Construction of a High-Resolution Temperature Dataset at 40–110 KM over China Utilizing TIMED/SABER and FY-4A Satellite Data
This study aims to develop a high-resolution temperature dataset from 40 km to 110 km over China by machine learning techniques, with a horizontal resolution of 0.5° × 0.5° and vertical resolution of 1 km, utilizing measurements from SABER onboard the Thermosphere, Ionosphere, Mesosphere Energetics,...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Atmosphere |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4433/16/7/758 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study aims to develop a high-resolution temperature dataset from 40 km to 110 km over China by machine learning techniques, with a horizontal resolution of 0.5° × 0.5° and vertical resolution of 1 km, utilizing measurements from SABER onboard the Thermosphere, Ionosphere, Mesosphere Energetics, and Dynamics (TIMED) and Fengyun 4A (FY-4A) satellites. Accurate temperature profiles play a critical role in understanding the atmospheric dynamics and climate change. However, because of the limitation of traditional detecting methods, the measurements of the upper stratosphere and mesosphere are rare. In this study, a new method is developed to construct a high-resolution temperature dataset over China in the middle atmosphere based on the XGBoost technique. The model’s performance is also validated based on rocket observations and ERA5 reanalysis data. The results indicate that the model effectively captures the characteristics of the vertical and seasonal variations in temperature, which provide a valuable opportunity for further research and improvement of climate models. The model demonstrates the highest accuracy below 80 km with RMSE < 12 K, while its performance decreases above 100 km, where RMSE can exceed 20 K, indicating optimal performance in the upper stratosphere and lower mesosphere regions. |
|---|---|
| ISSN: | 2073-4433 |