了初探,分析了种植结构调整的复杂性及CAS特征,搭建了基于区域农业水土资源复杂适应系统的种植结构优化调整模型框架,并根据整体模型特点,采用多目标协同进化遗传算法对所建模型进行了求解,在此基础上,从子系统间协调度和系统总协调度入手,对农业水土资源系统效益的协调性进行了评价,结果表明:种植结构优化调整后的研究区域农业水土资源系统向着更加协调的方向演化。(6)黑龙江省粮食总产量中长期多元回归预报模型预测结果表明:未来黑龙江省粮食总产量将表现出持续增长的态势,2015、2020、2025、2030年的粮食总产量分别为5193.54万t、5701.73万t、6208.20万t、6703.89万t;齐齐哈尔市中长期粮食总产量主成分分析预报模型预测结果表明:齐齐哈尔市未来粮食总产量也将呈现出增加的趋势,2015、2020、2025、2030年的粮食总产量分别为844.33万t、907.33万t、974.33万t、1044.8万t。更多还原

【Abstract】 Water and soil resources is not only the most basic material of the human society survival, but also the basic conditions for production and development. It also decides regional economic disparities and is an important factor of the national power size. Water and soil resources problem has been one of the most focus problems for a long time, especially since the 1970s, as the contradiction among the growing human demand for water and soil resources, the scarcity and limited of its own, the problem of soil and water resources development and utilization clearly become a global hot topic.This paper adopts the modern data processing methods and combine with the national "Eleventh Five-Year" science and technology support projects to systemically study water and soil resources trends, multiple time scale, carrying capacity, the grain yield increase potential under different spatial scales and planting structure adjustment in Qiqihar, Heilongjiang Province. The main research contents and production are presented as follows:(1)Some mathematical models are set up to predict annual precipitation, average annual groundwater depth, annual runoff and arable land, such as BP-ANN model, improved GM(1,1) model, PLS model and wavelet random coupling model. The results show that: the established models are better to forecast water and soil resources. The precipitation forecasting results show that there will be not severe drought in the next 10 years; The groundwater forecasting results show that if mining continued in the current pattern, the groundwater depth decline will be 6.45m in the next 10 years; The arable land forecasting results show that arable land area will be volatile in the next 10 years and present increase or decrease trends.(2)This paper mainly use wavelet theory to analyze the multiple time scales of annual precipitation, the main flood season precipitation, annual runoff, monthly runoff and arable land of Qiqihar. The research results show: The first, second and third main cycle of annual runoff are 29 years, 6 years and16 years. The first, second and third main cycle of the main flood season precipitation are 6 years, 11 years and 29 years. The main flood season precipitation change to a large extent controls the change of annual precipitation; The first, second and third main cycle of annual runoff are 32 years, 5 years(or 7 years) and 13 years. The first, second and third main cycle of monthly runoff are 384 months(32 years), 84 months(7 years) and 39 months(3.3 years). The monthly runoff change determines the annual runoff change; The first, second and third main cycle of arable land are 40 years(even greater), 17 years and 32 years,and these main cycles commonly control the dynamic change law of arable land in Qiqihar.(3)This paper adopts DPSIR model (Driving Force-Pressure-State-Impact-Response Framework) to construct water and soil resources carrying capacity evaluation index system. On the basis of index system reliability test,the RAGA-PPE model is built to evaluate water and soil resources carrying capacity of Qiqihar urban and its counties,the results show:the urban > Gannan > Longjiang > Fuyu > Kedong > Tailai > Yi’an > Nehe > Keshan > Baiquan.(4)Each grain production impact factor for the Sort of impact by grain output in Heilongjiang Province:Grain yield>Sown area of grain>Agricultural disaster area>Chemical Fertilizers>Total Power>Effective irrigation area>Pesticide use,but the key role impact factor of grain production in Qiqihar City were Effective irrigation area、Pesticide use、Total Power.Above-mentioned research can provide reference to formulate food development of different scales in Heilongjiang Province and Qiqihar City,At the same time,It illustrate that with the changes of research scale,Each grain production impact factor for the region’s grain output Sort of impact are also changing all the time.(5) From the basic idea of complex adaptive systems theory, this paper preliminarily explore the characteristics, agent and operating mechanism of CAS theory and analyze the complexity and CAS characteristics of planting structure adjustment, the planting structure adjustment model framework based on agriculture water and soil resources complex adaptive system is built. At the same time, co-evolutionary genetic algorithm is adopt to solve the built model. On this basis, from the coordination degree of the subsystem and overall system, the coordination degree of agriculture water and soil resources system benefit can be evaluated. The results show: agricultural water and soil resources system after planting structure optimization adjustment evolve toward a more coordinated direction.(6) The Long-term forecast grain output result by Multiple regression prediction model in Heilongjiang Province shows:the future grain output in Heilongjiang Province will express the continued growth Situation,Grain output in 2015、2020、2025、2030years Were 5193.54tons、5701.73tons、6208.20tons、6703.89tons;The Long-term forecast grain output result by Principal component analysis and prediction model in Qiqihar City shows:the future grain output in Qiqihar City will also express the growth Situation,Grain output in 2015、2020、2025、2030years Were 844.33tons、907.33tons、974.33tons、1044.8tons.更多还原