摘要：【中文】本发明公开了一种适用于多要素的高分辨率集合预报系统，包括数据采集模块、空间降尺度模块、时间插值模块、时间拆分模块、集合预报模块，所述数据采集模块采集集合预报成员以及订正的网格实况数据；所述空间降尺度模块使用空间降尺度插值方法；所述时间插值模块采用内插法处理除降水外的其他气象要素；所述时间拆分模块的处理方法采用时间拆分法；所述集合预报模块在不同气象要素的误差特征下集成了多个子模块。本发明进行降水时间拆分和其他要素时间插值时采用更为合理的方法，提高预报精细程度和预报的准确性；充分考虑到不同要素的误差特征，构建出多个集合预报子模块，针对不同要素采用不同集合预报算法，适用于多个气象要素的集合预报系统。 【EN】The invention discloses a high-resolution ensemble prediction system suitable for multiple elements, which comprises a data acquisition module, a spatial downscaling module, a time interpolation module, a time splitting module and an ensemble prediction module, wherein the data acquisition module acquires ensemble prediction members and corrected grid live data; the space downscaling module uses a space downscaling interpolation method; the time interpolation module adopts an interpolation method to process other meteorological elements except precipitation; the processing method of the time splitting module adopts a time splitting method; the ensemble forecasting module integrates a plurality of sub-modules under the error characteristics of different meteorological elements. The method adopts a more reasonable method when the precipitation time is split and other element time interpolation is carried out, so that the forecasting fineness and the forecasting accuracy are improved; and (3) fully considering the error characteristics of different elements, constructing a plurality of ensemble forecasting submodules, adopting different ensemble forecasting algorithms aiming at different elements, and being suitable for an ensemble forecasting system of a plurality of meteorological elements.

摘要：【中文】本发明公开了空气质量预报预警系统，包括空气质量监测分析、大气污染条件分析、大气污染预报预警，所述空气质量监测分析对污染物的实况数据进行采集，通过NB‑IoT网络传输到系统的数据仓库，将实况数据进行多维度的统计分析；所述大气污染条件分析获取气象实测数据，再将污染物预报以GIS色斑图的形式展示，以要素变化对污染物影响进行预报产品综合分析；所述大气污染预报预警将采集到的污染物实况数据针对不同的站点设置污染物数值提醒阈值，超过阈值后以图像高亮、声音提醒。本发明能够克服环境气象预报的预报有效时长不足，而且能够通过多种模式预报更加精准的预报区域范围内的污染物变化趋势，从而为环保业务部门提前预报、准确预报提供了有力保障。 【EN】The invention discloses an air quality forecasting and early warning system, which comprises air quality monitoring analysis, atmospheric pollution condition analysis and atmospheric pollution forecasting and early warning, wherein the air quality monitoring analysis is used for acquiring live data of pollutants, transmitting the live data to a data warehouse of the system through an NB-IoT network and carrying out multi-dimensional statistical analysis on the live data; analyzing the atmospheric pollution condition to obtain meteorological actual measurement data, displaying the pollutant forecast in a GIS (geographic information System) color spot diagram form, and performing forecast product comprehensive analysis on the influence of element change on the pollutant; the atmospheric pollution forecast early warning sets pollutant numerical reminding threshold values for collected pollutant live data aiming at different sites, and the pollutants are reminded by image highlighting and sound after the pollutant numerical reminding threshold values are exceeded. The invention can overcome the shortage of effective forecasting duration of environmental weather forecasting, and can forecast the pollutant variation trend in a more accurate forecasting area range through various modes, thereby providing powerful guarantee for advanced forecasting and accurate forecasting of environmental protection business departments.

摘要：【中文】一种硫化改性纳米零价铁的制备及在处理抗生素抗性基因中的应用，属于纳米材料及污水处理领域。本发明使用亚硫酸钠作为脱氧剂，使用脱氧后的亚硫酸钠溶液配制硫酸亚铁、硼氢化钾及硫化钠，得到相应脱氧溶液；使用硫化钠作为硫化改性剂，将硫化钠溶液与硼氢化钾溶液按比例混合，逐滴滴加硫酸亚铁溶液，得到硫化改性的纳米零价铁，使用去离子水洗涤。对多批次的实际水厂二级出水中总细菌与14种抗生素抗性基因均有着高于99.9％的去除效果，且能有效的抑制总细菌及抗生素抗性基因的再生，大大削减了污水处理厂二级出水中基因型污染物对受纳水体的生态威胁。 【EN】Preparation of sulfuration modified nano zero-valent iron and application of sulfuration modified nano zero-valent iron in treatment of antibiotic resistance genes, belonging to the field of nano materials and sewage treatment. The invention uses sodium sulfite as a deoxidizer, and uses the sodium sulfite solution after deoxidation to prepare ferrous sulfate, potassium borohydride and sodium sulfide to obtain corresponding deoxidation solution; sodium sulfide is used as a vulcanization modifier, a sodium sulfide solution and a potassium borohydride solution are mixed in proportion, a ferrous sulfate solution is dropwise added to obtain vulcanization modified nano zero-valent iron, and the nano zero-valent iron is washed by deionized water. The removal effect of more than 99.9 percent on total bacteria and 14 antibiotic resistance genes in secondary effluent of multiple batches of actual water plants is achieved, the regeneration of the total bacteria and the antibiotic resistance genes can be effectively inhibited, and the ecological threat of genotype pollutants in the secondary effluent of a sewage treatment plant to a receiving water body is greatly reduced.

摘要：【中文】本发明涉及气象监测技术领域，尤其涉及一种基于ResNet101的天气类型识别方法，包括在预处理过程，首先将图像采集系统采集到的需要训练的图像的数量按照6：2：2的比例划分成训练集、验证集和测试集；然后对训练集的图像信息，预处理系统在读入图像全部像素点数据后进行缩放、平移、翻转、剪切和旋转操作；再采用ResNet101模拟训练系统提取预处理系统处理后的图像特征信息，并根据特征进行分类任务训练，形成图像识别系统。本发明能够识别更加复杂的天气类型，包括但不仅限于雨凇、雾凇、雾霾、霜、露、结冰、降雨、降雪、冰雹等，且具有很强的可扩展性；能够从内容复杂的图像信息中准确提取关键的天气现象特征，对图像信息没有特定的要求，适应性强。 【EN】The invention relates to the technical field of meteorological monitoring, in particular to a weather type identification method based on ResNet101, which comprises the following steps that in the preprocessing process, firstly, the number of images which are acquired by an image acquisition system and need to be trained is 6: 2: 2, dividing the ratio into a training set, a verification set and a test set; then, for the image information of the training set, the preprocessing system performs operations of zooming, translation, turning, shearing and rotation after reading in all pixel point data of the image; and extracting image characteristic information processed by the preprocessing system by adopting a ResNet101 simulation training system, and performing classification task training according to the characteristics to form an image recognition system. The method can identify more complicated weather types including but not limited to rime, haze, frost, dew, ice, rainfall, snowfall, hail and the like, and has strong expandability; the method can accurately extract key weather phenomenon characteristics from the image information with complex content, has no specific requirements on the image information, and has strong adaptability.