Changes in nitrogen budget and potential risk to the environment over 20 years (1990–2010) in the agroecosystems of the Haihe Basin,China

The nitrogen balance can serve as an indicator of the risk to the environment of nitrogen loss from agricultural land. To investigate the temporal and spatial changes in agricultural nitrogen application and its potential threat to the environment of the Haihe Basin in China, we used a database of county-level agricultural statistics to calculate agricultural nitrogen input, output, surplus intensity, and use efficiency. Chemical fertilizer nitrogen input increased by 51.7% from 1990 to 2000 and by 37.2% from 2000 to 2010, concomitant with increasing crop yields. Simultaneously, the nitrogen surplus intensity increased by 53.5% from 1990 to 2000 and by 16.5% from 2000 to 2010, presenting a continuously increased environmental risk. Nitrogen use efficiency decreased from 0.46 in 1990 to 0.42 in 2000 and remained constant at 0.42 in 2010, partly due to fertilizer composition and type improvement. This level indicates that more than half of nitrogen inputs are lost in agroecosystems. Our results suggest that although the improvement in fertilizer composition and types has partially offset the decrease in nitrogen use efficiency, the environmental risk has still increased gradually over the past 20 years, along with the increase in crop yields and nitrogen application. It is important to achieve a better nitrogen balance through more effective management to significantly reduce the environmental risk, decrease nitrogen surplus intensity, and increase nitrogen use efficiency without sacrificing crop yields.     

Towards a new climate economics:research areas and prospects

International cooperation to address climate change now stands at a crossroads.With a new international regime for emissions reduction established by the Durban Platform, "New Climate Economics(NCE)" has become a research hotspot.The need for urgent action to combat climate change has prompted discussion on reforms of economic growth patterns and the energy system.The industrial civilization,therefore,now faces a transition towards a new pathway for ecological sustainability.NCE explores new economic concepts,theories,and analytical methods to design a balanced pathway for sustainable growth and emission reduction.Instead of getting trapped in discussions on allocation of emission reduction responsibilities and obligations among countries,NCE pays more attention to developing win-win multilateral cooperation mechanisms that facilitate collaborative RD and knowledge sharing.In addition,NCE studies incentives for low-carbon transition,turning carbon emission reduction into a domestic need for countries to increase their international competitiveness.To achieve the 2°C target,most countries around the world face challenges of insufficient emission allowances to cover expected emissions associated with their projected economic growth.As carbon emissions rights becomes an increasingly scarce resource,increasing the carbon productivity of the economy turns to be the critical path to address the dilemma of green or growth.NCE studies the historical evolution of carbon productivity for countries at different development stages as well as ways to enhance such carbon productivity.This type of study provides invaluable lessons for emerging economies to reach their own emission peaks without losing the momentum of growth.Replacing fossil fuels with new and renewable energy has proven to be an inevitable choice for reshaping the energy system and addressing climate change- it has already become a global trend.NCE studies incentives for new energy technology innovation and deployment provided by carbon pricing,and sheds light on the co-benefits of climate change mitigation,such as resource conservation,environmental protection,and energy security.The role of carbon pricing in promoting intemational RD cooperation and technical transfer will also be studied.The shift in consumption patterns is another key factor enabling a low-carbon transformation.Therefore,NCE also explores the theoretical work on new values of wealth,welfare and consumption,new lifestyles in the context of ecological civilization,concepts and implementation of low-carbon urban planning in developing countries,and the impacts of consumption pattern changes on social development,material production,and urban infrastructure construction.     

基于BERT-BILSTM-CRF模型的电力行业事故文本智能分析*

为解决电力行业事故报告文本较长、语义复杂,难以进行有效文本识别问题,提出1种以BERT作为底层的预训练模型,并设计1种双重注意力机制编码器,结合BILSTM-CRF深度挖掘事故文本语义特征,从而实现文本智能分析。首先构建电力词典,通过对BERT预训练,进行BIO标注,然后引入BILSTM-CRF模型实现对文本标签智能分类,最后将该模型与现行其他4种深度学习模型进行对比。研究结果表明:该模型智能识别精确率、召回率及F1值(查准率)均达到约97%,较其他4种模型中效果最好的模型分别提高0.02,0.03,0.02。研究结果可为电力行业事故报告文本分析提供1种新思路。     

烟花爆竹引发重大火灾

沈阳皇朝万鑫国际大厦火灾 事件背景： 2011年2月3日零时13分,兔年的钟声刚刚敲响．位于沈阳市和平区青年大街390号的皇朝万鑫酒店发生火灾,火灾因燃放烟花引燃酒店B座楼体外墙表面装饰材料所致,过火面积约1万余平方米。但由于建筑物太高,     

太原市污灌区农田土壤中多环芳烃污染特征及生态风险评价

通过对污灌区农田土壤多环芳烃(Polycyclic Aromatic Hydrocarbons,PAHs)的分布特征、污染程度及来源进行解析,深入了解污水灌溉引发的土壤污染问题,实现污水灌溉农田土壤的污染预警和科学合理利用.在太原污灌农田共采集110个土壤样品,使用气相色谱-质谱仪(Gas Chromatography...     

基于“人工智能”驱动的农村生活污水系统治理专栏序言

伴随流域水污染控制与治理工作的深入,农村分散生活污水治理逐渐成为制约流域水环境改善的关键.同时,农村生活污水治理是人居环境提升、美丽乡村建设以及乡村振兴战略实施的集中体现,事关广大农村居民的切身利益.本专栏是国家水专项“分散污水治理课题”团队近年来科研成果的总结,从污水治理的全生命周期角度出发,尝试通过人工智能赋能治理模式决策、“菜单式”工艺组合、设施运维与监管、长效管理等4个方面对研究和实践进行了梳理总结,构建基于人工智能驱动的农村生活污水治理决策支持系统,以期为北运河流域以及其他区域农村生活污水治理提供科技支撑.     

未来RCP4.5情景下黄海DMS浓度变化模拟与分析

使用CORDEX-EA过去气候态(2000-2009年)与RCP4.5情景下近未来气候态(2041-2050年)大气强迫结果驱动中国东部陆架海域耦合DMS模块生态模型,模拟了黄海过去及近未来表层DMS浓度(C_DMS),探究了黄海近未来C_DMS时空分布的变化及其影响因素.结果表明:近未来黄海C_DMS的年循环发生变化,北黄海C_DMS极高值出现月份由5、9月转变为4、10月,南黄海由4、9月转变为4、8月;局部C_DMS高值区也发生变化,春季山东半岛附近海域、夏季苏北浅滩、南黄海中东部、秋季南黄海东部C_DMS高值区加强,夏季山东半岛附近C_DMS高值区减弱.近未来热通量、风应力对山东半岛、南黄海中东部海域C_DMS影响较大;降水量、云量对西朝鲜湾C_DMS的影响占优;苏北浅滩C_DMS受多个气候因子共同作用.     

三峡水库水动力分区及总磷标准研究

采用EFDC模型计算2010~2018年三峡水库流域水龄.结果表明,三峡水库水龄由上游库尾至下游库首愈来愈大,尤其坝前水龄高达100余天,湖泊属性增强,坝前支流受库区影响大,大坝调度对水龄影响较大,高水位运行期水龄将增长至2~4倍.基于水龄空间分布将三峡水库及其17条主要支流划分为河流型、河湖过渡型和湖泊型.水龄影响因素定量分析表明,水龄与水位呈正相关,与流量呈负相关,泄水期变化速率较蓄水期明显,单位流量与水位内水龄变化幅度高达6~7d,变化区间介于30~100d之间.最后,提出基于水龄权重法的河湖过渡区总磷评价标准,该标准可揭示库区水体总磷的时空变化特征.     

Grazing greatly reduces the temporal stability of soil cellulolytic fungal community in a steppe on the Tibetan Plateau

Excessive livestock grazing degrades grasslands ecosystem stability and sustainability by reducing soil organic matter and plant productivity. However, the effects of grazing on soil cellulolytic fungi, an important indicator of the degradation process for soil organic matter, remain less well understood. Using T-RFLP and sequencing methods, we investigated the effects of grazing on the temporal changes of cellulolytic fungal abundance and community structure in dry steppe soils during the growing months from May to September, on the Tibetan Plateau using T-RFLP and sequencing methods. The results demonstrated that the abundance of soil cellulolytic fungi under grazing treatment changed significantly from month to month, and was positively correlated with dissolved organic carbon (DOC) and soil temperature, but negatively correlated with soil pH. Contrastingly, cellulolytic fungal abundance did not change within the fencing treatment (ungrazed conditions). Cellulolytic fungal community structure changed significantly in the growing months in grazed soils, but did not change in fenced soils. Grazing played a key role in determining the community structure of soil cellulolytic fungi by explaining 8.1% of the variation, while pH and DOC explained 4.1% and 4.0%, respectively. Phylogenetically, the cellulolytic fungi were primarily affiliated with Ascomycota (69.65% in relative abundance) and Basidiomycota (30.35%). Therefore, grazing substantially reduced the stability of soil cellulolytic fungal abundance and community structure, as compared with the fencing treatment. Our finding provides a new insight into the responses of organic matter-decomposing microbes for grassland managements.     

机器学习在材料服役性能预测中的应用

针对材料服役性能预测存在误差大、计算复杂、适用性差等问题,提出了基于数据挖掘的机器学习预测方法。首先阐述了机器学习的应用流程,并总结了常用模型原理及其在材料性能预测中的应用。然后采用多种机器学习模型对RPV钢的辐照性能进行预测,并通过Stacking集成方法提高了模型的预测精度。结果表明,机器学习可用于材料服役性能预测,具有较高的预测精度和可靠性。根据材料服役数据的不同特征选择合适的学习模型,同时进行模型融合和参数优化,可有效提高模型的预测精度及运算速度。