Historical CORONA imagery, a very high-resolution panchromatic imagery captured during 1960-72, has a lot of unexplored potentials to represent the land cover of the pre-Landsat era. This pre-Landsat land cover can be a valuable input to many long-term environmental modelling problems. While automated binary change detection and multi-class classification by manual photo-interpretation have been the two most popular applications of CORONA imagery, automated multi-class classification of CORONA imagery has not been explored. In this research, we propose two supervised classification methods capable of extracting spatial features for the classification of CORONA imagery. The first method utilizes two-dimensional convolutional neural networks (2D-CNN) for classification, whereas the second method involves stacking the CORONA imagery with its texture features obtained through geometric moments (GM) and performing pixel-wise classification using the random forest (RF) classifier and is termed as GM-RF. The proposed methods are tested on two study sites – the critical zone observatory (CZO) managed by the Indian Institute of Technology Kanpur in Uttar Pradesh (India) and a site near Chhapra in Bihar (India). For estimating the capabilities of the proposed methods and to compare the proposed methods with other methods, overall accuracy and Cohen’s kappa coefficient (k) along with z-scores statistics are used. The effects of various hyperparameters (e.g. percentage of training data, patch size, order of geometric moments) on accuracy are also evaluated. The classification accuracies achieved through the 2D-CNN and GM-RF method in terms of ĸ are higher than 90% for both study sites indicating that the proposed methods can classify CORONA imagery with little domain knowledge about the study areas, and hence can be used in a variety of applications.

Scalable plant monitoring methods are needed for the early detection of physiological changes induced by pathogens and abiotic stressors. Recent advances in sub-nanometer hyperspectral imaging enable the detection of Fraunhofer lines (FLs), which represent distinct narrow absorption lines in the solar spectrum. This study explored the use of FLs to detect subtle physiological changes associated with early stages of biotic stress. Airborne sensors comprising a sub-nanometer hyperspectral imager, a narrow-band hyperspectral imager, and a thermal camera were flown over a commercial olive orchard infected with the vascular pathogen Verticillium dahliae. Thermal imagery screened early-stage symptomatic trees under fixed water stress levels so that canopy temperature would remain insensitive to transpiration changes caused by pathogen colonizing the vascular system. We then assessed the sensitivity of FLs and oxygen absorption features (O₂-A and O₂-B) by comparing absorption depths between asymptomatic and symptomatic trees. The sub-nanometer O₂-A feature and six FLs differentiated early-stage symptomatic trees under constant (i.e., insensitive) water-stress levels, which the canopy temperature assessment failed to detect. A classification model integrating the O₂-A band depth with two sensitive FL depths (FL_680.98-nm and FL_744.581-nm) achieved a mean overall accuracy of 0.77 and a mean kappa value of 0.55. These findings underscore the potential of FLs derived from airborne sub-nanometer imagery for early detection of vascular diseases when canopy temperature is insensitive to subtle transpiration depletion. This exploratory work provides a foundation for future research towards the role of sub-nanometer imaging-derived fluorescence emission in monitoring plant biotic-induced physiological responses.

Atmospheric interference has been a major challenge in the remote sensing of sun-induced chlorophyll fluorescence (SIF) from airborne platforms. While existing algorithms use ground-reference targets to tune the O2-A band transmittance, they often rely on static atmospheric assumptions that ignore the influence of subtle topographical changes on the O2-A transmittance. Such simplifications can introduce systematic biases into the resultant SIF estimates, and the impact on the spatial pattern of SIF remains poorly understood. To address this spatial SIF artifact, we present a joint retrieval algorithm with full atmospheric model. This algorithm has been optimized for accurate SIF retrieval in simulation and in-situ data. Within this joint retrieval framework, we constrained the key atmospheric states (i.e., aerosol optical thickness and pressure elevation) instead of treating them as constants, and retrieved SIF from a benchmark airborne sub-nanometer hyperspectral image obtained in a maize field in 2021. We retrieved the spatial pattern of airborne SIF which best matched the SIF field observations at a stringent constraint on the distribution of aerosol optical thickness (with a standard deviation of 10-6) and a loose constraint on the distribution of pressure elevation (with a standard deviation of 10-1). The strong constraint on the aerosol optical thickness restricted its spatial variability, maintaining a physically plausible, smooth pattern. The weak constraint on the pressure elevation allows for tuning the O2-A band transmittance, enabling the spatial pattern to better adapt to local spatial variability in topography and temperature. This combination of prior constraints resulted in a balanced SIF retrieval for the airborne imagery, avoiding over-fitting to microtopography without sacrificing spatial integrity. Our findings demonstrated that carefully constraining these atmospheric states in the optimal estimation inversion considerably alleviate the spatial artifact stemming from the static atmospheric assumption and improve the SIF quantification. The optimal constraints identified in this study align with the physical variability of AOT and effective pressure elevation at the airborne scale, suggesting that this configuration is transferable to other airborne missions. This underscores that for airborne SIF missions, atmospheric parameters should not be treated as fixed constants but rather as variables within a regularized joint inversion.

Hyperspectral retrieval of nitrogen-related vegetation variables has undergone rapid methodological advances driven by the emergence of protein-sensitive radiative transfer models (RTMs), modern machine learning (ML), and operational imaging spectroscopy. This review synthesizes recent developments in hyperspectral retrieval of nitrogen-related vegetation variables across leaf and canopy scales, with particular emphasis on advances reported between 2020 and 2026. We examine the evolution from classical parametric regression and nonlinear ML approaches towards physically based RTM inversion and hybrid RTM–ML frameworks that integrate the complementary strengths of physical modeling and statistical learning. Particular attention is given to protein-sensitive RTMs, advanced ML approaches, and uncertainty-aware retrieval. Recent developments highlight the potential of hybrid RTM–ML frameworks to combine physical consistency with computationally efficient statistical learning, while probabilistic methods such as Gaussian Process Regression provide additional capabilities for uncertainty characterization. The review further discusses the transition from experimental studies to operational applications enabled by airborne and satellite imaging spectroscopy, including PRISMA, EnMAP, and forthcoming missions such as CHIME. Remaining challenges include the inherently ill-posed nature of nitrogen retrieval, limited and insufficiently representative calibration data, uncertainty characterization, and generalization across sensors, species, and ecosystems. Overall, the reviewed evidence points towards increasingly integrated retrieval frameworks, while emphasizing that robust transferability and operational implementation remain dependent on representative data, physical realism, and rigorous uncertainty assessment.