Harmonization and inter-calibration of data streams from UAVs, CubeSats, Sentinel-2, and Landsat 8
Abstract
The emergence of new sensing opportunities (e.g., CubeSats, unmanned aerial vehicles (UAVs)) in parallel with continued advances in conventional space-agency-based satellite platforms present a unique opportunity to develop new insights with potentially far reaching societal and economic benefits. While this high resolution geospatial information revolution could constitute a game changer in the ability to derive time-critical and location-specific insights into dynamic land surface processes, challenges prevail in combining and profiting from multi-sensor data streams. Advanced harmonization techniques are needed to ensure consistency in observations across sensors regardless of sensor design and platform, spectral responses, radiometric performance, absolute calibration accuracy, and view and illumination geometry.
This study presents an automated harmonization framework for ensuring interoperability and reflectance consistency across sensors. The CubeSat-Enabled Spatio-Temporal Enhancement Method (CESTEM) is first used to produce high resolution (daily, 3 m) data cubes of surface reflectance (SR) imagery leveraging rigorously calibrated 'gold standard' satellites (i.e., Sentinel-2, Landsat 8) in synergy with superior resolution CubeSats from Planet. The framework incorporates gap-filling techniques to produce spatially complete daily SR imagery at 3 m. The CESTEM data cubes of harmonized SR imagery then serve as the reference for calibrating cm-scale UAV imagery using a machine learning approach. The framework is largely insensitive to sensor input and calibration (i.e., it can work with DNs, radiances, or reflectances), and avoids the use of impractical ground-based SR collections for drone imagery radiometric calibration.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.B14A..01H
- Keywords:
-
- 0410 Biodiversity;
- BIOGEOSCIENCES;
- 0480 Remote sensing;
- BIOGEOSCIENCES;
- 1910 Data assimilation;
- integration and fusion;
- INFORMATICS