CDI-Type I: Collaborative Research: A Computational Thinking Approach to Mapping Critical Marine Mammal Habitat Through Readily-Deployable Video Systems

CDI-I 型：协作研究：通过易于部署的视频系统绘制关键海洋哺乳动物栖息地的计算思维方法

• 批准号: 1124936
• 负责人: G. Carleton Ray
• 金额: $ 6万
• 依托单位: University of Virginia Main Campus
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2011
• 资助国家: 美国
• 起止时间: 2011-09-15 至 2015-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1124936&HistoricalAwards=false
• 关键词: CDI; Type; Collaborative; Research; Computational

Unprecedented thinning and retreat of the Arctic sea ice cover together with recent climate modeling studies that predict the Arctic could be free or nearly free of sea ice in summertime within the next few decades have raised concern for the future of Arctic ice-­associated marine mammals. All species of ice-dependent marine mammals of Beringia have been subject to petitions to designate them as threatened or endangered under the Endangered Species Act of 1973. Accordingly, in 2008, the polar bear was listed as threatened after the U.S Fish and Wildlife Service [2008] found that ?polar bear habitat - principally sea ice - is declining throughout the species? range?. The case for the Pacific walrus is under review with the U.S. Fish and Wildlife Service Endangered Species Program and two ice seal species have been proposed for protective status. A major challenge facing these national policy decisions is lack of information concerning the extent and distribution of critical habitat within the overall ice pack. Most climate models and standard sea ice data products provide ice extent and concentration information, but these quantities only partially explain the distribution of Arctic marine mammals. The size and shape of ice floes and openings has also been shown to be important, but there are currently no standardized means of monitoring these properties of the sea ice cover. The work proposed here aims to develop new, powerful video processing techniques and bring them to bear on the problem of identifying and quantifying critical habitat areas within the Arctic ice pack. The project team brings together scientists and engineers with expertise in advanced image and video analysis and modeling, computational science, sea ice geophysics and marine mammal ecology. The research plan is centered on a computational-thinking approach to transforming high-volume video data into low-volume high-relevance information that is key to decision support in a range of settings. The overall aim is to develop and implement advanced video processing algorithms based on geophysical and ecological knowledge of sea ice to routinely map marine mammal habitat in ice covered waters. Data from the system will be disseminated to a cyber-enabled forum of experts, who will aid in habitat interpretation and provide guidance for data acquisition. The project will build on existing cyber-enabled forums such as the Sea Ice for Walrus Outlook (SIWO, co-organized by Eicken) as models for implementation. The techniques that will be developed will be readily extendible to other observing tasks such as marine hazard identification. Such information will become increasingly important in the near future with growing commercial activity and limited decision-making and support infrastructure in the Arctic. A readily-deployable, networkable system will allow increased marine traffic in the Arctic to be turned to an advantage.This project is supported by the NSF Directorate for Geosciences and the Experimental Program to Stimulate Competitive Research (EPSCoR).

北极海冰覆盖前所未有的变薄和消退，加上最近的气候模型研究预测未来几十年内北极可能在夏季没有或几乎没有海冰，这引起了人们对与北极冰有关的海洋哺乳动物的未来的担忧。根据 1973 年《濒危物种法》，白令陆桥所有依赖冰的海洋哺乳动物都已收到请愿书，将其指定为受威胁或濒危物种。因此，在 2008 年，美国鱼类和野生动物管理局 [2008] 发现“整个物种的北极熊栖息地（主要是海冰）正在减少”后，北极熊被列为受威胁物种。范围？。美国鱼类和野生动物管理局濒危物种计划正在审查太平洋海象的案件，并已提议对两种冰海豹物种进行保护。这些国家政策决策面临的一个主要挑战是缺乏有关整个冰层内关​​键栖息地的范围和分布的信息。大多数气候模型和标准海冰数据产品都提供冰范围和浓度信息，但这些数量只能部分解释北极海洋哺乳动物的分布。浮冰和开口的大小和形状也被证明很重要，但目前没有标准化的方法来监测海冰覆盖的这些特性。这里提出的工作旨在开发新的、强大的视频处理技术，并将其应用于识别和量化北极冰层内关键栖息地区域的问题。该项目团队汇集了在高级图像和视频分析与建模、计算科学、海冰地球物理学和海洋哺乳动物生态学方面拥有专业知识的科学家和工程师。该研究计划以计算思维方法为中心，将大量视频数据转换为少量高相关性信息，这对于一系列环境中的决策支持至关重要。总体目标是开发和实施基于海冰地球物理和生态知识的先进视频处理算法，以定期绘制冰覆盖水域中海洋哺乳动物栖息地的地图。来自该系统的数据将被传播到一个由专家组成的网络论坛，他们将帮助解释栖息地并为数据采集提供指导。该项目将以现有的网络论坛为基础，例如 Sea Ice for Walrus Outlook（SIWO，由 Eicken 联合组织）作为实施模型。将开发的技术将很容易扩展到其他观测任务，例如海洋灾害识别。随着北极商业活动的不断增长以及决策和支持基础设施的有限，此类信息在不久的将来将变得越来越重要。易于部署、可联网的系统将使北极地区增加的海上交通转化为优势。该项目得到了 NSF 地球科学理事会和刺激竞争性研究实验计划 (EPSCoR) 的支持。