支持向量机在水质评价中的应用

介绍了支持向量机算法的原理，建立了基于支持向量机的水质评价多层次分类检测模型，通过提取水质评价标准中的物质成分为特征参数，利用几个SVM分类器的串联组合，实现对水质的分类和识别，同时，引入类权重因子，解决训练样本类别数量不平衡而导致的错分问题，实验结果显示该方法提高了水质评价的准确性和效率。     

Effect of water deficit in different growth stages on stem sap flux of greenhouse grown pear-jujube tree

The effects of water deficit in different fruit growth stages on the variation of stem sap flux of 6-year old greenhouse-grown pear-jujube trees were investigated. Treatments included sufficient water supply during the whole fruit-growing period (T₁), mild water deficit during the flowering–fruit setting stage (T₂), moderate water deficit during the fruit rapid growth stage (T₃) and severe water deficit during the fruit maturing stage (T₄). Results showed that significant compensation effect on stem sap flux after re-watering was observed in T₂, but not in T₃ and T₄ stages. At the end of rapid growth stage, the diurnal variation of stomatal conductance generally had a similar trend as that of stem sap flux, but with a distinct midday depression from 12:00 to 14:00 p.m. In addition, a linear relationship between the relative available soil water content (RAWC) and the ratio of daily stem sap flux to that of sufficient water treatment was observed (R² = 0.4489).     

水肥一体化智能调控设备智能液位开关设计

在对设施蔬菜生产中使用的水肥一体化设备进行智能灌溉控制设计中,监测混肥罐和回液桶中液位高低是非常必要的。现有液位检测普遍采用机械开关式液位浮子,工程实践中发现当进出水口与浮子过近,或者水纹波动显著时,浮子上下剧烈抖动会导致水泵随浮子的抖动而经常启停,对于启动电流较大的泵容易造成损伤。本文在软硬件设计中采用了计数器数字滤波电路以及单片机定时器中断,设计了一种替代简单液位浮子的实用智能液位开关。为了检验该设计的性能,在同等条件下,进行了不同液面、不同压力和水流量的对比试验。试验结果表明,计数脉冲发生电路中充电电阻和放电电阻为4.7 kΩ时,滤波后的输出信号比现场实际的数控信号时间迟延了2.11 s,经数字电路滤波后的输出信号已经完全消除了宽度小于2 s的所有干扰信号,并且具有很高的稳定性。迟延时间与计数电路、计数脉冲发生电路有关,通过电路设计调整滤波脉宽可使输出信号有效地滤除干扰成分。同时,利用该方法也可以根据不同工况,如不同功率的水泵设定不同参数滤除干扰信号,提高智能液位开关适用性。     

Can we maintain turf to customers’ satisfaction with less water?

Science-based, holistic, site-specific water conservation practices can reduce water use on turfgrass sites without adversely affecting turfgrass performance. However, when water use is decreased below a certain threshold, performance declines. Water conservation measures that reduce turfgrass performance essentially decrease its economic, environmental, recreational, and aesthetic values, which can in turn adversely impact many ‘stakeholders’, including the local economy and those affected by increased wind erosion, water erosion, or fire hazard. On larger turfgrass sites, considerable costs are associated with some water conservation strategies, especially when the quality of an alternative irrigation water source is poor or redesign of the landscape and/or irrigation system is involved.     

An integrated modelling toolbox for water resources assessment and management in highland catchments: Model description

Water and land resource competition and environmental degradation pose difficult questions for resource managers. In particular, the ensuing trade-offs between economic, environmental, and social factors and their spatiotemporal variability must be considered when implementing management policies. This paper describes an integrated modelling toolbox that has been developed for highland catchments – specifically the Mae Chaem catchment in Northern Thailand. This toolbox contains models of crop growth, erosion and rainfall-runoff, as well as household decision and socioeconomic impact models. The approach described advances and complements previous approaches by: considering more complex interactions between land-use decisions and the hydrological cycle; modelling household decisions based on uncertain expectations; and assessing impacts of changes not only on flows and household income, but also on subsistence production and erosion. An example of the types of trade-offs and scenarios that can be assessed using the integrated modelling toolbox is also presented. This demonstrates that for the scenarios presented, the magnitude and direction of impacts simulated by the model is not dependent on climate. Further testing of the model is demonstrated in a companion paper. Overall, the plausibility of the model is shown.     

A tool for community-based assessment of the implications of development on water security in hillside watersheds

Development and population growth in Latin American countries with steep slope farming are likely to further increase pressures on water and land resources. A methodology was developed for assessing water availability and use under different development pathways at a watershed scale to determine whether water security is a potential problem, and if so, under what conditions it is likely to occur. This methodology makes use of a GIS-based spatial water budget model for simulating stream water availability, water use and stream flow control on a daily basis at a watershed scale. Here, we analysed water availability under three plausible development scenarios for the 3246 ha Cabuyal River watershed in southwest Colombia in the year 2025: Corporate Farming (CF), Ecological Watershed (EW), and Business as Usual (BU). Simulated average river flows at the watershed outlet were, respectively, 874, 796 and 925 l s⁻¹ for the CF, EW and BU scenarios. The contribution of base flow to river flow (base flow index) was on average, 80.8, 85.6 and 77.9%, respectively, for the three scenarios. The watershed had the potential to meet the anticipated increase in water use under each explorative scenario. However, dams were necessary to store irrigation water in the CF scenario, otherwise over 60% of the available water would have been used during the dry season. Such a high figure raises concerns about effects on aquatic and riparian ecology, concentrations of potential contaminants, water reserves for especially low rainfall years, and the watershed resilience to meet temporarily higher water needs during the day. Analyses indicated that current water-use conflicts in the watershed can be resolved if irrigation water supply is separated from drinking water supply. This study helped reduce some of the complexity associated with the interdependencies between land and water resources, the impact of using them, and spatial linkages within the watershed. Results of this study can be used for teaching local stakeholders about basic landscape responses and helping multi-institutional alliances to become proactive and to guide development to the benefit of local communities.     

节水灌溉实时决策系统

为寻求利用有限水资源取得最高作物产量或最大效益的灌溉方案,应根据不同供给水源的时空分布,作物品种与生长期和农田水分状况进行综合分析,系统分析方法的线性或非线性规划方法可望得到最优结果。     

河北二季作区马铃薯节水灌溉研究初报

为研究河北二季作区马铃薯整个生育期不同灌水量对其生长发育的影响,以该地主栽品种为材料,根据不同生育期需水规律,设定5个不同水量梯度的滴灌处理和1个常规沟灌处理(CK);调查分析了不同处理间物候期、重要农艺性状、病害发生情况和产量等相关指标。结果表明,处理3(采用滴灌模式,水量为110 m3)与常规沟灌相比,马铃薯生长势、总产量和商品薯产量无显著差异,生育期提前,病害较轻,商品薯率提高7.2%,节水35.41%,综合表现优于其他处理,为试验地区最佳灌水量。     

Root exudation of prometryn and its metabolites from Chinese celery (Apium graveolens)

Root exudates from Chinese celery (Apium graveolens) and Chinese cabbage (pak choi, Brassica chinensis) plants treated by prometryn, an herbicide, were qualitatively and quantitatively investigated and compared under hydroponic cultivation. Prometryn and its metabolites released into the nutrient solution were analyzed by ultra-performance liquid chromatograph coupled with orbitrap mass spectrometer to investigate whether this xylem-mobile herbicide is exuded from the roots. The results showed that celery and pak choi had different root exudation profiles. Celery metabolized prometryn to prometryn sulfoxide and released both compounds from the roots. In contrast, pak choi barely metabolized or actively released prometryn from the roots. The concentration of prometryn sulfoxide released from celery after 96 hr was 21 µg/L, which was nearly one-third that of released prometryn. Our results indicate that the root exudation and translocation of xylem-mobile herbicides could be significant in plants and are highly species dependent compared with phloem-mobile herbicides.     

Gene Discovery and Characterization from Non-sequenced Medicinal Plants

Plant-derived secondary metabolites still constitute important resources for currently prescribed drugs and for the discovery of active principles which are effective in new indication areas. Promising compounds are often exclusively found in wild medicinal plants and the classical plant-based pharmaceutical production approach has been hampered by low yields, lack of reproducibility and the high complexity of plant extracts （Rischer and Oksman-Caldentey, 2005）. Influencing the accumulation of target compounds by breeding or genetic engineering is greatly impeded because in most cases the general biosynthetic routes are incompletely understood.