时间：2022年10月28日（星期五）14:10-16:30

会议号：腾讯会议 921 428 531

主办：华中农业大学

承办：资源与环境学院    

报告人1：Dr. Muhammad Asif 14:10-14:45

题目：层状结构纳米材料在电化学传感中的应用

Presentation:Layered nanoarchitectures in electrochemical sensing applications

摘要：层状双氢氧化物（LDHs）材料属于阴离子层状粘土，也称为类水滑石化合物。层状结构的LDH材料具有良好的吸附和插层性能、高电荷密度和阴离子交换能力，极具应用前景。然而，较低的电导率限制了LDHs基材料在氧化还原反应中的催化性能。将LDHs材料与一些导电聚合物或者导电碳材料（如石墨烯氧化物、碳纳米管（CNT）和MXenes）集成有望实现其性能的优化。我们聚焦于将LDHs材料作为电化学传感应用于环境污染物和小分子代谢物的检测，对准确监测环境中的有害分子和代谢紊乱具有重要意义。通过快速检测过氧化氢、硝酸盐和硫化氢等小分子物质为环境及临床分析提供理论支持。过氧化氢分子可以作为早期疾病诊断的生物标记物，硝酸盐和H₂S被用作环境样品细菌检测的标记物。目前，传统的仪器分析耗时且复杂，无法用于活样本分析。本次报告将概述我们合成LDH杂化纳米结构的各种策略及其作为检测环境污染物和小生物分子的电化学传感器的应用。

Layered double hydroxides (LDHs) materials belong to family of anionic lamellar clays, also known as hydrotalcite-like compounds. LDHs are promising materials with layered structure which have some outstanding intrinsic features such as admirable adsorption and intercalation behaviours, high charge density and anion exchangeabilities. Although great efforts have been devoted for the advancement in LDHs based materials but they still suffer from low conductivity that limits their efficacy for oxidation reduction reactions. The integration of LDHs material with some conductive polymers and conductive carbon materials like graphene oxides, carbon nanotubes (CNTs) and MXenes can be the possible solution to accomplish an optimized electrochemical performance. We are now focusing more on electrochemical sensing of environmental pollutants (resorcinol, catechol, etc…) as well as small molecules metabolites which are of great significance for accurate monitoring of hazardous molecules in environment and metabolic disorders respectively. Rapid detection of small molecules such as hydrogen peroxide, nitrates and H₂S can provide theoretical support for environmental and clinical analysis. These biomolecules can act as biomarkers for early diseases diagnosis. Moreover, nitrates and H₂S are used as markers in bacterial detection from environmental samples. At present, traditional instrumental analysis is time consuming and complex, and cannot be used for live sample analysis. This talk will, therefore, be an overview of our various strategies to synthesize LDHs hybrid nanostructures and their applications as electrochemical sensors for the detection of environmental pollutants and small biomolecules.

报告人2：蔡耀明 14:45-15:20  

题目：面向高光谱遥感图像信息提取的无监督深度学习方法

Presentation: Unsupervised Deep Learning for Hyperspectral Remote Sensing Image Processing

摘要：高光谱遥感是当前遥感科技发展的一个前沿领域，它能同时获取地物空间结构信息与反映地物材质属性的精细光谱信息，具有“图谱合一”的特点，被广泛应用于精准农业、地质勘探、环境监测、军事侦察等重要领域。高光谱图像信息提取是高光谱遥感应用的重要基础，随着以深度学习为代表的人工智能技术的快速发展，高光谱图像中标记样本获取困难的问题也日益凸显。本报告在介绍高光谱成像与深度学习关键技术的基础上，讨论了当前高光谱图像智能解译中的若干挑战，并针对其中的无监督降维、无监督聚类和模型优化等核心任务，介绍了报告人在此方面的一系列研究成果，最后开放性地讨论了未来潜在的研究方向。

Hyperspectral remote sensing is a frontier field in the development of current remote sensing science and technology, which can simultaneously capture rich spatial structure information and spectral information reflecting the material properties of land covers. Hyperspectral remote sensing technology has been widely used in various fields including precision agriculture, geological exploration, environmental monitoring, military reconnaissance, etc. Information extraction of hyperspectral images is the foundation for applications of hyperspectral images in practice. With the rapid development of artificial intelligence technology, especially deep learning technology, the availability of labeled samples in hyperspectral images has become a bottleneck problem restricting the development of information extraction. In this presentation, we first briefly introduce key technologies in the context of hyperspectral imaging and deep learning, followed by a discussion of challenges in the current intelligent interpretation of hyperspectral images. Then, we introduce a series of our recent works for unsupervised learning, including unsupervised dimensionality reduction, unsupervised classification, and model optimization. Finally, we discuss some potential directions for future research.

Presentation: Application of remote sensing technology on biomass burning and air pollution monitoring

报告人3：殷帅 15:20-15:55

题目：遥感在生物质燃烧与大气污染监测方面的应用

Application of remote sensing technology on biomass burning and air pollution monitoring

摘要：随着遥感技术的发展与完善，地理空间观测为农业、林业、水利、海洋、生态环境等领域的研究提供了大量不可或缺的基础数据。同时，这些遥感数据与地面观测网络相结合为缓解大气污染与全球温暖化的相关政策制定提供了充分的理论依据。我将结合我的研究方向与过去的研究成果从以下三个实例介绍遥感及空间数据在生物质燃烧和大气污染监测方面的应用。

1. 中国秸秆焚烧的时空分布格局与变化趋势。由于不同的作物耕作制度，秸秆焚烧在我国各个地区呈现不同的空间特征与季节性，例如东部地区以夏季（6月份）焚烧为主，而东北地区则以春季（3月–4月）与秋季焚烧（10月–11月）为主。值得注意的是遥感观测数据显示在过去的二十年里，由于经济发展，污染控制，气候变化等方面的的影响，我国秸秆焚烧的强度与燃烧模式发生了明显的变化。

2. 东南亚地区生物质燃烧的特征及其所带来的健康风险。东南亚地区的生物质燃烧主要受火耕农业的影响，同时印尼地区的泥炭地火灾在厄尔尼诺年都会大幅加强，释放大量的温室气体与空气污染物，引发严重的雾霾污染事件。据GEMM暴露-反应模型估计，1990-2019年期间婆罗洲与苏门答腊岛的泥炭地火灾所造成的PM_2.5相关超额死亡数约为32（28–35）万人。

3. 中国雾霾污染治理所取得成就及其所带来的公共健康益处。地面观测网络与遥感观测气溶胶指数（例如气溶胶光学厚度）都表明我国的雾霾污染在过去十年里得到了有效的缓解。自2013年以来，《大气污染防治行动计划》的实施大大降低了居民的PM_2.5暴露风险。结合动态趋势模型与暴露-反应模型估算，此防治方案使2014年至2019年全国PM_2.5相超额死亡数降低了约69（63–75）万人。但随着中国人口结构的改变，以及气候变化的影响，未来PM_2.5所造成的超额死亡人数仍面临大幅上升的风险。

Abstract: With the development and improvement of remote sensing technology, satellite observations have provided a large amount of essential data for research on agriculture, forestry, hydrology, the ecological environment, etc. Simultaneously, the combination of the remote sensing data and the ground monitoring network provides theoretical support for creating mitigation pathways to fight against air pollution and global warming. Based on my research interest and the findings of my past research, I will introduce the application of remote sensing technology in the observation of biomass burning and air pollution from the following three aspects.

1. Spatiotemporal characteristics and variation of crop residue burning in China. As the various crop farming systems and climatic conditions, crop residue burning presents distinctive spatial and seasonal patterns in each part. For example, in east China the burning mainly occurred in summer (June) after the harvest of winter wheat, while in the northeast region the burning mainly burned in two seasons, the spring (March-April) and autumn (October-November). Noteworthily, remote sensing observations have revealed that owing to the impact of economic growth, pollution control, climate change, etc., the intensity and pattern of crop residue burning in China have changed dramatically in the past two decades.

2. Characteristics of biomass burning in Southeast Asia and the related health burden. Biomass burning in Southeast Asia is mainly influenced by slash-and-burn agriculture. Meanwhile, peatland fires in Indonesia intensified substantially during El Niño years, which emitted enormous greenhouse gases and air pollutants, and simultaneously caused severe haze pollution episodes. According to the results of the Global Exposure Mortality Model, the intense peatland fires on Sumatra and Borneo Island were estimated to have induced 317 (282–348) thousand excess deaths from 1990 to 2019, with excess deaths mainly occurring in the El Niño years.

3. Achievements in China's haze pollution control and the corresponding public health benefits. Both the ground observation network and satellite-retrieved aerosol indices (e.g., aerosol optical depth) indicated that the haze pollution of China has been effectively mitigated in the past decade. The launch of the "Air Pollution Prevention and Control Action Plan" has greatly reduced the public exposure risk. The result from the dynamic trend model and exposure-response model showed that this action is estimated to reduce the nationwide PM_2.5-related deaths by 687 (626–743) thousand from 2014 to 2019. However, with the changes in China's demographic structure and the impact of climate change, the number of premature mortality attributable to PM_2.5 exposure is still likely to rise to an unprecedented level in the future.

报告人4：徐坤 15:55-16:30

题目：全球变化对加拿大北方森林树木生长和林分密度的影响建模

Modeling the Impacts of Global Change on Tree Growth and Stand Density of Boreal Forests in Canada

摘要：全球变化深远地影响着北方森林并对加拿大森林功能完整性造成挑战。然而，我们对全球变化对北方森林生长、生物量和林分密度的影响还不甚了解。在这项研究中，我整合了超过3万个永久森林样地数据，进行了以下建模：（1）通过整合对生长起到关键作用的气候因素，建立了五个主要树种生物量的异速生长模型；（2）建立了阿尔伯塔省永久森林样地的火灾风险模型，从而定量分析其火灾存活情况；（3）对北美洲北方森林林分密度进行了模型估计。我的研究在以下三个方面体现了对北方森林生态研究的贡献：（1）建立基于气候的异速生长模型能有助于评估气候对森林生物量积累的影响情况；（2）指出阿尔伯塔省永久样地的平均火灾存活时间为28.7年这一重要长期样地管理指标；（3）估计北美洲北方森林现存3513亿颗树。这些发现预计会增进我们对加拿大北方森林受气候变化影响的了解程度，并且有助于制定缓解气候变化的有关决策。

Global change has profoundly impacted boreal forests, and challenges the functional integrity of forests in Canada. However, its impacts on the growth, biomass and tree density of boreal forests remains poorly understood. In this research, I compiled over 30,000 permanent sample plots (PSPs) to: (1) develop tree biomass allometric equations for five major timber species by incorporating climatic factors critical to growth, (2) build a fire hazards model to quantify the fire risks of PSPs in Alberta, and (3) estimate boreal tree density in North America. My research advances boreal forest ecology on several fronts, including (1) building climate-based allometric models that allow assessing the impact of climate on forest biomass accumulation, (2) finding the mean survival time of PSPs in Alberta as 28.7 years, a parameter critical to maintaining long-term forest inventory plots, and (3) estimating a total of 351.3 billion boreal trees in North America. These findings are expected to contribute to understanding impacts of climate change on Canadian boreal forests and informing climate change mitigation policy-making.