Speaker
Description
Land Surface Temperature (LST) is recognized as an Essential Climate Variable (ECV) by the Global Climate Observing System. It serves as a critical metric for assessing the global surface energy balance, monitoring evapotranspiration, and managing agricultural water use. Consequently, satellite remote sensing remains the primary tool for deriving LST at global scales. However, current thermal infrared missions face a trade-off between spatial and temporal resolution. Most systems offering high revisit frequencies provide data at low spatial resolutions (1 km or greater), while high-resolution missions (60–100 m) suffer from long acquisition intervals, hampering the generation of consistent, quality time series required for agricultural monitoring.
To address this gap, the TRISHNA mission (Thermal infraRed Imaging Satellite for High-resolution Natural resource Assessment), a joint collaboration between the French space agency (CNES) and the Indian Space Research Organisation (ISRO), is scheduled for launch in 2027. TRISHNA aims to provide global thermal data at a 60-meter spatial resolution with a high-frequency 3-day revisit interval. To achieve these ambitious spatiotemporal characteristics, the satellite will observe ground scenes at varying viewing angles over an 8-day orbital cycle. While this viewing strategy ensures frequent coverage, it introduces significant challenges regarding surface anisotropy.
Surface temperature is not directionally uniform; the measured temperature varies depending on the viewing geometry relative to the sun. These directional effects arise from the complex 3D structure of vegetation and soil, which creates varying proportions of sunlit and shaded components within a pixel. Consequently, a "hotspot" effect occurs when the sensor views the scene from the same direction as the sun, necessitating the normalization of these angular measurements to a standard viewing geometry for valid time-series analysis. While kernel-driven models derived from the visible domain (BRDF) have been successfully adapted for thermal measurements, the physical drivers in the thermal domain—specifically heat storage and emission—differ from reflective properties. Therefore, standard models require refinement.
Recent studies suggest that incorporating thermal inertia—the resistance of a material to temperature change—can substantially improve directional modeling. To validate this hypothesis, a field experiment was conducted in a vineyard in Maharashtra, India. Vineyards were selected as they represent complex, row-structured crops that exhibit strong anisotropic effects. The experiment utilized a multi-angular setup with an Altum-PT camera, capturing synchronized data in five visible spectral bands and broadband Thermal Infrared (8–14 µm).
The resulting dataset is being coupled with advanced radiative transfer models to disentangle the drivers of directional anisotropy. This includes the 1D SCOPE model (Soil Canopy Observation, Photochemistry and Energy fluxes) and the 3D DART model (Discrete Anisotropic Radiative Transfer). By integrating field measurements with these simulations, the study aims to quantify the benefits of introducing a thermal inertia component into directional algorithms.
The authors propose to present the experimental field setup, the analysis of the multi-angular camera outputs, and the comparative results of the modeling efforts. This work contributes directly to the algorithmic development for the upcoming TRISHNA mission.
An oral presentation is requested.
