本研究采用多层感知器（MLP）神经网络融合多源海洋大地测量数据。构建了一套覆盖180°E–180°W、80°S–80°N，网格分辨率为1′×1′的全球海洋新水深模型——山东科技大学2023全球海图（SDUST2023BCO）。所采用的多源海洋大地测量数据包括：山东科技大学发布的重力异常数据、斯克里普斯海洋研究所发布的垂向重力梯度与垂向偏转数据，以及法国国家空间研究中心发布的平均动力地形数据。首先，基于多源海洋大地测量数据组织MLP模型的输入与输出训练数据；其次，将目标点位的输入数据输入训练完成的MLP模型以获取预测水深；最终构建了覆盖全球海域、分辨率达1′×1′的高精度海底地形模型。通过将SDUST2023BCO模型与船载单波束测深数据以及GEBCO_2023和topo_25.1模型进行对比，评估了该模型的有效性与可靠性。结果表明，SDUST2023BCO模型精确可靠，能有效捕捉并反映全球海洋地形信息。", "ds_source": "
数据来源于https://doi.org/10.5281/zenodo.13341896 。", "ds_process_way": "
本研究重点是建立一个新的全球（南纬80°–北纬80°，东经180°-西经180°）水深模型，名为山东理工大学2023年海洋水深图（SDUST2023BCO）。该模型基于MLP神经网络构建，利用训练/预测点的多源海洋大地测量数据（重力异常、垂直重力梯度、垂直偏转的主分量和素数、平均动态地形）与其周围网格点之间的差异构建。通过与GEBCO_2023和topo_25.1模型进行比较SDUST2023BCO模型的可靠性得到了验证。", "ds_quality": "
SDUST2023BCO已达到国际先进的全球水深模型水平。SDUST2023BCO模型的准确性优于GEBCO_2023和topo_25.1模型，尤其是在较深水域。", "ds_acq_start_time": null, "ds_acq_end_time": null, "ds_acq_place": "全球", "ds_acq_lon_east": 180.0, "ds_acq_lat_south": -90.0, "ds_acq_lon_west": -180.0, "ds_acq_lat_north": 90.0, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "login-access", "ds_total_size": 560524774, "ds_files_count": 2, "ds_format": "*.nc", "ds_space_res": "", "ds_time_res": "", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "89d706f9-b18f-40d6-8890-1ca4297ecdee.png", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": null, "organization_id": "0a4269e1-65f4-45f1-aeba-88ea3068eebf", "ds_serv_man": "敏玉芳", "ds_serv_phone": "0931-4967596", "ds_serv_mail": "ncdc@lzb.ac.cn", "doi_value": "", "subject_codes": [ "170.45" ], "quality_level": 3, "publish_time": "2026-01-30 17:50:48", "last_updated": "2026-01-30 17:50:48", "protected": false, "protected_to": null, "lang": "zh", "cstr": "11738.11.NCDC.ZENODO.DB7080.2026", "license": null, "i18n": { "en": { "title": "SDUST2023BCO: a global seafloor model determined from multi-layer perceptron neural network using multi-source differential marine geodetic data", "ds_format": "*.nc", "ds_source": "
The data is sourced from https://doi.org/10.5281/zenodo.13341896 .", "ds_quality": "
SDUST2023BCO has reached the international advanced global water depth model level. The accuracy of the SDUST2023BCO model is superior to that of the GEBCO2023 and topo-25.1 models, especially in deeper waters.", "ds_ref_way": "", "ds_abstract": "
This study uses a multi-layer perceptron (MLP) neural network to fuse multi-source ocean geodetic data. A new global ocean depth model covering 180 ° E-180 ° W, 80 ° S-80 ° N, with a grid resolution of 1 ′ × 1 ′ has been constructed - Shandong University of Science and Technology 2023 Global Chart (SDUST2023BCO). The multi-source marine geodetic data used includes gravity anomaly data released by Shandong University of Science and Technology, vertical gravity gradient and vertical deflection data released by Scripps Institution of Oceanography, and average dynamic terrain data released by the French National Centre for Space Research. Firstly, the input and output training data of the MLP model are organized based on multi-source marine geodetic data; Secondly, input the input data of the target point into the trained MLP model to obtain the predicted water depth; Finally, a high-precision seabed terrain model covering the global sea area with a resolution of 1 ′ × 1 ′ was constructed. The effectiveness and reliability of the SDUST2023BCO model were evaluated by comparing it with shipborne single beam depth measurement data, as well as the GEBCO2023 and topo-25.1 models. The results indicate that the SDUST2023BCO model is accurate and reliable, and can effectively capture and reflect global ocean terrain information.", "ds_time_res": "", "ds_acq_place": "global", "ds_space_res": "", "ds_projection": "", "ds_process_way": "
The focus of this study is to establish a new global (80 ° S – 80 ° N, 180 ° E – 180 ° W) water depth model called the Shandong University of Technology 2023 Ocean Depth Map (SDUST2023BCO). This model is constructed based on MLP neural network, utilizing the differences between multi-source ocean geodetic data (gravity anomalies, vertical gravity gradients, principal and prime components of vertical deflection, average dynamic terrain) from training/prediction points and their surrounding grid points. The reliability of the SDUST2023BCO model was validated by comparing it with the GEBCO2023 and topo-25.1 models.", "ds_ref_instruction": "" } }, "submit_center_id": "ncdc", "data_level": 0, "license_type": "CC BY 4.0", "ds_topic_tags": [ "SDUST2023BCO", "测量数据", "海底模型" ], "ds_subject_tags": [ "地理学" ], "ds_class_tags": [], "ds_locus_tags": [ "全球" ], "ds_time_tags": [], "ds_contributors": [ { "true_name": "郭金运", "email": "jinyunguo1@126.com", "work_for": "山东科技大学大地测量与地球信息学院", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "郭金运", "email": "jinyunguo1@126.com", "work_for": "山东科技大学大地测量与地球信息学院", "country": "中国" } ], "ds_managers": [ { "true_name": "郭金运", "email": "jinyunguo1@126.com", "work_for": "山东科技大学大地测量与地球信息学院", "country": "中国" } ], "category": "水文" }