基于粒子群算法优化BP神经网络的土壤含水量短期预测模型

为提高土壤含水量预测精度,基于物联网监测数据,提出了粒子群算法(PSO)优化BP神经网络的土壤含水量预测方法。首先应用主成分分析法筛选出影响土壤含水量的关键影响因子,然后构建8-5-1的BP神经网络拓扑结构,应用粒子群算法优化BP神经网络的初始权值和阈值。结果表明:与传统BP神经网络相比,新模型优化了网络结构,避免了陷入局部最优解,具有良好的预测效果;模型的评价指标平均绝对误差、平均绝对百分误差、误差均方根分别为0.259 2、0.010 5和0.135 6,与单一BP神经网络相比,预测精度更高,可满足实际的土壤含水量预测的需要。     

清、洪水灌溉条件下土壤中盐分运移规律的试验研究

在大同县罗庄村玉米地进行洪水灌溉与清水灌溉的大田灌水试验,研究在不同的灌水定额下,用洪水与清水灌溉后,土壤剖面上盐分的运移规律。研究结果表明:用洪水与清水灌溉后,土壤中盐分的运移规律基本相同;盐分随水分运动而向下运移的深度随灌水定额的增大而增大。     

节水农业的形成与灌溉水价改革

我国人均水资源占有量仅为世界人均水资源占有量的1/4,但目前在水资源的使用上存在着惊人的浪费现象,水资源的利用率极低。为解决水资源严重短缺的问题,可行的途径是大力普及各项节水技术,以提高水的利用率,努力建成节水型社会。当务之急是尽快建成节水型农业。而节水农业的形成首先必须依赖于水价制度的改革,利用经济杠杆以提高广大用水者的节水意识。深化水价制度的改革必须走出两个思想误区:其一是误认为农业供水成本远低于工业供水成本,所以农业水价必须远低于工业水价;其二是认为农业是弱质产业,所以农业供水已没有涨价的空间。     

艾比湖流域水资源变化与区域响应

艾比湖流域地处干旱的北疆地区,人类活动中最突出的建设水利设施和土地开发利用,加之全球变化的影响,改变了艾比湖及其水系原生自然生态系统,水资源和水环境发生了很大的变化。在综合研究流域内水资源变化成因的基础上,提出通过开源和节流并举,保证生态用水量,调整产业结构,依靠工程性和非工程性措施响应水资源变化,降低脆弱性,实现流域效益的最大化。     

河北省土壤水资源划区探讨

综述了土壤水和土壤水资源的概念以及土壤水资源划区的意义,以土壤的地带性规律为基础,与农业综合区划相协调并保持乡级行政区界的完整性为划区原则,提出了河北省土壤水资源的划区方案.即河北省土壤水资源分为区、亚区和片3级,共划分为7个土壤水资源区、14个土壤水资源亚区和28个土壤水资源片.     

基于地表水目标水质的畜禽粪便水环境承载力估算

针对南方水网地区畜禽养殖业发展受限问题,以浙江省江山港流域为研究对象,实地调研核定流域水污染物来源途径与入河污染负荷量;基于QUAL2K模型,建立地表水水质响应模型,评估江山港流域在达到目标水质时的畜禽粪便水环境承载力;运用情景分析法探究畜禽养殖与地表水水质响应关系。结果表明：1)2020年江山港流域入河负荷量为190.564 t（以氨氮计）,土壤流失、污水处理厂、化肥施用、有机肥施用和水产养殖分别占总负荷的70.0%、14.8%、11.0%、2.8%和1.4%。2）当水质目标设定为达到Ⅰ类和优于Ⅱ类时,江山港流域能承载畜禽最大养殖量分别为133.5万和1 622.1万头猪当量,当前养殖规模未超载。3）江山港流域地表水水质对畜禽养殖量的响应关系弱,畜禽养殖不是流域水环境保护的制约因素。建议江山港流域加强生态沟渠的建设与管理,协同实现畜禽养殖业发展与水环境保护。     

Does root-sourced ABA play a role for regulation of stomata under drought in quinoa (Chenopodium quinoa Willd.)

The Andean seed crop quinoa (Chenopodium quinoa Willd.) is traditionally grown under drought and other adverse conditions that constrain crop production in the Andes, and it is regarded as having considerable tolerance to soil drying. The objective of this research was to study how chemical and hydraulic signalling from the root system controlled gas exchange in a drying soil in quinoa. It was observed that during soil drying, relative g_s and photosynthesis A_max (drought stressed/fully watered plants) equalled 1, until the fraction of transpirable soil water (FTSW) decreased to 0.82 ± 0.152 and 0.33 ± 0.061, respectively, at bud formation, indicating that photosynthesis was maintained after stomata closure. The relationship between relative g_s and relative A_max at bud formation was represented by a logarithmic function (r² = 0.79), which resulted in a photosynthetic water use efficiency WUE_Amax_/gs$\text{WU}{\text{E}}_{A_{\text{max}}/g_{\text{s}}}$ of 1 when FTSW > 0.8, and increased by 50% with soil drying to FTSW 0.7–0.4. Mild soil drying slightly increased ABA in the xylem. It is concluded that during soil drying, quinoa plants have a sensitive stomatal closure, by which the plants are able to maintain leaf water potential (ψ_l) and A_max, resulting in an increase of WUE. Root originated ABA plays a role in stomata performance during soil drying. ABA regulation seems to be one of the mechanisms utilised by quinoa when facing drought inducing decrease of turgor of stomata guard cells.     

Einfluss von grundwasser und trittbelastung auf den prozess der torfmineralisation und die stoffbilanzen: Influence of ground water and footprints on process of peat mineralization and nutrient balances

Pasture with burdening on peat-bog done four times a year has more positive effects on the water balance than twice grassland use. The saving of water that has been determined by lysimeters evokes no positive effects on the N-, P-, K- or Ca-balance. Due to the lower depth of the plant roots under lowland bog used for pasture more nutrients were leached into the ground water. The soil compression leads to anaerobic conditions. Therefore the phosphorus becomes more mobile and migrates into the outlet. The effect on the peat mineralization and the nutrient balances was not as clear as assumes, because other parameters like the combination of the vegetation formation and the age of the plants also had a strong influence on the factor. The younger plants from the pastures have show higher nutrient contents so that the nutrient export is higher and the balance becomes more negative.     

土壤及泥沙颗粒组成与养分流失的研究

利用自然降雨影响下的农田为试验样地,分析了降雨前后土壤颗粒组成与养分含量的变化,探讨了流失泥沙颗粒组成与养分的变化特征。结果表明,在降雨后,西瓜地与油菜地表土颗粒组成呈现粘粒与粉粒含量减少,砂粒含量相对增加的趋势;土壤养分含量都有不同程度的下降,其中水解氮与速效磷下降的幅度很大,西瓜地与油菜地水解氮降雨后降幅分别为9.67%和7.99%,速效磷的降幅分别为9.20%和8.76%;全氮养分降幅很小,西瓜地与油菜地分别为2.17%和1.54%;流失泥沙细小颗粒含量与养分含量随时间基本都呈降低的变化趋势,并且细小颗粒含量与携带流失的养分含量有显著甚至极显著的正相关性;流失泥沙具有富集粘粒和富集养分的特性,油菜地与西瓜地粘粒富集率分别为1.23和1.20,两样地全氮与速效磷的富集率都在1.3以上,而全磷和水解氮富集率基本处在1.1-1.3之间。