【作者】 刘战东；

【导师】 段爱旺；

【作者基本信息】 中国农业科学院 ， 农业水土工程， 2012， 博士

【摘要】 随着我国社会经济的快速发展,水资源短缺越来越严重,农业水资源供需矛盾也在不断加剧。灌溉用水是农业用水的主体,在整个社会的水资源消耗中占有很大的比重,但目前我国的灌溉水利用效率与效益都还较低,用水浪费现象仍较普遍。因此,发展节水高效的灌溉技术与管理模式对于缓解水资源供需矛盾,保障国家粮食安全都具有十分重要的意义。在这样的背景下,深入研究作物生育期间的降雨利用过程,分析降水利用影响因子及其作用机制,并对相关因子的影响进行数值模拟,可为科学估计作物生育期间降雨的有效转化程度,以及制定优化灌溉制度提供重要的理论依据,对于农业水资源的高效利用也具有重要的理论和现实意义。研究工作于2009-2011年在河南焦作市广利灌区灌溉试验站进行。选取冬小麦作为研究对象,通过三年的田间试验,对冬小麦的降雨冠层截留特性,降雨产流、入渗特征,降水的蒸发及进入土壤后的再分布规律等内容进行了试验观察及定量模拟,并通过试验探讨分析了不同覆盖方式和降雨特性对降雨利用过程和利用效率的影响,取得的主要研究结果如下：(1)采用“简易吸水法”分别研究了冬小麦单株和群体的降雨截留性能。结果表明：冬小麦抽穗前,叶片数量对单株截留量影响显著(P<0.05)；而在相同叶片数的情况下,不同生育时期的单株截留量也有显著的差异(P<0.01)。冬小麦单株截留量分别与株高、叶面积、鲜重呈线性正相关关系,吸水率则与株高、鲜重呈线性负相关关系。冬小麦群体截留量分别与LAI、地上部生物量呈线性正相关关系,吸水率则分别与它们呈线性负相关关系。从拔节至成熟,冬小麦冠层截留量呈现先增加后减少的趋势,最大冠层截留量出现在抽穗期(1.28mm),不同生育期的冠层截留量之间差异极显著(P<0.01)。模拟降雨试验结果显示,冬小麦生育期透过冠层落在棵间的雨量与降雨总量呈显著的正相关关系(P<0.01)；冠层截留量与降雨总量呈显著的幂函数关系(P<0.01)。降雨强度与落在棵间的雨量占总降水量的百分比呈负指数函数关系(P<0.01),与冠层截留量所占百分比呈负幂函数关系(P<0.01)。不同降雨强度下冬小麦冠层截留过程趋势一致,降雨强度越小,其达到冠层截留容量所需时间越长。降雨强度对冠层截留容量没有明显影响。在雨量恒定条件下,截留量随雨强的增加而减小,呈明显的负相关关系。通过分析试验数据构建了冬小麦冠层降雨截留过程模型：基于实测数据,确定了模型中表征冬小麦降雨蒸发能力的参数α为0.008。模型的模拟值和实测值有较好的一致性,显示建立的模型适用于冬小麦冠层截留的计算。(2)通过模拟降雨试验,研究了冬小麦田降雨强度(RI)、冠层覆盖度(用Leaf area index表示,LAI)及0-40cm初始土壤剖面含水量(θ40)对降雨产流、入渗特征的影响。结果表明：在其他影响因子保持稳定的条件下,径流强度、累积径流量、入渗速率和累积入渗量分别与降雨历时(t)具有显著的负指数函数、幂函数、幂函数和对数函数关系(P<0.01)。产流时间随RI增大而提前,两者呈显著幂函数关系(P<0.01)；径流强度、累计径流量和径流系数随RI增大而增大；平均入渗率、稳定入渗率和入渗量随RI增大而增大,入渗率趋于稳定值的时间随RI增大而提前,降雨蓄积系数则随RI增大而减小。产流时间随LAI的减少而提前,两者呈显著线性函数关系(P<0.01)；径流强度、径流量及径流系数随LAI的增加而减小；平均入渗率、稳定入渗率、入渗量和降雨蓄积系数随LAI增大而增大,入渗率趋于稳定值的时间则随LAI的增大而延长；当降雨强度增大时,LAI对麦田产流、入渗过程的影响减弱。在RI和LAI保持不变时,产流时间随θ40增大而提前,两者呈显著线性函数关系(P<0.01),而40cm以下土层的含水量对产流时间影响相对较小；径流强度、径流量及径流系数随θ40增大而增大,但稳定径流强度基本相同；入渗率趋于稳定值的时间随θ40增大而提前,平均入渗率、入渗量和降雨蓄积系数随θ40增大而减小,但稳定入渗率基本相同。通过多元回归分析,建立了基于RI、LAI和θ40的产流时间计算模型：tp=20.3070RI-1.0761LAI1.5209θ0-1.1844经检验,模型具有较好的模拟效果。径流强度、累积径流量、入渗率、累积入渗量可最终表示为t、RI、LAI和θ40等四因素的函数；建立的径流系数和降雨蓄积系数多元回归计算模型如下RC=-0.19188+0.2282RI-0.00785LAI+0.009231θ40RSC=1.5754RI-0.5437LAI0.0430θ40-0.2339(3)通过人工模拟小雨至特大暴雨6个降雨级别的降水过程,研究了不同降雨条件下的麦田土壤蒸发和土壤水再分布规律,结果表明：在相同的气象条件下,不同降水强度处理下的土壤蒸发过程具有相同的变化趋势。日土壤蒸发量和累积蒸发量均随着降雨量的提高呈对数函数方式增加；不同处理下白天土壤蒸发量有着显著的差异,但晚间的差异不明显；一次降水处理后,麦田土壤累计蒸发量占降雨总量的比例(E/P)随降雨总量的增加呈幂函数减小。土壤日蒸发量与气象因子的相关关系说明,降雨后麦田土壤蒸发与日照时数、20cm蒸发皿水面蒸发、日最高气温、日平均气温相关性程度较高,均达到显著和极显著水平。土壤蒸发量对各气象因子的响应程度随降雨级别(降雨量和降雨强度)的增加而变大。土壤日蒸发量与0-100cm土体内各土层的含水量呈现显著的相关性(P<0.01),且相关系数随降雨级别的提高呈不断增加的趋势。降雨级别(降雨量和降雨强度)越大,土壤水再分布影响的土层就越深,土壤含水量变化幅度越大,同时土壤水再分配过程所需的时间越长。湿润锋运移速率随降雨级别提升而增大,同时湿润锋运移深度也相应增加。在同一降雨级别下,湿润锋运移和土壤水再分布过程随着土壤初始含水量的增加而加快。从整体上看,受作物根系生长发育的影响,返青期0-100cm土层水分变化幅度不如拔节期、灌浆期明显。降雨级别越大,降水转化为土壤水的量也越多,但从转化效率上看,中等级别降雨最高。利用HYDRUS-1D模型对受降雨、蒸发和作物根系吸水影响的土壤不饱和区含水量分布的变化进行模拟,结果表明HYDRUS-1D模型的模拟结果可以较好地反映不同降雨条件下麦田土壤水分变化的真实情况,因此可以作为实际管理中对土壤水分变化的一种预测手段使用。(4)在大田中设置地膜覆盖(PM)、4种秸秆覆盖(覆盖量分别为1500,4500,7500,10500kg/hm2,分别标记为SM15、SM45、SM75和SM105和无覆盖处理(CK),研究不同覆盖方式和降雨特性对降雨后冬小麦棵间蒸发量、土壤剖面水分分布和降雨土壤蓄积量的影响。结果表明：模拟降雨后,各处理的日土壤蒸发量和土壤蒸发累积量有着相同的变化趋势；同一时间段内不同覆盖处理的土壤蒸发量差异明显,均表现为SM105<SM75<PM<SM45<SM15<CK冬小麦群体冠层结构对雨后土壤蒸发量影响显著,同等降雨条件下,同一覆盖处理拔节期前的日土壤蒸发量明显大于拔节期后的相应数值。不同覆盖处理间0-100cm土层剖面的含水量差异明显，其中PM的最小，CK的次之，SM105的最大；降雨土壤蓄积量整体也表现为：SM105＞SM75＞SM45＞SM15＞CK＞PM；受植株冠层降雨截留量增大的影响，冬小麦拔节期后各处理的土壤含水量和降雨土壤蓄积量要明显低于拔节期前的数值。同一覆盖处理，60mm/h降雨强度条件下的降雨入渗深度和入渗量要明显高于40mm/h下的降雨入渗深度和入渗量。与60mm/h降雨强度相比，40mm/h降雨强度下各处理拔节前与拨节后的降雨土壤蓄积量变化幅度要明显的大。更多还原

【Abstract】 With rapid social and economic development, shortage of water resources is becoming more andmore serious, and competition for limited agricultural water resources also is becoming more and morebitterly in China. Water consumption for irrigation is the majority of agricultural water resourcesconsumption and also occupies a very important position in whole social water resources utilization.However, water use efficiency and return is still relative low and water waste is still very common andserious. So that, development of water-saveing irrigation techniquies and irrigation management modelsis very helpful for relieving severe competition for water resources and ensuring national security ofgrain supply. With this background, it is necessary and urgent to study systematically the transformationand utilization of rainfall during crop growth period, to analyse factors related with rainfall utilizationand their affecting mechanisms, and simulate numerically the relationships between relavant factors andrainfall utilization. The research results may provide an important theoretical basis for estimatingeffective conversion level of rainfall during crop growing period and a helpful tool for designingoptimal irrigation schedule, which has theoretical and practical significance for improving the utilizationof limited agricultural water resources in China.Field experiments were conducted at Guangli Irrigaiton Experimental Station located in Qinyangcounty, Jiaozuo city, Henan province in2009-2011. Characteristics of canopy interception, surfacerunoff and soil infiltration during a rainfall, and soil surface evaporation and redistribution of soil waterafter the rainfall were investgated and studied in winter wheat season, The effcets of soil surfacemulching patterns and characteristics of rainfall on utilization of rainfall on winter wheat field were alsostudied. The main results are as followings:（1）The capacity of intercepting rainfall in winter wheat at plant and canopy level was measuredand factors affected the capacity were analyzed with wet-absorption method. The results showed thatintercepted rainfall amount（IRA） by a winter wheat plant was effected significantly by leaf number（P<0.05） before heading. The IRA by a plant with same leaf number varied significantly at differentgrowing stages （P<0.01）. The IRA per plant increased linearly with leaf area, plant height and freshweight increasing, respectively. However, absorption rate decreased linearly with leaf area, plant heightand fresh weight increase. The IRA by the whole winter wheat canopy presented linear positivecorrelation with leaf area index （LAI）, and aboveground biomass（AB） separately, but absorption rateexpressed a negative correlation with LAI and aboveground biomass. From jointing stage to maturity,IRA by winter wheat canopy increased slowly, and then decreased gradually. The peak value of IRA（1.28mm） occurred at heading stage. The IRA values varied significantly in different growing stages（P<0.01）,Simulated rainfall experiment showed that water amount fell on interrow soil surface in winterwheat was positively correlated significantly with total rainfall （P<0.01）, while the relationship betweencanopy interception and total rainfall fitted significantly a power function （P<0.01）. There existed anegative exponential correlation between rainfall intensity and interrow throughfall percentage （P<0.01） and a negative power correlation between rainfall intensity and canopy interception percentage （P<0.01）. The canopy interception under different rainfall intensities changed similarly with growing stages of winter wheat, while the time from starting of rainfall to reaching canopy interception capacity was longer with lower rainfall intensity. Rainfall intensity did not show significant influence on canopy interception capacity of winter wheat. Rainfall Interception decreased with the increase of rainfall intensity, and showing a significant negative correlation with rainfall intensity under a fixed rainfall amount. A mechanism model for simulating rainfall interception process of winter wheat canopy was developed, and relavent parameters were determined based on experimental data.A parameter, a, indicating evaporative capacity of rainfall intercepted by winter wheat canopy, was set to0.008. Simulated values fitted well to measured values, indicating that the model is suitable for estimating rainfall interception of winter wheat canopy.（2） The effects of rainfall intensity （RI）, canopy coverage （expressed with Leaf area index, LAI） and initial water content of soil layer of0-40cm （θ4O） on surface runoff and water infiltration characteristics was investigated and analyzed with a simulated artificial rainfall experiment in winter wheat field. Results indicated that under other factors were fixed, the surface runoff intensity, cumulative runoff, infiltration rate and cumulative infiltration fitted obviously as negative exponent function, power function, power function, and logarithmic function of the rainfall duration（t）（P<0.01）, respectivally The time that surface runoff began to be seen was advanced with RI increased and can be fitted well as a power function （P<0.01）. Runoff intensity, accumulative runoff and runoff coefficient increased with RI increasing. As RI increasing, the time reached stable infiltration was advanced, and average infiltration rate, stable infiltration rate, and accumulative infiltration gradually increased, but rainfall storage coefficient decreased. As LAI decreasing, surface runoff appeared earlier and can be fitted significantly with a linear function （P<0.01）. Runoff intensity, accumulative runoff and runoff coefficient decreased with LAI increasing. Average infiltration rate, stable infiltration rate, accumulative infiltration volume and rainfall storage coefficient increased with LAI increasing, and the time reached stable infiltration was postponed with LAI increasing. As RI increasing, the effects of LAI on surface runoff and infiltration during a rainfall process diminished. Under a stable RI and LAI, the time surface runoff appeared was advanced with the increase of θ40, and can be fitted well as a positive linear function （P<0.01）, but there was little effects of initial moisture in soil layer of40-100cm on the time surface runoff appered. Runoff intensity, runoff volume and runoff coefficient increased with the increasing of θ40, but with a similar stable runoff intensity. Under fixed RI and LAI, the time that infiltration rate reached stable point was advanced, and infiltration rate, infiltration volume and rainfall storage coefficient decreased with the increasing of θ40, but with a similar stable infiltration rate.With multiple regression analysis, a numerical model describing the relationship among the time surface runoff appears and RI, LAI, θ40in winter wheat was established as: tp=2O.3O7ORI-1.0761LAI1.52O9θ0-1.1844 The practical use of the model showed that the modificated values fitted well with the measuredvalues.Runoff intensity, cumulative runoff volume, infiltration rate and cumulative infiltration volume canall be fitted as a function of t, RI, LAI and θ40, Respectively. The regression models for calculatingrunoff coefficient and rainfall storage coefficient were established respectively as:RC=0.19188+0.2282RI-0.00785LAI+0.009231θ₄0RSC=1.5754RI^-0.5437LAI^0.0430θ₄₀^-0.2339（3）Soil evaporation and soil water redistribution in winter wheat field under different rainfallsituations were explored with simulating6rainfall grades, from light rain to torrential rainfall. Resultsshowed that under similar meteorological conditions, soil evaporation varied with growing stages, buton a similar pattern for different rainfall grades. Both daily soil evaporation and cumulative evaporationcan be fitted well as a positive logarithmic function of the rainfall volume. The soil evaporation volumesin different rainfall grades in daytime was significantly different, but the differences in night time werenot obvious. The ratio of soil evaporation to rainfall （E/P） in winter wheat was fitted well as a negativepower function of rainfall. Correlating meteorological factors and soil evaporation showed that soilevaporation after a rain was signicantly correlative with sunshine hours,20cm Pan evaporation, dailymaximum temperature and mean temperature in winter wheat field （P<0.05or P<0.01）, respectively.Effects of meteorological factors on soil evaporation increased with the rising of rainfall grades （rainfallvolume and rainfall intensity）. The correlation between daily soil evaporation and water content in soillayer of0-100cm was very significant （P<0.01）, and the correlation coefficients for each soil layers allincreased with rainfall grade rising.The higher a rainfall grade （rainfall volume and rainfall intensity） was, the deeper a soil profileinfluenced by soil water redistribution was, the larger soil water content change, and the longer the timeneeded to finish soil water redistribution was. Movement velocity of wetting front increased withrainfall grade rising, while final depth of wetting front reached also increased. In addition, with a fixedrainfall grade, wetting front movement velocity increased, but the time needed to finish soil waterredistribution shorted with the increasing of initial soil water content. The changing range of soilmoisture within0-100cm at reviving stage was smaller than those at jointing stage and grain fillingstage, showed the influence of crop root growth. The higher rainfall grade was, the greater the ratio ofrainfall volume to water stored finally in soil was. However, the transformation efficiency of mediumgrade rainfall to soil water was relatively higher.With HYDRUS-1D model, changes of soil water content in unsaturated zone caused by rainfall,evaporation and crop root water uptake was simulated. Results showed that simulated values fitted verywell to the measured valued in all rainfall situations in winter wheat field, and the model is a suitabletool of forecasting soil moisture in practical field water management.（4）Six mulching theatments, consisting of one plastic mulching （PM）, four straw mulching（with1500,4500,7500, and10500kg/hm2, and labeled as SM15, SM45, SM75and SM105, respectively）,and one no mulching（CK）, was set in winter wheat field to study the effectes mulching models and rainfall characteristics on interraw soil surface evaporation, soil moisture distribution and soil waterstorage after a rainfall under simulated rainfall situation. Results indicated that the changing trends ofdaily soil evaporation and cumulative evaporation are very similar for all mulching treatments, but thedifferences of cumlative soil evaporation during a same period among six mulching treatments wereobvious, and ranged as SM105<SM75<PM<SM45<SM15<CK. Canopy structure of winter wheataffected significantly soil evaporation rate after a rain. Under same rainfall situation,（daily orcumulative） soil evaporation before jointing stage was significantly greater than that after the jointingstage for a same treatment. There were significant differences for soil moisture in0-100cm soil layeramong six treatments, and smallest for PM, and greatest for M105. For soil water storage after a rainfall,the results can be ranged as: SM105＞SM75＞SM45＞SM15＞CK＞PM. By canopy interceptionincreasing, soil moisture and soil water storage after jointing stage for six treatments were significantlylower than that before the jointing stage. Both rainfall infiltration depth and infiltration volume underrainfall intensity of60mm/h was significantly greater than that under rainfall intensity of40mm/h atsame mulching treatment and same rainfall duration. Compared those under rainfall intensity of60mm/h, the differences of soil water storage before and after jointing stage was more greater for all sixmulching treatments under rainfall intensity40mm/h.更多还原