Spectral Near-Infrared and Thermal Infrared Imaging for Advanced Estimation of Thermal and Geochemical Soil-Plant-Water Properties
The quantification of hydro-biogeochemical properties and fluxes with relevant spatial and temporal resolution requires advances in remote sensing and sensor network technologies, and in linking remotely sensed signals to hydro-biogeochemical properties of interest. The hydro-biogeochemical properties are essentially the information can reveal soil, land qualities, for energy conservation, and agricultures, et al. Such advances are critical for breakthroughs in the management of terrestrial ecosystems, precision agriculture and remediation of contaminated sites. While Unmanned Aerial Vehicles (UAVs) have shown promise for monitoring plant and ground surface properties, practitioners are still lacking an integrated approach that coupled UAV, advanced spectral sensors, ground based measurements and automated processing algorithms to quickly deliver reliable estimates of soil and plant biogeochemical properties. One particular challenge is to couple current UAV-based sensing of spectral reflectance in the visible-short wave infrared (VSWIR) range with multi-spectral thermal infrared (TIR) imaging and distributed sensor networks to enable the quantification of ecosystem properties and fluxes with high spatial and temporal resolution. The idea behind is that the more the spectral coverages, the more the biogeochemical properties/information one can detect. The TIR region in the wavelength range of 8-14 μm is of strong interest for numerous applications. For example, long wavelengths nearly equal to the surface temperature of leaves and plants is a key parameter to estimate stomatal conductance, evapotranspiration, and drought stress. Other applications consist in identifying surface properties, assessing contaminant spills and water quality, and identify particular thermal behaviors controlled by gas concentration, geochemical compounds or solids (incl., algae bloom). Current limitation of UAV-based thermal imaging is the accuracy of measurement, the lack of integration with other sensors, and difficulties in inferring thermal properties or fluxes across space and time. In addition, the current thermal infrared sensors operate in a single broad spectral band while the acquisition of multiple bands is critical for improving the estimation of surface emissivity, surface temperature and for detecting and investigating complex patterns. The objective of this STTR project is to develop an integrated UAV-based remote sensing strategy, including hyperspectral VSWIR and multi-spectral TIR imaging coupled with ground-based sensor network to provide accurate estimate of properties critical for improving the estimation of heat, water and gas fluxes. The project involves the development of a multi-component sensing platform, algorithms and a data integration framework for improving the estimation of various hydro-biogeochemical properties. In Phase 1, we designed and initiated the development of a multi-spectral TIR imaging system based on Fourier Transform Infrared (FTIR) technology. Current results are promising for developing a UAV-based system providing data with unprecedented spatial and spectral resolution. In addition, we developed algorithms to process TIR data, including the estimation of surface temperature, separation of plant canopy from soil pixels and TIR data classification based on plant type and landscape position. These algorithms provide new capabilities in extracting or classifying patterns in remote sensing imagery and represent a critical step for improving the analysis of time-lapse datasets. Further, in order to improve UAV-based measurements analysis and validation, in Phase 1 we developed a prototype of a low-cost, duplicable leaf surface temperature sensor for ground-based deployments. This sensor is aimed at being deployed at numerous locations during UAV-TIR survey in order to improve the calibration and/or validation of UAV-based measurements. Finally, we investigated various surface temperature datasets and developed an approach to disentangle the impact of various processes (incl., radiation, leaf/ground ratio, surface moisture) on the measured surface temperature. This approach has the potential to be included with the other developments in this project to create an integrated UAV based monitoring approach that will provide high spatiotemporally resolved information on plant and soil properties. The developed capability will be highly valuable to improve ecosystem understanding and will have a strong potential for commercial application in ecosystem management, agriculture, and water resource management among others.
- Research Organization:
- BaySpec, Inc.; Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States)
- Sponsoring Organization:
- USDOE Office of Science (SC)
- DOE Contract Number:
- SC0020478
- OSTI ID:
- 1902353
- Type / Phase:
- STTR (Phase I)
- Report Number(s):
- Phase1_Final_Report_Update_12_08_22
- Resource Relation:
- Related Information: Updated final report to include the Data Rights Clause
- Country of Publication:
- United States
- Language:
- English
Similar Records
A Multi-Sensor Unoccupied Aerial System Improves Characterization of Vegetation Composition and Canopy Properties in the Arctic Tundra: Supporting Data
Using soil library hyperspectral reflectance and machine learning to predict soil organic carbon: Assessing potential of airborne and spaceborne optical soil sensing
Related Subjects
58 GEOSCIENCES
09 BIOMASS FUELS
13 HYDRO ENERGY
15 GEOTHERMAL ENERGY
29 ENERGY PLANNING, POLICY, AND ECONOMY
32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATION
Thermal infrared
spectral signature
hyperspectral
wireless sensor network
remote sensing
energy and water fluxes
plant health
Fourier Transform Infrared