茂名市农业环境质量常规性调查监测与评价

为向今后农业生产和农业环境保护提供科学依据,开展了农业环境质量长期定点监测调查工作,介绍了监测布点原则和方法,抽采样和监测项目及选用标准。农业环境质量监测结果显示,土壤监测点超标率为15.71%,灌溉水监测点超标率为14.3%,空气监测点各监测项目全部达标,总体上看茂名市农业环境质量问题不容忽视。在此基础上提出了对策与建议。     

RBF神经网络在线学习方式及其在大坝位移监测中的应用

针对目前RBF神经网络训练算法存在的问题,提出了一种模拟人类学习方式的自动调整隐层节点数的在线训练方法,对其理论依据进行了分析,并用实例对其进行了验证。结果表明,此种学习方法速度快、拟合精度高、新旧知识均可记忆,克服了以往算法的不足,具有很大的实用性。     

Organic input quality is more important than its quantity: C turnover coefficients in different cropping systems

Annual C input to soil is a major factor affecting soil organic carbon (SOC) dynamics. However different types of C-sources can have different behaviour, in relation to their chemical characteristics and how they interact with soil. Root-derived C, in particular, should be more efficient than other organic materials as a result of the physicochemical and biological characteristics of the surrounding environment, leading to a reduction in the C decomposition rate.To test this hypothesis, we considered a long-term experiment underway in Northern Italy since 1962, comparing permanent meadow and 6 different crop rotations over a wide range of nutrient inputs, in both organic and inorganic forms. C inputs from amendments were measured and those from crops were calculated using allometric functions and crop and residues yields. The time evolution of SOC was studied through a single-pool, first-order kinetic model, allowing the estimation of humification coefficients for residues, roots, farmyard manure and cattle slurries.The highest value of the humification coefficient was estimated for farmyard manure, which confirmed its high efficiency in stabilising SOC content. Root C presented a humification coefficient 1.9 times higher than above-ground plant materials while slurries were intermediate, with a humification coefficient roughly half that of farmyard manure and even lower that of roots.The quality of C input thus seems of fundamental importance for evaluating the sustainability of different cropping systems in terms of SOC dynamics.     

The positive relationship between soil quality and crop production: A case study on the effect of farm compost application

In order to ensure sustainable agriculture, and for evaluating the effects of management practices on soil processes, tools for assessing soil quality are required. The development and use of a multiparameter index, which includes a wide range of soil properties, have been tested and found useful by several studies. However, soil quality measurements are ‘stand-alone’ tools unless they are either linked to important soil functions, used to characterize (agro)ecosystems or used to predict sustainability or productivity. In our study, the relationship between crop production and soil quality was assessed in a six year old field experiment studying the effect of farm compost (FC) amendment in a crop rotation of potato, fodder beet, forage maize and Brussels sprouts. To justify the hypothesis that repeated FC amendment results in both improved soil quality and consequently higher crop yields, a wide range of chemical, biological and physical soil properties were measured and integrated into a soil quality index (SQI). Next, crop yields were used as a functional goal to verify the causal relationship between SQI and crop production. Our results showed that there were significant changes in chemical, physical and biological soil quality as a result of repeated FC amendment. This was evidenced for example by a remarkable increase in both soil organic carbon (SOC) and total N content. Microbial biomass, the relative amount of bacterivorous nematodes and earthworm number were significantly increased as well and, together with SOC and total N, indicated as the dominant factors in assessing soil quality. The integration of these key indicators into the SQI revealed higher SQI values when FC was applied. In addition, crop yields were increased in all FC treated plots by which SOC was pointed out as the most important indicator influencing crop production. Finally, a causal relationship was observed between soil quality and the yield of potato and fodder beet. We conclude that our SQI may be a promising and useful tool to compare different (soil) management practices in relation to a strategic, regional goal, e.g., sustainable high yields. Before generalizing, we recommend a thorough validation of our SQI in other long-term field experiments.     

农业技术和气候变化对农作物产量和蒸散量的影响

随着农业生产条件的改善、品种改进和有利的气象条件的变化, 世界各地的作物产量得到大幅度提高, 但作物的蒸散量却未出现大幅度提高。本文以石家庄气象站1955~2007 年的气象资料为基础, 分析了河北省冬小麦和夏玉米生长期间主要气象因素变化, 结合中国科学院栾城农业生态系统试验站长期定位灌溉试验的研究结果, 分析了农业生产条件和气象因子变化对冬小麦和夏玉米产量及耗水量的影响。结果表明,1955~2007 年冬小麦和夏玉米生长季的气象因子发生了变化, 日照时数、相对湿度、风速、气温日较差显著降低, 最低气温、平均气温和积温显著升高, 气象因子的变化对作物总蒸散量未产生明显影响, 但由于降水减少,作物生长期间的灌溉需水量呈增加趋势。长期灌溉试验结果表明, 随着农业生产条件的变化和品种的改良, 冬小麦和夏玉米的产量不断增加, 而耗水量的增加幅度小于产量增加幅度, 夏玉米的耗水量呈稳定状态。节水技术的推广和应用对维持耗水量稳定起着非常关键的作用。     

华北平原冬小麦总初级生产力的遥感监测

该文以华北平原冬小麦为研究区,验证MODIS-GPP产品数据在下垫面复杂的农田生态系统的适用性,及农田生态系统的碳源/汇功能。首先采用MODIS-NDVI产品数据,根据冬小麦物候特征,解译出冬小麦种植面积;然后与经过验证并校正后的MODIS-GPP产品数据估算华北平原冬小麦2010年生长季内的GPP总量;最后,由涡度相关实测数据估算出冬小麦的生态系统呼吸量（RE）和净生态系统生产力(NEP)。结果显示：MODIS-NDVI解译出的冬小麦面积结果与TM解译结果的用户精度高达90.6%,总体精度达87.2%;MODIS-GPP产品数据与涡度相关数据估算结果的复相关系数高达0.9227;2010年华北平原冬小麦生长季GPP约为6.24′108 kgC,约占该区所有植被GPP的21.3%;华北平原冬小麦固碳能力较强,净生态系统生产力（NEP）占GPP的58.7%。研究表明,MODIS-GPP产品数据满足华北平原冬小麦GPP统计精度需求,冬小麦生态系统对陆地生态系统起到了重要的碳汇功能。     

便携式作物生长监测诊断仪的设计与试验

为了解决当前作物生长信息实时、快速、无损测量手段瓶颈问题,研发了一种基于多光谱传感器的便携式作物生长监测诊断仪。该仪器由多光谱传感器系统、处理器系统及附属机构组成,能实时无损地获取作物叶层氮含量、叶层氮积累量、叶面积指数、生物量等主要生长指标。对水稻的试验结果表明,便携式作物生长监测诊断仪测得的DVI值与水稻叶层氮含量、氮积累量、叶面积指数、叶干重的决定系数R2分别为0.711,0.8231,0.7698,0.7212。该仪器结构简单,集成度高,性价比好,携带方便,易于田间操作。     

农用地等级监测指标体系与方法

通过农用地等级质量监测,实时获取农用地等级质量信息是实现农用地可持续开发利用和耕地占补平衡的基础。在已有农用地分等定级指标体系成果的基础上,通过指标易变性分析选出引起农用地等级质量变化的易变性指标,进一步分析导致易变性指标变化的驱动力,并结合当前相关指标或驱动力因子的监测技术现状,筛选出了能够直接或间接监测并评价农用地等级质量的监测指标,建立了农用地等级质量监测指标体系,并针对各种指标提出了相应的监测方法。分析结果表明,由于充分考虑了驱动力的区域差异性,构建的监测指标体系也具有很强的区域性,相应的监测方法更为实用。     

夏玉米田作物蒸腾与棵间土壤蒸发模拟计算方法研究

本文依据能量平衡原理和质量守恒定律，提出了作物蒸腾与棵间土壤蒸发的估算模式，并对参数的估算做了详细的分析和讨论。最后利用试验区实测资料，通过田间水热耦合运移的数值模拟，对夏玉米田作物蒸腾棵间土壤蒸发模型及其模拟计算方法进行了验证，结果表明，本文提出的模型和模拟计算方法是有效而实用的。     

基于MPEG-4的网络视频监控系统拥塞控制方法研究

有效可靠的的传输,是视频监控系统的一个重要研究课题。分析了基于MPEG-4的网络视频监控系统可以采用的拥塞控制方法,提出了一种基于RTP协议,从分组丢失率、时延、时延抖动三方面判断网络拥塞,综合考虑发送、接收和中间节点三方面影响的拥塞控制方法。