本数据集为2016-2024年汉江流域丹江口水库以上区域集合降水预报产品，采用基于SSA-CNN-LSTM-MHA耦合深度学习模型构建，实现了高精度的降水集合预报。数据集包含完整的多置信度降水预报结果，覆盖50%、75%、90%三种典型置信水平的区间上下限预报数据，可全面反映降水预报结果的不确定性范围。
\n数据集包含50%、75%、90%三种典型置信水平的区间上下限月尺度预报数据（0.1°*0.1°）均为tif格式。", "ds_source": "
基于降水基准数据集采用SSA-CNN-LSTM-MHA模型得到的集合预报降水数据。
", "ds_process_way": "SSA-CNN-LSTM-MHA模型
", "ds_quality": "数据比较准确，达到精度要求。
", "ds_acq_start_time": "2016-01-01 00:00:00", "ds_acq_end_time": "2024-12-31 00:00:00", "ds_acq_place": "汉江流域丹江口以上流域", "ds_acq_lon_east": 111.83333333333333, "ds_acq_lat_south": 31.333333333333332, "ds_acq_lon_west": 106.25, "ds_acq_lat_north": 34.333333333333336, "ds_acq_alt_low": null, "ds_acq_alt_high": null, "ds_share_type": "apply-access", "ds_total_size": 5913000, "ds_files_count": 649, "ds_format": "*.tif", "ds_space_res": "0.1°", "ds_time_res": "月", "ds_coordinate": "无", "ds_projection": "", "ds_thumbnail": "174ead68-52b7-41db-9204-613d50af332d.png", "ds_thumb_from": 2, "ds_ref_way": "", "paper_ref_way": "", "ds_ref_instruction": "", "ds_from_station": "", "organization_id": "44547705-9513-4685-a641-661bdf406520", "ds_serv_man": "", "ds_serv_phone": "", "ds_serv_mail": "", "doi_value": "", "subject_codes": [ "170.15" ], "quality_level": 0, "publish_time": "2026-07-14 10:09:04", "last_updated": "2026-07-20 09:41:16", "protected": false, "protected_to": "2027-01-01 00:00:00", "lang": "zh", "cstr": "11738.11.ncdc.reservoir.db7532.2026", "i18n": { "en": { "title": "Precipitation forecast dataset based on the SSA-CNN-LSTM-MHA model for the upper reaches of the Danjiangkou Reservoir in the Han River basin from 2016 to 2024", "ds_format": "*.tif", "ds_source": "Ensemble forecast precipitation data obtained based on the precipitation benchmark data set and using the SSA-CNN-LSTM-MHA model.
", "ds_quality": "The data is relatively accurate and meets the accuracy requirements.
", "ds_ref_way": "", "ds_abstract": "This dataset is a collective precipitation forecast product for the area above the Danjiangkou Reservoir in the Hanjiang River Basin from 2016 to 2024. It is constructed using a coupled deep learning model based on SSA-CNN-LSTM-MHA to achieve high-precision collective precipitation forecast. The data set contains complete multi-confidence precipitation forecast results, covering interval upper and lower limit forecast data with three typical confidence levels of 50%, 75%, and 90%, which can fully reflect the uncertainty range of precipitation forecast results.
\r\nThe data set includes interval upper and lower limit monthly scale forecast data (0.1°*0.1°) with three typical confidence levels of 50%, 75%, and 90%, all in tif format.", "ds_time_res": "month", "ds_acq_place": "Hanjiang Basin and Danjiangkou Basin", "ds_space_res": "0.1°", "ds_projection": "", "ds_process_way": "
SSA-CNN-LSTM-MHA model
", "ds_ref_instruction": "None" } }, "submit_center_id": "ncdc", "data_level": 0, "recommendation_value": 0, "license_type": "https://creativecommons.org/licenses/by/4.0/", "doi_reg_from": "reg_local", "cstr_reg_from": "reg_local", "doi_not_reg_reason": null, "cstr_not_reg_reason": null, "is_paper_in_submitting": false, "belong_to_nieer": false, "allow_update_data": false, "ds_topic_tags": [ "集合降水预报" ], "ds_subject_tags": [ "大气科学" ], "ds_class_tags": [], "ds_locus_tags": [ "汉江流域丹江口以上流域" ], "ds_time_tags": [ 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024 ], "ds_contributors": [ { "true_name": "刘勇", "email": "yongliu@nhri.cn", "work_for": "水利部交通运输部国家能源局南京水利科学研究院", "country": "中国" } ], "ds_meta_authors": [ { "true_name": "刘勇", "email": "yongliu@nhri.cn", "work_for": "水利部交通运输部国家能源局南京水利科学研究院", "country": "中国" } ], "ds_managers": [ { "true_name": "刘勇", "email": "yongliu@nhri.cn", "work_for": "水利部交通运输部国家能源局南京水利科学研究院", "country": "中国" } ], "category": "气象" }