How do local-scale phenomena relate to large-scale phenomena and vice versa? This is one of the key questions for improving climate models. Because the atmosphere, plants, and soils play important roles in the Earth's carbon and water cycles, a better understanding of atmosphere-biosphere interactions near the surface is crucial for the accurate forecasting of future climate, monitoring carbon fluxes, and understanding ecosystem services. The EcoSpec project is a study of how plants and ecosystems respond and contribute to environments using hyperspectral remote sensing, a method independent from the traditional eddy covariance method.
Environmental conditions such as temperature, moisture, and light strongly influence photosynthesis, respiration, and evapotranspiration in ecosystems. These functions vary with the composition and properties of the land surface and their changes. Spectral reflectance of land surface is a function of the composition, abundance, and configuration of surface featuressuch as plants and soilsand their physical and chemical properties.
By applying the principles of biophysical remote sensing, scientists from the Environmental Science and Mathematics and Computer Science Divisions have prototyped a method using a hyperspectral reflectance measurements of land surfaces to investigate local-scale, high-temporal-frequency interactions between the near-surface atmosphere and the terrestrial biosphere and dynamics of ecosystem functions.
The EcoSpec optical tower system is equipped with a hyperspectral sensor (or spectrometer) with over 2,000 spectral channels, thermal infrared sensors, an RGB (red, green, blue) camera, a diffuse radiometer, and an albedometer. The system can be 100% solar powered and is controlled by a single-board computer that supports synchronized movement of instruments, collects data every minute from dawn to dusk, and transmits the data wirelessly throughout growing seasons.
The EcoSpec team first determined the indicative power of hyperspectral reflectance for plant activities and ecosystem functions across the season. Hyperspectral data collected from soybeans throughout the growing season indicated that plant properties such as green biomass, light use efficiency, and chlorophyll content primarily control photosynthesis during different parts of the season. Ancillary true-color photos recording land surface conditions during hyperspectral measurements showed intricate variation of foliar cover, soil exposure, shadow, and surface soil wetness throughout the growing seasons, which could help describing the observed spectral reflectance values.
The EcoSpec system was deployed to the Fermi agricultural site during two growing seasons. With the agricultural system as a case study, the EcoSpec team has prototyped data collection and quality control protocols for high-temporal-frequency hyperspectral remote sensing and integrated analytics for hyper-dimensional time-series forecasting.
Paper: Hamada, Y.; Cook, D.; Bales, D. EcoSpec: Highly Equipped Tower-Based Hyperspectral and Thermal Infrared Automatic Remote Sensing System for Investigating Plant Responses to Environmental Changes. Sensors 2020, 20, 5463.
Poster: Tarpey, P; Hamada Y. Feature selection for high dimensional Time series forecasting with artificial neural networks. Poster presentation at: Yale Undergraduate Research Conference. February 11-12, 2017; New Haven, CT.
Poster: Hamada, Y. et al. Tower-based optical sensing architecture for facilitating the investigation of fine scale biosphere-atmosphere interactions via optical-flux data integration. Poster presented at: The 2014 AGU Fall Meeting; December 15-19, 2014; San Francisco, CA.
EcoSpec Web Site
For the latest information on the project, see the EcoSpec Project web site.
Related Research Areas
See the Research Highlights Index for a complete list of EVS Research Areas.