【作者】 韩勇鸿；

【导师】 樊贵盛；

【作者基本信息】 太原理工大学 ， 环境工程， 2013， 博士

【摘要】 土壤持水参数传输函数的研究是土壤墒情自动监测预报、农田作物实时灌溉预报和农田节水灌溉技术参数确定的主要依据,为灌区实现综合节水和水资源的可持续利用提供强有力的技术支撑。研究成果对我国农业地区土壤水资源的开发利用与管理具有重要的理论意义和实际应用价值。本文以国家自然科学基金资助项目(40671081)和山西省科技攻关基金资助项目(20100311124)为依托,基于晋中地区、临汾地区和工大试验小区的400多组大田试验数据,分析了不同类型土壤的持水参数的变化特性,揭示土壤质地、土壤结构、有机质含量、含盐量对土壤持水参数的影响,建立了影响因素与土壤持水参数间的定量模型；利用多元线性、多元非线性和BP神经网络等土壤传输函数,对给定土壤质地、不同土壤质地、考虑含盐量等三种条件下的土壤持水参数进行预测,并对预测结果进行验证和分析,最后确定土壤持水参数的最优的土壤传输模型。主要研究成果和创新点如下：(1)影响土壤持水参数的主要因素有：土壤质地、土壤结构、土壤有机质含量和十壤含盐量。(2)田间持水率和凋萎系数两个土壤持水参数随着土壤结构由疏松变为密实而减少,它们之间符合二次函数关系；随着土壤粘粒含量的增加而增大,它们之间符合对数函数关系；随土壤砂粒含量的增大而减小,它们之间符合对数函数关系；随着土壤有机质含量的增加而增大,它们之间符合对数函数关系；随着土壤含盐量的增加而增加,它们之间符合对数函数关系。(3)利用多元线性、多元非线性和BP神经网络传输模型预测土壤田间持水参数的预测误差都在10%以下,三种土壤传输模型都是可行的,BP神经网络传输模型的预测结果最好。本研究是以指导农田灌溉和水资源可持续利用为主要出发点,对不同类型的土壤持水参数的变化特性、影响因素、预测模型等内容进行了较为全面深入的研究,实现了用多种土壤传输函数来预测土壤持水参数。但是由于数据观测误差、试验方法或预测方法等原因,使得研究成果在区域范围内使用,还需要在大范围内进一步试验和验证。更多还原

【Abstract】 The study on transfer functions of soil water retention parameter is the primary reason for automatic monitoring and forecasting of soil moisture, forecasting of irrigating crop in real time and determining technological parameters of water-saving irrigation, which provides strong technical supports for realizing comprehensive water-saving in irrigated area and the sustainable use of water resources. The results of study have great significance in theory and application on utilization and management of soil water resources in agricultural district of China.This paper is supported by the National Natural Science Foundation of China (40671081) and Key Programs for Science and Technology Development Foundation of Shanxi Province, China (20100311124).It is based on more than400field test data of experimental plots in Jinzhong district, Linfen district and Technology University of Taiyuan. Variation characteristics of different types of soil water-holding parameter were analyzed. It is revealing that the effects of soil water-holding parameter on soil texture, soil structure, organic content and salinity, establishing quantitative model of influencing factors and soil water-holding parameter. Utilizing soil transmission function-multiple linear, multiple nonlinear and BP neural network, soil water-holding parameter is forecasted under three conditions-given soil texture, different soil texture and consideration of salinity and the forecast results were analyzed. Finally, the optimal soil transmission model of soil water-holding parameter was determined. Main study findings and innovation points are as follows:(1)The main influence factors of soil water-holding parameter:soil texture, soil structure, organic content of soil and salinity. (2)Field moisture and wilting coefficient decrease with the increase of soil dry bulk density and soil sand grain content, increase with the increase of soil clay content, organic content and salinity. Field moisture and wilting coefficient fit quadratic function with soil dry bulk density. Field moisture and wilting coefficient fit logarithm function with soil clay content, so do soil sand grain content, organic content and salinity.(3)All forecast errors are under10%by utilizing soil transmission models-multiple linear, multiple nonlinear and BP neural network to forecast soil field water-holding parameter. Three soil transmission models are feasible and the forecast results of BP neural network transmission model are the best.The main starting point of this study is the guidance of field irrigation and sustainable utilization of water resources, a thorough study of variation characteristics, influence factors and forecasting model of different types of soil water-holding parameter was conducted. Forecasting soil water-holding parameter with many kinds of soil transmission function was realized. However, because of data error of observation, test method or forecast method, the study findings can’t be used at the regional scale until there is a further test and verify in a large scale.更多还原