A Subset of the Seamless Hourly Human-Perceived Temperature Dataset of China (0.01°, 12 Indices)
收藏资源简介：
Overview This repository provides a representative 95-hour national-level sample of the 1-km hourly human-perceived temperature (thermal index) dataset across China, along with a standalone Python visualization script (visualize_1km_demo.py). This subset is provided for peer review. Due to the massive storage size of the complete archive (3.63 TB), the full dataset is stored on local servers. To request the complete dataset or for any questions, please contact Zhaohua Liu (9120040026@jxust.edu.cn) and Zhaoliang Zeng (zhaoliang.zeng@whu.edu.cn). Repository File Structure To comprehensively demonstrate the dataset's performance across different climatic seasons, this 95-hour subset is organized into four seasonal representative archives in 2024:Dataset_20240101.zip: Winter representative subset (January 1, 2024) Dataset_20240401.zip: Spring representative subset (April 1, 2024) Dataset_20240701.zip: Summer representative subset (July 1, 2024) Dataset_20241001.zip: Autumn representative subset (October 1, 2024) visualize_1km_demo.py: Standalone Python script for fast 2D grid conversion and national map rendering. Data Specifications Spatial Coverage: China (80°E–136°E, 15°N–54°N, corresponding to the Himawari-8/9 coverage over China) Spatial Resolution: 0.01° × 0.01° (approx. 1 km) Temporal Resolution: Hourly (95 hours in total across the four seasonal packages) Variables / Column Descriptions Each .parquet file contains the following columns: lon: Longitude (degrees East, WGS84)lat: Latitude (degrees North, WGS84) final_[INDEX]_1km: Estimated 1-km hourly values for 12 thermal indices (Unit: °C), where corresponds to TEM, WBT, ATin, ATout, DI, NET, sWBGT, WCT, HMI, MDI, ET, and HI (e.g., final_TEM_1km, final_HI_1km). Quick Visualization After unzipping any of the dataset archives, place the .parquet files in the same directory as visualize_1km_demo.py. Running the script will automatically convert the tabular data into 2D matrices and generate high-resolution (300 dpi) national spatial distribution maps (including the South China Sea inset) for all 12 indices using solely pandas, numpy, and matplotlib
