Related Experiment Video
Updated: May 18, 2026
08:47
Computer Vision-Based Biomass Estimation for Invasive Plants
Published on: February 9, 2024
Estimation of broadband emissivity (8-12 um) from ASTER data by using RM-NN
1Key Laboratory of Agri-informatics, MOA, and Hulunber Grassland Ecosystem Observation and Research Station, Institute of Agricultural Resources and Regional Planning, Chinese Academy of Agricultural Sciences, Beijing 100081, China. maokebiao@126.com
Optics Express
|October 6, 2012
Summary
A new radiative transfer model (RM) with neural network (NN) algorithm accurately estimates land surface emissivity using ASTER satellite data. This method improves upon existing products, offering higher precision for climate and remote sensing applications.
Area of Science:
- Earth Science
- Remote Sensing
- Atmospheric Science
Background:
- Land surface window emissivity is crucial for calculating the longwave radiative budget.
- Existing ASTER Standard Data Product (AST05) emissivity has an accuracy of ±0.015, limited by atmospheric correction.
- Accurate emissivity estimation is vital for climate modeling and understanding Earth's energy balance.
Purpose of the Study:
- To develop and validate a novel algorithm for directly estimating window (8-12 µm) emissivity from Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) brightness temperatures.
- To compare the performance of the new algorithm against existing methods for emissivity estimation.
- To assess the accuracy and robustness of the proposed method.
Main Methods:
- Utilizing a combined radiative transfer model (RM) and neural network (NN) algorithm.
- Directly estimating broadband emissivity from ASTER 1B brightness temperatures using simulated radiance transfer (MODTRAN 4).
- Employing brightness temperatures from ASTER bands 11, 12, 13, and 14 as input nodes for the dynamic neural network.
Main Results:
- The RM-NN algorithm demonstrates higher competence in estimating broadband emissivity compared to other methods.
- Achieved an average estimation accuracy of approximately 0.009.
- The estimation results showed minimal sensitivity to instrument noise.
- Successful application to extract emissivity from ASTER 1B data in China, with results validated against higher-resolution data.
Conclusions:
- The RM-NN algorithm provides a more accurate and robust method for estimating land surface window emissivity.
- This approach overcomes limitations associated with atmospheric correction in standard products.
- The method holds significant potential for improving climate studies and remote sensing applications requiring precise emissivity data.
