NOAA-Navy Sanctuary Soundscape Monitoring Project, Fin Whale Sound Producion, Florida Keys, SanctSound_FK02_01_finwhale_1d

NOAA and the U.S. Navy are working to better understand underwater sound within the U.S. National Marine Sanctuary System. From 2018 to 2021, these agencies will work with numerous scientific partners to study sound within seven national marine sanctuaries and one marine national monument, which includes waters off Hawai'i and the east and west coasts. Standardized measurements will assess sounds produced by marine animals, physical processes (e.g., wind and waves), and human activities. Collectively, this information will help NOAA and the Navy measure sound levels and baseline acoustic conditions in sanctuaries. This work is a continuation of ongoing Navy and NOAA research, including efforts by NOAA's Office of National Marine Sanctuaries This dataset represents the derived products from the raw acoustic data that are archived at NOAA National Centers for Environmental Information.
abstract=This record represents fin whale sound production detected from raw passive acoustic data. The Low Frequency Detection and Classification System (LFDCS) call library for fin whale 20-Hz pulses was built for the data sampled at 120 Hz. All fin whale detections with a Mahalanobis distance of 3.0 or less were manually verified for true detections. A logistic regression was applied to these results to facilitate reducing the size of the dataset that ultimately needed to be manually verified for confident species detection. This analysis revealed that a minimum number of 29 detections per time window used (hour or day) need to be detected to ensure that a fin whale was truly detected with a confidence of 90%. All days with at least 29 detections were manually verified for daily presence of fin whale 20-Hz pulses. From days with 29 or more detections, fin whales were considered present for that day if a true detection was found within a regular inter-pulse interval pattern of at least three other 20-Hz pulses. These data were recorded at SanctSound Site FK02_01 between December 18, 2018 and April 21, 2019.
acknowledgement=This project received funding from the U.S. Navy.
cdm_data_type=TimeSeries
citation=Cite as: NOAA Office of National Marine Sanctuaries and U.S Navy. 2021. Fin Whale Sound Production Recorded at SanctSound Site FK02_01, SanctSound Data Products. NOAA National Centers for Environmental Information. Accessed [date]. DOI: https://doi.org/http://doi.org/10.25921/b6vh-7t47
comment=Data quality: Quality data were recorded for the duration of the deployment.
contributor_name=Simone Baumann-Pickering, Scripps Institution of Oceanography; Leila Hatch, NOAA Stellwagen Bank National Marine Sanctuary; John Joseph, U.S. Naval Postgraduate School; Anke Kuegler, Hawai'i Institute of Marine Biology, University of Hawai'i at Manoa; Marc Lammers, NOAA Hawaiian Islands Humpback Whale National Marine Sanctuary; Tetyana Margolina, U.S. Naval Postgraduate School; Karlina Merkens, NOAA Pacific Islands Fisheries Science Center; Lindsey Peavey Reeves, NOAA Channel Islands National Marine Sanctuary; Timothy Rowell, NOAA Northeast Fisheries Science Center; Jenni Stanley, Woods Hole Oceanographic Institution; Alison Stimpert, Moss Landing Marine Laboratories; Sofie Van Parijs, NOAA Northeast Fisheries Science Center; Eden Zang,NOAA Hawaiian Islands Humpback Whale National Marine Sanctuary
contributor_role=Principal Investigator
Conventions=COARDS, CF-1.6, ACDD-1.3
featureType=TimeSeries
geospatial_bounds=POINT (24.4888 -81.666316)
history=All acoustic data were processed using the Low Frequency Detection and Classification System (LFDCS; Baumgartner and Mussoline, 2011), which creates conditioned spectrograms using a short-time Fourier transform with a data frame of 512 samples and 75% overlap (80% overlap for the 120 Hz decimated data (blue and fin whales)), resulting in a time step of 64 ms and frequency resolution of 3.9 Hz (for 120 Hz data: 853 ms time step and 0.23 Hz frequency resolution). After tracing contour lines, or “pitch tracks”, through tonal sounds, the program uses multivariate discriminant function analysis to classify the pitch tracks into species-specific call types based on a call library. Each detection is assigned a Mahalanobis distance (MD), which measures the deviation of a sound’s pitch track from the assigned call type (see Baumgartner and Mussoline (2011) for a more complete description). A lower MD indicates a closer match to the assigned call type. For a well-developed call type in the LFDCS (i.e., the seven attributes used in the discriminant function analysis are multivariate normal), 75% of pitch-tracks for the call type will have a MD of 3.0 or less (Baumgartner et al., 2013). Setting a MD threshold is necessary to minimize the false detection rates, but in doing so causes some true detections to be missed in the analysis. The MD threshold of 3.0 was chosen for all vocalizations detected and classified in the humpback, sei, and fin whale call library. However, for blue whales, false detection rates were lower than any of the other species, thus a MD of 5.0 was chosen to decrease the probability of missing true detections. All LFDCS detections were manually reviewed by trained acoustic analysts to determine daily presence of each of the four baleen whale species. A true detection was defined as a pitch track that correctly classified a call or song unit to the species that produced it (Bonnell et al., 2016). Given the variability of each species' call type, the specific methodology to determine daily acoustic presence was different for each species. The LFDCS call library for fin whale 20-Hz pulses was built for the data sampled at 120 Hz. All fin whale detections with a MD of 3.0 or less were manually verified for true detections. A logistic regression was applied to these results to facilitate reducing the size of the dataset that ultimately needed to be manually verified for confident species detection. This analysis revealed that a minimum number of 29 detections per time window used (hour or day) need to be detected to ensure that a fin whale was truly detected with a confidence of 90%. All days with at least 29 detections were manually verified for daily presence of fin whale 20-Hz pulses. From days with 29 or more detections, fin whales were considered present for that day if a true detection was found within a regular inter-pulse interval pattern of at least three other 20-Hz pulses. Data were processed with LFDCS
id=http://doi.org/10.25921/b6vh-7t47
infoUrl=https://ncei.noaa.gov
institution=NOAA
instrument=SoundTrap ST500
keywords_vocabulary=GCMD Science Keywords
naming_authority=NOAA-Navy
project=NOAA-Navy Sanctuary Soundscape Monitoring Project
sourceUrl=(local files)
standard_name_vocabulary=CF Standard Name Table v55

{'abstract': '国家海洋与大气管理局（NOAA）与美国海军合作，旨在深入探究美国国家海洋保护区系统内的水下声音。自2018年至2021年，这些机构将与众多科学合作伙伴共同研究七个国家海洋保护区和一个海洋国家纪念地内的声音，包括夏威夷群岛以及美国东海岸和西海岸周边海域。标准化测量将评估由海洋动物、物理过程（例如，风和波浪）以及人类活动产生的声音。综合这些信息，将有助于NOAA和海军评估保护区内的声音水平和声学基线条件。这项研究是海军和NOAA持续研究工作的延续，包括NOAA国家海洋保护区办公室的各项工作。本数据集代表了对存档于NOAA国家环境信息中心的原始声学数据所得的衍生产品。', 'acknowledgement': '本项目得到了美国海军的资助。', 'cdm_data_type': '时间序列数据', 'citation': '引用格式：NOAA国家海洋保护区办公室和美国海军。2021年。在SanctSound FK02_01站点记录的抹香鲸声音生产，SanctSound数据产品。NOAA国家环境信息中心。访问日期。[日期]。DOI：https://doi.org/http://doi.org/10.25921/b6vh-7t47', 'comment': '数据质量：部署期间记录了高质量数据。', 'contributor_name': 'Simone Baumann-Pickering，斯克里普斯海洋研究所；Leila Hatch，NOAA Stellwagen Bank国家海洋保护区；John Joseph，美国海军研究生院；Anke Kuegler，夏威夷大学海洋生物学研究所；Marc Lammers，NOAA夏威夷群岛座头鲸国家海洋保护区；Tetyana Margolina，美国海军研究生院；Karlina Merkens，NOAA太平洋岛屿渔业科学中心；Lindsey Peavey Reeves，NOAA Channel Islands国家海洋保护区；Timothy Rowell，NOAA东北渔业科学中心；Jenni Stanley，伍兹霍尔海洋研究所；Alison Stimpert，莫斯兰丁海洋实验室；Sofie Van Parijs，NOAA东北渔业科学中心；Eden Zang，NOAA夏威夷群岛座头鲸国家海洋保护区', 'contributor_role': '主要研究员', 'Conventions': 'COARDS, CF-1.6, ACDD-1.3', 'featureType': '时间序列数据', 'geospatial_bounds': '点（24.4888 -81.666316）', 'history': '所有声学数据均使用低频检测与分类系统（LFDCS；Baumgartner and Mussoline, 2011）进行处理，该系统通过短时傅里叶变换和数据帧为512样本、75%重叠（120 Hz降采样数据为80%重叠，蓝色和抹香鲸）创建条件谱图，从而实现64毫秒的时间步长和3.9赫兹的频率分辨率（120赫兹数据：853毫秒时间步长和0.23赫兹频率分辨率）。通过在音调声音中追踪轮廓线，或称为“音调轨迹”，程序使用多元判别函数分析将音调轨迹分类为基于呼叫库的物种特异性呼叫类型。每个检测都被分配一个马氏距离（MD），该距离衡量声音音调轨迹与分配的呼叫类型之间的偏差（参见Baumgartner and Mussoline (2011) 以获得更详细的描述）。较低的MD表示与分配的呼叫类型更接近的匹配。对于LFDCS中发育良好的呼叫类型（即，用于判别函数分析的七个属性是多元正态），75%的呼叫类型的音调轨迹将具有3.0或更低的MD（Baumgartner et al., 2013）。设置MD阈值是必要的，以最大限度地减少误检率，但这样做会导致分析中遗漏一些真实检测。因此，对于所有检测和分类的呼叫，选择了3.0的MD阈值。然而，对于蓝鲸，误检率低于其他任何物种，因此选择了5.0的MD来降低遗漏真实检测的概率。所有LFDCS检测均由经过培训的声学分析师手动审查，以确定四种须鲸物种的每日存在情况。真实检测被定义为正确地将呼叫或歌曲单元分类为产生它的物种的音调轨迹（Bonnell et al., 2016）。鉴于每种物种的叫声类型的可变性，确定每日声学存在的具体方法因物种而异。为120赫兹采样数据构建了抹香鲸20赫兹脉冲的LFDCS呼叫库。所有MD为3.0或更低的抹香鲸检测均进行了手动验证，以确认真实检测。对这些结果应用逻辑回归，以简化最终需要手动验证以确信物种检测的数据库集的大小。分析表明，每个时间窗口（小时或天）至少需要检测29次，以确保以90%的置信度检测到真实的抹香鲸。所有至少有29次检测的日子都进行了每日抹香鲸20赫兹脉冲存在的手动验证。从有29次或更多检测的日子来看，如果在至少三个其他20赫兹脉冲的常规脉冲间隔模式中找到真实检测，则认为当天有抹香鲸的存在。数据使用LFDCS进行处理。', 'id': 'http://doi.org/10.25921/b6vh-7t47', 'infoUrl': 'https://ncei.noaa.gov', 'institution': 'NOAA', 'instrument': 'SoundTrap ST500', 'keywords_vocabulary': 'GCMD科学关键词', 'naming_authority': 'NOAA-Navy', 'project': 'NOAA-Navy sanctuary soundscape monitoring project', 'sourceUrl': '(本地文件)', 'standard_name_vocabulary': 'CF标准名称表v55'}