GLORIA - A globally representative hyperspectral in situ dataset for optical sensing of water quality.

The development of algorithms for remote sensing of water quality (RSWQ) requires a large amount of in situ data to account for the bio-geo-optical diversity of inland and coastal waters. The GLObal Reflectance community dataset for Imaging and optical sensing of Aquatic environments (GLORIA) includes 7,572 curated hyperspectral remote sensing reflectance measurements at 1 nm intervals within the 350 to 900 nm wavelength range. In addition, at least one co-located water quality measurement of chlorophyll a, total suspended solids, absorption by dissolved substances, and Secchi depth, is provided. The data were contributed by researchers affiliated with 59 institutions worldwide and come from 450 different water bodies, making GLORIA the de-facto state of knowledge of in situ coastal and inland aquatic optical diversity. Each measurement is documented with comprehensive methodological details, allowing users to evaluate fitness-for-purpose, and providing a reference for practitioners planning similar measurements. We provide open and free access to this dataset with the goal of enabling scientific and technological advancement towards operational regional and global RSWQ monitoring.

Algorithms for remote estimation of chlorophyll-a in coastal and inland waters using red and near infrared bands.

Remote sensing algorithms that use red and NIR bands for the estimation of chlorophyll-a concentration [Chl] can be more effective in inland and coastal waters than algorithms that use blue and green bands. We tested such two-band and three-band red-NIR algorithms using comprehensive synthetic data sets of reflectance spectra and inherent optical properties related to various water parameters and a very consistent in situ data set from several lakes in Nebraska, USA. The two-band algorithms tested with MERIS bands were Rrs(708)/Rrs(665) and Rrs(753)/Rrs(665). The three-band algorithm with MERIS bands was in the form R3=[Rrs(-1)(665)-Rrs(-1)(708)]×Rrs(753). It is shown that the relationships of both Rrs(708)/Rrs(665) and R3 with [Chl] do not depend much on the absorption by CDOM and non-algal particles, or the backscattering properties of water constituents, and can be defined in terms of water absorption coefficients at the respective bands as well as the phytoplankton specific absorption coefficient at 665 nm. The relationship of the latter with [Chl] was established for [Chl]>1 mg/m3 and then further used to develop algorithms which showed a very good match with field data and should not require regional tuning.

Efficiency of chlorophyll in gross primary productivity: A proof of concept and application in crops.

One of the main factors affecting vegetation productivity is absorbed light, which is largely governed by chlorophyll. In this paper, we introduce the concept of chlorophyll efficiency, representing the amount of gross primary production per unit of canopy chlorophyll content (Chl) and incident PAR. We analyzed chlorophyll efficiency in two contrasting crops (soybean and maize). Given that they have different photosynthetic pathways (C3 vs. C4), leaf structures (dicot vs. monocot) and canopy architectures (a heliotrophic leaf angle distribution vs. a spherical leaf angle distribution), they cover a large spectrum of biophysical conditions. Our results show that chlorophyll efficiency in primary productivity is highly variable and responds to various physiological and phenological conditions, and water availability. Since Chl is accessible through non-destructive, remotely sensed techniques, the use of chlorophyll efficiency for modeling and monitoring plant optimization patterns is practical at different scales (e.g., leaf, canopy) and under widely-varying environmental conditions. Through this analysis, we directly related a functional characteristic, gross primary production with a structural characteristic, canopy chlorophyll content. Understanding the efficiency of the structural characteristic is of great interest as it allows explaining functional components of the plant system.

Apple flavonols during fruit adaptation to solar radiation: spectral features and technique for non-destructive assessment.

Spectral properties of flavonols of three varieties (Golden Delicious, Antonovka, and Renet Simirenko) of anthocyanin-free apple fruit were investigated with reflectance spectroscopy. The results of spectral and biochemical analyses suggested that fruit reflectance in a broad spectral range 365-430 nm is strongly dependent on and, in sunlit fruit surfaces, governed by flavonols. The build up of peel flavonols (mainly rutin and other quercetin glycosides) resulted in a dramatic decrease of fruit reflectance in this range, flattening of the spectrum, and extending the region with low reflectance (4-5%) to ca. 410 nm. The spectral features observed suggest that flavonols contribute significantly to screening of excessive radiation, not only UV-A, but in the short-wave bands of chlorophyll and carotenoid absorption in the visible part of the spectrum as well. To retrieve quantitatively flavonol content from reflectance spectra, we tested the applicability of an inversion technique developed for non-destructive leaf pigment assessment. The model for flavonol content assessment was suggested in the form (R(-1)410 - R(-1)460)R800, where Rlambda is reflectance at wavelength lambda. The model was linearly related to flavonol content between 8 and 220nmol/cm2 with the coefficient of determination r2=0.92 and root mean square error of flavonol estimation of 20 nmol/ cm2 regardless of cultivar, chlorophyll, and carotenoid content.

Non-destructive estimation of foliar carotenoid content of tree species using merged vegetation indices.

Leaf pigment content is an important indicator of plant status and can serve to assess the vigor and photosynthetic activity of plants. The application of spectral information gathered from laboratory, field and remote sensing-based spectrometers to non-destructively assess total chlorophyll (Chl) content of higher plants has been demonstrated in earlier studies. However, the precise estimation of carotenoid (Car) content with non-destructive spectral measurements has so far not reached accuracies comparable to the results obtained for Chl content. Here, we examined the potential of a recently developed angular vegetation index (AVI) to estimate total foliar Car content of three tree species. Based on an iterative search of all possible band combinations, we identified a best candidate AVIcar. The identified index showed quite close but essentially not linear relation with Car contents of the examined species with increasing sensitivity to high Car content and a lack of sensitivity to low Car content for which earlier proposed vegetation indices (VI) performed better. To make use of the advantages of both VI types, we developed a simple merging procedure, which combined the AVIcar with two earlier proposed carotenoid indices. The merged indices had close linear relationship with total Car content and outperformed all other examined indices. The merged indices were able to accurately estimate total Car content with a percental root mean square error (%RMSE) of 8.12% and a coefficient of determination of 0.88. Our findings were confirmed by simulations using the radiative transfer model PROSPECT-5. For simulated data, the merged indices again showed a quasi linear relationship with Car content. This strengthens the assumption that the proposed merged indices have a general ability to accurately estimate foliar Car content. Further examination of the proposed merged indices to estimate foliar Car content of other plant species is desirable to prove the general applicability of the index for non-destructive estimation of Car from leaf reflectance data.

Productivity, absorbed photosynthetically active radiation, and light use efficiency in crops: implications for remote sensing of crop primary production.

Vegetation productivity metrics such as gross primary production (GPP) at the canopy scale are greatly affected by the efficiency of using absorbed radiation for photosynthesis, or light use efficiency (LUE). Thus, close investigation of the relationships between canopy GPP and photosynthetically active radiation absorbed by vegetation is the basis for quantification of LUE. We used multiyear observations over irrigated and rainfed contrasting C3 (soybean) and C4 (maize) crops having different physiology, leaf structure, and canopy architecture to establish the relationships between canopy GPP and radiation absorbed by vegetation and quantify LUE. Although multiple LUE definitions are reported in the literature, we used a definition of efficiency of light use by photosynthetically active "green" vegetation (LUE(green)) based on radiation absorbed by "green" photosynthetically active vegetation on a daily basis. We quantified, irreversible slowly changing seasonal (constitutive) and rapidly day-to-day changing (facultative) LUE(green), as well as sensitivity of LUE(green) to the magnitude of incident radiation and drought events. Large (2-3-fold) variation of daily LUE(green) over the course of a growing season that is governed by crop physiological and phenological status was observed. The day-to-day variations of LUE(green) oscillated with magnitude 10-15% around the seasonal LUE(green) trend and appeared to be closely related to day-to-day variations of magnitude and composition of incident radiation. Our results show the high variability of LUE(green) between C3 and C4 crop species (1.43 g C/MJ vs. 2.24 g C/MJ, respectively), as well as within single crop species (i.e., maize or soybean). This implies that assuming LUE(green) as a constant value in GPP models is not warranted for the crops studied, and brings unpredictable uncertainties of remote GPP estimation, which should be accounted for in LUE models. The uncertainty of GPP estimation due to facultative and constitutive changes in LUE(green) can be considered as a critical component of the total error budget in the context of remotely sensed based estimations of GPP. The quantitative framework of LUE(green) estimation presented here offers a way of characterizing LUE(green) in plants that can be used to assess their phenological and physiological status and vulnerability to drought under current and future climatic conditions and is essential for calibration and validation of globally applied LUE algorithms.

Assessing carotenoid content in plant leaves with reflectance spectroscopy.

Spectral reflectance of maple, chestnut and beech leaves in a wide range of pigment content and composition was investigated to devise a nondestructive technique for total carotenoid (Car) content estimation in higher plant leaves. Reciprocal reflectance in the range 510 to 550 nm was found to be closely related to the total pigment content in leaves. The sensitivity of reciprocal reflectance to Car content was maximal in a spectral range around 510 nm; however, chlorophylls (Chl) also affect reflectance in this spectral range. To remove the Chl effect on the reciprocal reflectance at 510 nm, a reciprocal reflectance at either 550 or 700 nm was used, which was linearly proportional to the Chl content. Indices for nondestructive estimation of Car content in leaves were devised and validated. Reflectances in three spectral bands, 510+/-5 nm, either 550+/-15 nm or 700+/-7.5 nm and the near infrared range above 750 nm are sufficient to estimate total Car content in plant leaves nondestructively with a root mean square error of less than 1.75 nmol/cm2.

Wide Dynamic Range Vegetation Index for remote quantification of biophysical characteristics of vegetation.

The Normalized Difference Vegetation Index (NDVI) is widely used for monitoring, analyzing, and mapping temporal and spatial distributions of physiological and biophysical characteristics of vegetation. It is well documented that the NDVI approaches saturation asymptotically under conditions of moderate-to-high aboveground biomass. While reflectance in the red region (rho(red)) exhibits a nearly flat response once the leaf area index (LAI) exceeds 2, the near infrared (NIR) reflectance (PNIR) continue to respond significantly to changes in moderate-to-high vegetation density (LAI from 2 to 6) in crops. However, this higher sensitivity of the rho(NIR) has little effect on NDVI values once the rho(NIR) exceeds 30%. In this paper a simple modification of the NDVI was proposed. The Wide Dynamic Range Vegetation Index, WDRVI = (a * rho(NIR-rho(red))/(a * rho(NIR) + rho(red)), where the weighting coefficient a has a value of 0.1-0.2, increases correlation with vegetation fraction by linearizing the relationship for typical wheat, soybean, and maize canopies. The sensitivity of the WDRVI to moderate-to-high LAI (between 2 and 6) was at least three times greater than that of the NDVI. By enhancing the dynamic range while using the same bands as the NDVI, the WDRVI enables a more robust characterization of crop physiological and phenological characteristics. Although this index needs further evaluation, the linear relationship with vegetation fraction and much higher sensitivity to change in LAI will be especially valuable for precision agriculture and monitoring vegetation status under conditions of moderate-to-high density. It is anticipated that the new index will complement the NDVI and other vegetation indices that are based on the red and NIR spectral bands.

Relationships between leaf chlorophyll content and spectral reflectance and algorithms for non-destructive chlorophyll assessment in higher plant leaves.

Leaf chlorophyll content provides valuable information about physiological status of plants. Reflectance measurement makes it possible to quickly and non-destructively assess, in situ, the chlorophyll content in leaves. Our objective was to investigate the spectral behavior of the relationship between reflectance and chlorophyll content and to develop a technique for non-destructive chlorophyll estimation in leaves with a wide range of pigment content and composition using reflectance in a few broad spectral bands. Spectral reflectance of maple, chestnut, wild vine and beech leaves in a wide range of pigment content and composition was investigated. It was shown that reciprocal reflectance (R lambda)-1 in the spectral range lambda from 520 to 550 nm and 695 to 705 nm related closely to the total chlorophyll content in leaves of all species. Subtraction of near infra-red reciprocal reflectance, (RNIR)-1, from (R lambda)-1 made index [(R lambda)(-1)-(RNIR)-1] linearly proportional to the total chlorophyll content in spectral ranges lambda from 525 to 555 nm and from 695 to 725 nm with coefficient of determination r2 > 0.94. To adjust for differences in leaf structure, the product of the latter index and NIR reflectance [(R lambda)(-1)-(RNIR)-1]*(RNIR) was used; this further increased the accuracy of the chlorophyll estimation in the range lambda from 520 to 585 nm and from 695 to 740 nm. Two independent data sets were used to validate the developed algorithms. The root mean square error of the chlorophyll prediction did not exceed 50 mumol/m2 in leaves with total chlorophyll ranged from 1 to 830 mumol/m2.

Estimation of chlorophyll-a concentration in turbid productive waters using airborne hyperspectral data.

Algorithms based on red and near infra-red (NIR) reflectances measured using field spectrometers have been previously shown to yield accurate estimates of chlorophyll-a concentration in turbid productive waters, irrespective of variations in the bio-optical characteristics of water. The objective of this study was to investigate the performance of NIR-red models when applied to multi-temporal airborne reflectance data acquired by the hyperspectral sensor, Airborne Imaging Spectrometer for Applications (AISA), with non-uniform atmospheric effects across the dates of data acquisition. The results demonstrated the capability of the NIR-red models to capture the spatial distribution of chlorophyll-a in surface waters without the need for atmospheric correction. However, the variable atmospheric effects did affect the accuracy of chlorophyll-a retrieval. Two atmospheric correction procedures, namely, Fast Line-of-sight Atmospheric Adjustment of Spectral Hypercubes (FLAASH) and QUick Atmospheric Correction (QUAC), were applied to AISA data and their results were compared. QUAC produced a robust atmospheric correction, which led to NIR-red algorithms that were able to accurately estimate chlorophyll-a concentration, with a root mean square error of 5.54 mg m(-3) for chlorophyll-a concentrations in the range 2.27-81.17 mg m(-3).

NIR-red reflectance-based algorithms for chlorophyll-a estimation in mesotrophic inland and coastal waters: Lake Kinneret case study.

A variety of models have been developed for estimating chlorophyll-a (Chl-a) concentration in turbid and productive waters. All are based on optical information in a few spectral bands in the red and near-infra-red regions of the electromagnetic spectrum. The wavelength locations in the models used were meticulously tuned to provide the highest sensitivity to the presence of Chl-a and minimal sensitivity to other constituents in water. But the caveat in these models is the need for recurrent parameterization and calibration due to changes in the biophysical characteristics of water based on the location and/or time of the year. In this study we tested the performance of NIR-red models in estimating Chl-a concentrations in an environment with a range of Chl-a concentrations that is typical for coastal and mesotrophic inland waters. The models with the same spectral bands as MERIS, calibrated for small lakes in the Midwest U.S., were used to estimate Chl-a concentration in the subtropical Lake Kinneret (Israel), where Chl-a concentrations ranged from 4 to 21 mg m(-3) during four field campaigns. A two-band model without re-parameterization was able to estimate Chl-a concentration with a root mean square error less than 1.5 mg m(-3). Our work thus indicates the potential of the model to be reliably applied without further need of parameterization and calibration based on geographical and/or seasonal regimes.

Nondestructive estimation of anthocyanins and chlorophylls in anthocyanic leaves.

The anthocyanin and chlorophyll contents in leaves provide valuable information about the physiological status of plants. Thus, there is a need for accurate, efficient, and practical methodologies to estimate these biochemical parameters of vegetation. In this study, we tested the performance and accuracy of several nondestructive, reflectance-based techniques for estimating anthocyanin and chlorophyll contents in leaves of four unrelated species, European hazel (Corylus avellana), Siberian dogwood (Cornus alba =Swida alba), Norway maple (Acer platanoides), and Virginia creeper (Parthenocissus quinquefolia), with widely variable pigment content and composition. An anthocyanin reflectance index, which uses reflectances in the green and red edge spectral bands, and a modified anthocyanin reflectance index, employing, in addition, the near-infrared (NIR) band, were able to accurately estimate leaf anthocyanin for all species taken together with no reparameterization of algorithms. Total chlorophyll content was accurately estimated by a red edge chlorophyll index that uses spectral bands in the red edge and the NIR. These approaches can be used to estimate anthocyanin and chlorophyll nondestructively and allow the development of simple handheld field instrumentation.

Effect of bio-optical parameter variability and uncertainties in reflectance measurements on the remote estimation of chlorophyll-a concentration in turbid productive waters: modeling results.

Most algorithms for retrieving chlorophyll-a concentration (Chla) from reflectance spectra assume that bio-optical parameters such as the phytoplankton specific absorption coefficient (aPhi*) or the chlorophyll-a fluorescence quantum yield (eta) are constant. Yet there exist experimental data showing large ranges of variability for these quantities. The main objective of this study was to analyze the sensitivity of two Chla algorithms to variations in bio-optical parameters and to uncertainties in reflectance measurements. These algorithms are specifically designed for turbid productive waters and are based on red and near-infrared reflectances. By means of simulated data, it is shown that the spectral regions where the algorithms are maximally sensitive to Chla overlap those of maximal sensitivity to variations in the above bio-optical parameters. Thus, to increase the accuracy of Chla retrieval, we suggest using spectral regions where the algorithms are less sensitive to Chla, but also less sensitive to these interferences. aPhi* appeared to be one of the most important sources of error for retrieving Chla. However, when the phytoplankton backscattering coefficient (bb,Phi) dominates the total backscattering, as is likely during algal blooms, variations in the specific bb,Phi may introduce large systematic uncertainties in Chla estimation. Also, uncertainties in reflectance measurements, which are due to incomplete atmospheric correction or reflected skylight removal, seem to affect considerably the accuracy of Chla estimation. Instead, variations in other bio-optical parameters, such as eta or the specific backscattering coefficient of total suspended particles, appear to have minor importance. Suggestions regarding the optimal band locations to be used in the above algorithms are finally provided.

Effect of bio-optical parameter variability on the remote estimation of chlorophyll-a concentration in turbid productive waters: experimental results.

The analytical development and underlying hypothesis of a three-band algorithm for estimating chlorophyll-a concentration ([Chla]) in turbid productive waters are presented. The sensitivity of the algorithm to the spectral location of the bands used is analyzed. A large set of experimental observations ([Chla] varied between 4 and 217 mg m(-3) and turbidity between 2 and 78 nephelometric turbidity units) was used to calibrate and validate the algorithm. It was found that the variability of the chlorophyll-a fluorescence quantum yield and of the chlorophyll-a specific absorption coefficient can reduce considerably the accuracy of remote predictions of [Chla]. Instead of parameterizing these interferences, their effects were minimized by tuning the spectral regions used in the algorithm. This allowed us to predict [Chla] with a relative root-mean-square error of less than 30%.