Estimating the outgoing longwave radiation from the FY-3B satellite visible infrared radiometer Channel 5 radiance observations

Xiao Wu, Jun-Jie YanChinese 2011 · DOI

Abstract

Based on the infrared radiation transfer model, the outgoing longwave radiation (OLR) and Channel 5 radiance of Fengyun-3B (FY-3B) satellite visible infrared radiometer (VIRR) were simulated for 3812 global soundings. Using regression analysis of the simulations, an inverse model, which connected the flux equivalent brightness temperature with the channel brightness temperature, was derived. By applying the model to the FY-3B VIRR L1 data, the global OLR data at the time of the passing of the FY-3B were processed. The quality of the data was validated by comparing it with that of the NOAA-18 satellite’s advanced very high resolution radiometer (AVHRR). The validation results show root mean square errors in the range 10–13 W/M2 when comparing the daily average OLR of the VIRR with that of the NOAA-18 AVHRR, and the correlation coefficients were in the range 0.97–0.98. The larger RMSE is mainly due to the different passing times of the two satellites for the specific locations on the Earth. An example of the OLR data and its preliminary applications are given.

Introduction

Outgoing Longwave Radiation (OLR) represents the total energy emitted by the Earth-atmosphere system into space. As a critical component of the Earth's radiation budget, accurate OLR measurements are essential for understanding climate dynamics, monitoring monsoon systems, and predicting extreme weather events.

While broadband radiometers provide direct OLR measurements, many operational satellites, such as China’s Fengyun-3B (FY-3B), carry narrow-band instruments like the Visible Infrared Radiometer (VIRR). The challenge lies in accurately estimating the broadband OLR flux from these narrow-band radiance measurements. This study develops a robust inverse model to bridge this gap using VIRR Channel 5 data.

Methodology

The researchers developed a systematic approach to derive OLR from narrow-band satellite data, as illustrated in the following flowchart:

Flowchart of the OLR estimation process

The process involved four key stages:

  1. Simulation: An infrared radiation transfer model was used to simulate OLR and Channel 5 radiance for 3,812 global atmospheric sounding profiles, covering a diverse range of meteorological conditions.
  2. Model Derivation: A statistical relationship was established between the flux equivalent brightness temperature ($T_F$) and the channel brightness temperature ($T_{B5}$).
  3. Limb Darkening Correction: To account for varying satellite viewing angles, a correction was applied to normalize observations taken at different zenith angles.
  4. Validation: The model was applied to FY-3B VIRR Level 1 data and compared against the industry-standard NOAA-18 AVHRR OLR product.

Results and Validation

The core of the estimation method is the regression model. The researchers found that a quadratic relationship provided the most accurate fit for converting narrow-band brightness temperature to flux equivalent temperature.

Relationship between flux equivalent brightness temperature and Channel 5 brightness temperature

The resulting model demonstrated high precision and reliability:

  • Statistical Performance: The correlation coefficient between the simulated $T_F$ and $T_{B5}$ was exceptionally high (0.97–0.98).
  • Global Validation: When compared with NOAA-18 AVHRR data, the Root Mean Square Error (RMSE) for daily average OLR ranged from 10 to 13 $W/m^2$.
  • Error Analysis: The primary source of discrepancy was attributed to the difference in satellite overpass times. Variations in cloud cover and surface temperature between the passing of FY-3B and NOAA-18 naturally lead to differences in observed OLR.
  • Systematic Bias: A minor bias was noted in warm regions ($T_{B5} > 290 K$), where VIRR-derived OLR was approximately 2 $W/m^2$ lower than AVHRR values.

Conclusion and Future Work

The study successfully demonstrates that FY-3B VIRR Channel 5 data can be used to produce high-resolution global OLR maps ($0.01^\circ \times 0.01^\circ$). This data is valuable for tracking large-scale weather systems, such as cold waves and tropical storm development.

Future improvements will focus on updating atmospheric composition parameters in the radiation transfer models, utilizing more recent global sounding datasets, and enhancing cross-platform calibration to minimize systematic biases between different satellite instruments.

Summary updated

Abstract

Based on the infrared radiation transfer model, the outgoing longwave radiation (OLR) and Channel 5 radiance of Fengyun-3B (FY-3B) satellite visible infrared radiometer (VIRR) were simulated for 3812 global soundings. Using regression analysis of the simulations, an inverse model, which connected the flux equivalent brightness temperature with the channel brightness temperature, was derived. By applying the model to the FY-3B VIRR L1 data, the global OLR data at the time of the passing of the FY-3B were processed. The quality of the data was validated by comparing it with that of the NOAA-18 satellite’s advanced very high resolution radiometer (AVHRR). The validation results show root mean square errors in the range 10–13 W/M2 when comparing the daily average OLR of the VIRR with that of the NOAA-18 AVHRR, and the correlation coefficients were in the range 0.97–0.98. The larger RMSE is mainly due to the different passing times of the two satellites for the specific locations on the Earth. An example of the OLR data and its preliminary applications are given.
BibTeX
@inproceedings{wu2011estimating,
  author = {Xiao Wu and Jun-Jie Yan},
  title = {{Estimating the outgoing longwave radiation from the FY-3B satellite visible infrared radiometer Channel 5 radiance observations}},
  booktitle = {Chinese},
  year = {2011},
  doi = {10.1007/s11434-011-4686-6},
}
Figure