Mapping canopy traits over Québec using airborne and spaceborne imaging spectroscopy.

The advent of new spaceborne imaging spectrometers offers new opportunities for ecologists to map vegetation traits at global scales. However, to date most imaging spectroscopy studies exploiting satellite spectrometers have been constrained to the landscape scale. In this paper we present a new method to map vegetation traits at the landscape scale and upscale trait maps to the continental level, using historical spaceborne imaging spectroscopy (Hyperion) to derive estimates of leaf mass per area, nitrogen, and carbon concentrations of forests in Québec, Canada. We compare estimates for each species with reference field values and obtain good agreement both at the landscape and continental scales, with patterns consistent with the leaf economic spectrum. By exploiting the Hyperion satellite archive to map these traits and successfully upscale the estimates to the continental scale, we demonstrate the great potential of recent and upcoming spaceborne spectrometers to benefit plant biodiversity monitoring and conservation efforts.

Evolutionary history explains foliar spectral differences between arbuscular and ectomycorrhizal plant species.

Leaf spectra are integrated foliar phenotypes that capture a range of traits and can provide insight into ecological processes. Leaf traits, and therefore leaf spectra, may reflect belowground processes such as mycorrhizal associations. However, evidence for the relationship between leaf traits and mycorrhizal association is mixed, and few studies account for shared evolutionary history. We conduct partial least squares discriminant analysis to assess the ability of spectra to predict mycorrhizal type. We model the evolution of leaf spectra for 92 vascular plant species and use phylogenetic comparative methods to assess differences in spectral properties between arbuscular mycorrhizal and ectomycorrhizal plant species. Partial least squares discriminant analysis classified spectra by mycorrhizal type with 90% (arbuscular) and 85% (ectomycorrhizal) accuracy. Univariate models of principal components identified multiple spectral optima corresponding with mycorrhizal type due to the close relationship between mycorrhizal type and phylogeny. Importantly, we found that spectra of arbuscular mycorrhizal and ectomycorrhizal species do not statistically differ from each other after accounting for phylogeny. While mycorrhizal type can be predicted from spectra, enabling the use of spectra to identify belowground traits using remote sensing, this is due to evolutionary history and not because of fundamental differences in leaf spectra due to mycorrhizal type.

Predicting leaf traits across functional groups using reflectance spectroscopy.

Plant ecologists use functional traits to describe how plants respond to and influence their environment. Reflectance spectroscopy can provide rapid, non-destructive estimates of leaf traits, but it remains unclear whether general trait-spectra models can yield accurate estimates across functional groups and ecosystems. We measured leaf spectra and 22 structural and chemical traits for nearly 2000 samples from 103 species. These samples span a large share of known trait variation and represent several functional groups and ecosystems, mainly in eastern Canada. We used partial least-squares regression (PLSR) to build empirical models for estimating traits from spectra. Within the dataset, our PLSR models predicted traits such as leaf mass per area (LMA) and leaf dry matter content (LDMC) with high accuracy (R2 > 0.85; %RMSE < 10). Models for most chemical traits, including pigments, carbon fractions, and major nutrients, showed intermediate accuracy (R2  = 0.55-0.85; %RMSE = 12.7-19.1). Micronutrients such as Cu and Fe showed the poorest accuracy. In validation on external datasets, models for traits such as LMA and LDMC performed relatively well, while carbon fractions showed steep declines in accuracy. We provide models that produce fast, reliable estimates of several functional traits from leaf spectra. Our results reinforce the potential uses of spectroscopy in monitoring plant function around the world.

Spatial response resampling (SR2): Accounting for the spatial point spread function in hyperspectral image resampling.

With the increased availability of hyperspectral imaging (HSI) data at various scales (0.03-30 m), the role of simulation is becoming increasingly important in data analysis and applications. There are few commercially available tools to spatially degrade imagery based on the spatial response of a coarser resolution sensor. Instead, HSI data are typically spatially degraded using nearest neighbor, pixel aggregate or cubic convolution approaches. Without accounting for the spatial response of the simulated sensor, these approaches yield unrealistically sharp images. This article describes the spatial response resampling (SR2) workflow, a novel approach to degrade georeferenced raster HSI data based on the spatial response of a coarser resolution sensor. The workflow is open source and widely available for personal, academic or commercial use with no restrictions. The importance of the SR2 workflow is shown with three practical applications (data cross-validation, flight planning and data fusion of separate VNIR and SWIR images).•The SR2 workflow derives the point spread function of a specified HSI sensor based on nominal data acquisition parameters (e.g., integration time, altitude, speed), convolving it with a finer resolution HSI dataset for data simulation.•To make the workflow approachable for end users, we provide a MATLAB function that implements the SR2 methodology.

Landscape-scale concordance between local ecological knowledge for tropical wild species and remote sensing of land cover.

Monitoring the status of species is crucial for biodiversity conservation and sustainable resource management in tropical forests, but conventional in situ monitoring methods are impractical over large scales. Scientists have resorted to two potentially complementary approaches: local ecological knowledge (LEK) and remote sensing. To gauge the potential of combining LEK and remote sensing for assessing species status at landscape scales, a large-scale assessment of the reliability of both measures is critical but hampered by the lack of ground-level data. We conducted a landscape-scale assessment of LEK and remote sensing, using a survey of over 900 communities (a near census in our study area) and nearly 4,000 households in 235 randomly selected communities in the Peruvian Amazon-the largest LEK survey as yet undertaken in tropical forests. The survey collected LEK data on the presence of 20 indicator species from both community leaders/elders and randomly sampled households. We assessed LEK and remotely sensed land cover-forest cover and nonmain channel open water-as proxies for species habitat, across species (game, fish, and timber), over time (current and historical), and by indigeneity (Indigenous peoples and mestizos). Overall, LEK and remotely sensed land cover corroborate each other well. Concordance is highest for the current status of game species reported by sampled households, as is the concordance of historical LEK from Indigenous community leaders/elders. The results point to the promise of combining LEK and remote sensing in monitoring the status of species in data-poor tropical forests.

Water column compensation workflow for hyperspectral imaging data.

Our article describes a data processing workflow for hyperspectral imaging data to compensate for the water column in shallow, clear to moderate optical water types. We provide a MATLAB script that can be readily used to implement the described workflow. We break down each code segment of this script so that it is more approachable for use and modification by end users and data providers. The workflow initially implements the method for water column compensation described in Lyzenga (1978) and Lyzenga (1981), generating depth invariant indices from spectral band pairs. Given the high dimensionality of hyperspectral imaging data, an overwhelming number of depth invariant indices are generated in the workflow. As such, a correlation based feature selection methodology is applied to remove redundant depth invariant indices. In a post-processing step, a principal component transformation is applied, extracting features that account for a substantial amount of the variance from the non-redundant depth invariant indices while reducing dimensionality. To fully showcase the developed methodology and its potential for extracting bottom type information, we provide an example output of the water column compensation workflow using hyperspectral imaging data collected over the coast of Philpott's Island in Long Sault Parkway provincial park, Ontario, Canada.•Workflow calculates depth invariant indices for hyperspectral imaging data to compensate for the water column in shallow, clear to moderate optical water types.•The applied principal component transformation generates features that account for a substantial amount of the variance from the depth invariant indices while reducing dimensionality.•The output (both depth invariant index image and principal component image) allows for the analysis of bottom type in shallow, clear to moderate optical water types.

Implementation of the directly-georeferenced hyperspectral point cloud.

Before pushbroom hyperspectral imaging (HSI) data can be applied in remote sensing applications, it must typically be preprocessed through radiometric correction, atmospheric compensation, geometric correction and spatial resampling procedures. After these preprocessing procedures, HSI data are conventionally given as georeferenced raster images. The raster data model compromises the spatial-spectral integrity of HSI data, leading to suboptimal results in various applications. Inamdar et al. (2021) developed a point cloud data format, the Directly-Georeferenced Hyperspectral Point Cloud (DHPC), that preserves the spatial-spectral integrity of HSI data more effectively than rasters. The DHPC is generated through a data fusion workflow that uses conventional preprocessing protocols with a modification to the digital surface model used in the geometric correction. Even with the additional elevation information, the DHPC is still stored with file sizes up to 13 times smaller than conventional rasters, making it ideal for data distribution. Our article aims to describe the DHPC data fusion workflow from Inamdar et al. (2021), providing all the required tools for its integration in pre-existing processing workflows. This includes a MATLAB script that can be readily applied to carry out the modification that must be made to the digital surface model used in the geometric correction. The MATLAB script first derives the point spread function of the HSI data and then convolves it with the digital surface model input in the geometric correction. By breaking down the MATLAB script and describing its functions, data providers can readily develop their own implementation if necessary. The derived point spread function is also useful for characterizing HSI data, quantifying the contribution of materials to the spectrum from any given pixel as a function of distance from the pixel center. Overall, our work makes the implementation of the DHPC data fusion workflow transparent and approachable for end users and data providers.•Our article describes the Directly-Georeferenced Hyperspectral Point Cloud (DHPC) data fusion workflow, which can be readily implemented with existing processing protocols by modifying the input digital surface model used in the geometric correction.•We provide a MATLAB function that performs the modification to the digital surface model required for the DHPC workflow. This MATLAB script derives the point spread function of the hyperspectral imager and convolves it with the digital surface model so that the elevation data are more spatially consistent with the hyperspectral imaging data as collected.•We highlight the increased effectiveness of the DHPC over conventional raster end products in terms of spatial-spectral data integrity, data storage requirements, hyperspectral imaging application results and site exploration via virtual and augmented reality.

Influence of Color-to-Gray Conversion on the Performance of Document Image Binarization: Toward a Novel Optimization Problem.

This paper presents a novel preprocessing method of color-to-gray document image conversion. In contrast to the conventional methods designed for natural images that aim to preserve the contrast between different classes in the converted gray image, the proposed conversion method reduces as much as possible the contrast (i.e., intensity variance) within the text class. It is based on learning a linear filter from a predefined data set of text and background pixels that: 1) when applied to background pixels, minimizes the output response and 2) when applied to text pixels, maximizes the output response, while minimizing the intensity variance within the text class. Our proposed method (called learning-based color-to-gray) is conceived to be used as preprocessing for document image binarization. A data set of 46 historical document images is created and used to evaluate subjectively and objectively the proposed method. The method demonstrates drastically its effectiveness and impact on the performance of state-of-the-art binarization methods. Four other Web-based image data sets are created to evaluate the scalability of the proposed method.

A wood density and aboveground biomass variability assessment using pre-felling inventory data in Costa Rica.

BACKGROUND: The high spatio-temporal variability of aboveground biomass (AGB) in tropical forests is a large source of uncertainty in forest carbon stock estimation. Due to their spatial distribution and sampling intensity, pre-felling inventories are a potential source of ground level data that could help reduce this uncertainty at larger spatial scales. Further, exploring the factors known to influence tropical forest biomass, such as wood density and large tree density, will improve our knowledge of biomass distribution across tropical regions. Here, we evaluate (1) the variability of wood density and (2) the variability of AGB across five ecosystems of Costa Rica. RESULTS: Using forest management (pre-felling) inventories we found that, of the regions studied, Huetar Norte had the highest mean wood density of trees with a diameter at breast height (DBH) greater than or equal to 30 cm, 0.623 ± 0.182 g cm-3 (mean ± standard deviation). Although the greatest wood density was observed in Huetar Norte, the highest mean estimated AGB (EAGB) of trees with a DBH greater than or equal to 30 cm was observed in Osa peninsula (173.47 ± 60.23 Mg ha-1). The density of large trees explained approximately 50% of EAGB variability across the five ecosystems studied. Comparing our study's EAGB to published estimates reveals that, in the regions of Costa Rica where AGB has been previously sampled, our forest management data produced similar values. CONCLUSIONS: This study presents the most spatially rich analysis of ground level AGB data in Costa Rica to date. Using forest management data, we found that EAGB within and among five Costa Rican ecosystems is highly variable. Combining commercial logging inventories with ecological plots will provide a more representative ground level dataset for the calibration of the models and remotely sensed data used to EAGB at regional and national scales. Additionally, because the non-protected areas of the tropics offer the greatest opportunity to reduce rates of deforestation and forest degradation, logging inventories offer a promising source of data to support mechanisms such as the United Nations REDD + (Reducing Emissions from Tropical Deforestation and Degradation) program.

The application of remote sensing for detecting mass graves: an experimental animal case study from Costa Rica.

Detection of mass graves utilizing the hyperspectral information in airborne or satellite imagery is an untested application of remote sensing technology. We examined the in situ spectral reflectance of an experimental animal mass grave in a tropical moist forest environment and compared it to an identically constructed false grave which was refilled with soil, but contained no cattle carcasses over the course of a 16-month period. The separability of the in situ reflectance spectra was examined with a combination of feature selection and five different nonparametric pattern classifiers. We also scaled up the analysis to examine the spectral signature of the same experimental mass grave from an air-borne hyperspectral image collected 1 month following burial. Our results indicate that at both scales (in situ and airborne), the experimental grave had a spectral signature that was distinct and therefore detectable from the false grave. In addition, we observed that vegetation regeneration was severely inhibited over the mass grave containing cattle carcasses for up to a period of 16 months. This experimental study has demonstrated the real utility of airborne hyperspectral imagery for the detection of a relatively small mass grave (5 m(2)) within a specific climatic zone. Other climatic zones will require similar actualistic modeling studies, but it is clear that the applications of this technology provide the international community with both an early detection tool and a tool for ongoing monitoring.