中科天机公里级融合数据集2025年华北区域

中国区域公里级融合数据（TJ-CN）依托中科天机全球-区域一体化数值天气模式生成，采用“全球自由变焦”技术对中国区域进行聚焦网格加密，在不增加总网格点和计算量的前提下实现0.025°（约2.5公里）的高分辨率输出；同时，模式具备超精细化的次网格地形参数化能力，能够刻画米级尺度地形特征及其拖曳效应，更加真实地反映下垫面对大气的影响，从而更精准地描述中小尺度的天气现象；在此基础上结合AI技术，将多套天机模式生成的模式数据进行融合，并针对行业最关注的气象要素开展算法改进与智能优化，最终生成逐小时高分辨率、精细化且精确可靠的气象数据。相较于单一模式数据，这套融合数据具有以下核心优势：行业适配性更强；能够更好地满足科研AI方向对高分辨率数据的需求；模拟准确性更高；模拟分辨率更精细化。这些特点使其能更好满足科研与行业应用需求。分辨率：0.025度。要素：10米纬向风、10米经向风、100米纬向风、100米经向风、2米气温、总降水率、2米相对湿度、地表气压、地表向下短波通量共计9个要素。空间范围：华北地区，包含山西、河北、北京、天津四个省份。经度：110.2E ~ 119.8E， 纬度：34.6N ~ 42.7N。中科天机公里级融合数据共享实时获取信息请参考：https://www.tjweather.com/info/Download。中科天机公里级融合数据历史数据集获取信息请参考：https://www.tjweather.com/info/HistoryDownload。

The China-region kilometer-level fusion dataset (TJ-CN) is generated based on the Zhongke Tianji global-regional integrated numerical weather prediction (NWP) model. It adopts the "global free zoom" technology to perform focused grid refinement over the Chinese territory, achieving high-resolution output at 0.025° (approximately 2.5 km) without increasing total grid points and computational load.

Additionally, the model features ultra-refined sub-grid terrain parameterization capabilities, which can characterize meter-scale terrain features and their drag effects, more realistically reflect the impacts of the underlying surface on the atmosphere, and thus more accurately describe mesoscale and microscale weather phenomena.

On this basis, by integrating AI technologies, the dataset fuses multiple sets of Tianji NWP model outputs, carries out algorithm improvements and intelligent optimizations targeting the meteorological variables most concerned by industries, and finally generates hourly high-resolution, refined, accurate and reliable meteorological data.

Compared with single-model data, this fusion dataset has the following core advantages: better industry adaptability, better meeting the demand for high-resolution data in AI for scientific research, higher simulation accuracy, and finer simulation resolution. These characteristics enable it to better meet the needs of scientific research and industrial applications.

Spatial resolution: 0.025°.
Meteorological variables: 10-meter zonal wind, 10-meter meridional wind, 100-meter zonal wind, 100-meter meridional wind, 2-meter air temperature, total precipitation rate, 2-meter relative humidity, surface air pressure, and surface downward shortwave flux, totaling 9 variables.
Spatial scope: North China, including four provincial-level administrative regions: Shanxi, Hebei, Beijing, and Tianjin.
Longitude: 110.2°E ~ 119.8°E, Latitude: 34.6°N ~ 42.7°N.

For real-time access to the Zhongke Tianji kilometer-level fusion dataset, please refer to: https://www.tjweather.com/info/Download. For access to the historical dataset of the Zhongke Tianji kilometer-level fusion dataset, please refer to: https://www.tjweather.com/info/HistoryDownload.

搜集汇总

背景与挑战

背景概述

该数据集基于CAS-TJ全球-区域集成数值天气预报模型生成，利用全球自适应变焦技术在华北区域实现了约2.5公里的高空间分辨率，无需增加网格点或计算成本。它提供了2025年华北地区9个关键气象变量（包括辐射、温度、降水、风速、湿度和气压）的历史模拟数据，以小时为时间分辨率，服务于可再生能源、电力、水资源和农业等领域的高分辨率气象信息需求。