Atmospheric Lengthscales for Global VSWIR Imaging Spectroscopy

Area of Science:

• Earth Science
• Remote Sensing
• Ecosystem Monitoring

Background:

• Global imaging spectrometers like the Surface Biology and Geology (SBG) mission require accurate atmospheric correction for ecosystem trait mapping.
• Current methods face computational challenges with large data volumes, exceeding processing capacity.

Purpose of the Study:

• To develop a computationally efficient atmospheric correction method for large-scale Earth imaging missions.
• To maintain high accuracy in surface reflectance and atmospheric state estimation.

Main Methods:

• Utilizing the spatial continuity of atmospheric fields to overcome computational bottlenecks.
• Employing spatially sparse solutions to train local linear emulators for rapid inversions.
• Estimating atmospheric conditions at 200m scales.

Main Results:

• The local emulator approach achieves one to two orders of magnitude speed improvement over traditional methods.
• Accuracy is maintained, with reflectance accuracies of 1.1% RMSE or better compared to ground measurements.
• The method's errors are statistically consistent with retrieval uncertainty budgets.

Conclusions:

• Local emulators provide an efficient solution for atmospheric correction in global imaging missions.
• This approach makes rigorous model inversion algorithms feasible for missions like SBG.
• Accurate ecosystem trait mapping is achievable at unprecedented scales.