Fast and nondestructive method for leaf level chlorophyll estimation using hyperspectral LiDAR
Nevalainen, O. ; Hakala, T. ; Suomalainen, J.M. ; Mäkipää, R. ; Peltoniemi, M. ; Krooks, A. ; Kaasalainen, S. - \ 2014
Agricultural and Forest Meteorology 198-199 (2014). - ISSN 0168-1923 - p. 250 - 258.
supercontinuum laser source - vegetation indexes - reflectance spectra - precision agriculture - canopy reflectance - red edge - airborne - model - spectroscopy - validation
We propose an empirical method for nondestructive estimation of chlorophyll in tree canopies. The first prototype of a full waveform hyperspectral LiDAR instrument has been developed by the Finnish Geodetic Institute (FGI). The instrument efficiently combines the benefits of passive and active remote sensing sensors. It is able to produce 3D point clouds with spectral information included for every point, which offers great potential in the field of environmental remote sensing. The investigation was conducted by using chlorophyll sensitive vegetation indices applied to hyperspectral LiDAR data and testing their performance in chlorophyll estimation. The amount of chlorophyll in vegetation is an important indicator of photosynthetic capacity and stress, and thus important for monitoring of forest condition and carbon sequestration on Earth. Performance of chlorophyll estimation was evaluated for 27 published vegetation indices applied to waveform LiDAR collected from ten Scots pine shoots. Reference data were collected by laboratory chlorophyll concentration analysis. The performance of the indices in chlorophyll estimation was determined by linear regression and leave-one-out cross-validation. The chlorophyll estimates derived from hyperspectral LiDAR linearly correlate with the laboratory analyzed chlorophyll concentrations, and they are able to represent a range of chlorophyll concentrations in Scots pine shoots (R2 = 0.88, RMSE = 0.10 mg/g). Furthermore, they are insensitive to measurement scale as nearly the same values of vegetation indices were measured in natural setting while scanning the whole canopy and from clipped shoots re-measured with hyperspectral LiDAR in laboratory. The results indicate that the hyperspectral LiDAR instrument has the potential to estimate vegetation biochemical parameters such as the chlorophyll concentration. The instrument holds much potential in various environmental applications and provides a significant improvement over single wavelength LiDAR or passive optical systems for environmental remote sensing.
Assessing water stress of desert Tamarugo trees using in situ data and very high spatial resolution remote sensing
Chávez Oyanadel, R.O. ; Clevers, J.G.P.W. ; Herold, M. ; Acevedo, E. ; Ortiz, M. - \ 2013
Remote Sensing 5 (2013)10. - ISSN 2072-4292 - p. 5064 - 5088.
oriented image-analysis - northern chile - red edge - chlorophyll concentration - canopy - plants - soil - reflectance - vegetation - photoinhibition
The hyper-arid Atacama Desert is one of the most extreme environments for life and only few species have evolved to survive its aridness. One such species is the tree Prosopis tamarugo Phil. Because Tamarugo completely depends on groundwater, it is being threatened by the high water demand from the Chilean mining industry and the human consumption. In this paper, we identified the most important biophysical variables to assess the water status of Tamarugo trees and tested the potential of WorldView2 satellite images to retrieve these variables. We propose green canopy fraction (GCF) and green drip line leaf area index (DLLAIgreen) as best variables and a value of 0.25 GCF as a critical threshold for Tamarugo survival. Using the WorldView2 spectral bands and an object-based image analysis, we showed that the NDVI and the Red-edge Chlorophyll Index (CIRed-edge) have good potential to retrieve GCF and DLLAIgreen. The NDVI performed best for DLLAIgreen (RMSE = 0.4) while the CIRed-edge was best for GCF (RMSE = 0.1). However, both indices were affected by Tamarugo leaf movements (leaves avoid facing direct solar radiation at the hottest time of the day). Thus, monitoring systems based on these indices should consider the time of the day and the season of the year at which the satellite images are acquired.
Changes in plant defense chemistry (pyrrolizidine alkaloids) revealed through high-resolution spectroscopy
Almeida De Carvalho, S. ; Macel, M. ; Schlerf, M. ; Moghaddam, F.E. ; Mulder, P.P.J. ; Skidmore, A.K. ; Putten, W.H. van der - \ 2013
ISPRS Journal of Photogrammetry and Remote Sensing 80 (2013). - ISSN 0924-2716 - p. 51 - 60.
near-infrared spectroscopy - senecio-jacobaea - red edge - nitrogen - leaf - reflectance - forest - regression - vegetation - prediction
Plant toxic biochemicals play an important role in defense against natural enemies and often are toxic to humans and livestock. Hyperspectral reflectance is an established method for primary chemical detection and could be further used to determine plant toxicity in the field. In order to make a first step for pyrrolizidine alkaloids detection (toxic defense compound against mammals and many insects) we studied how such spectral data can estimate plant defense chemistry under controlled conditions. In a greenhouse, we grew three related plant species that defend against generalist herbivores through pyrrolizidine alkaloids: Jacobaea vulgaris, Jacobaea erucifolia and Senecio inaequidens, and analyzed the relation between spectral measurements and chemical concentrations using multivariate statistics. Nutrient addition enhanced tertiary-amine pyrrolizidine alkaloids contents of J. vulgaris and J. erucifolia and decreased N-oxide contents in S. inaequidens and J. vulgaris. Pyrrolizidine alkaloids could be predicted with a moderate accuracy. Pyrrolizidine alkaloid forms tertiary-amines and epoxides were predicted with 63% and 56% of the variation explained, respectively. The most relevant spectral regions selected for prediction were associated with electron transitions and CH, OH, and NH bonds in the 1530 and 2100 nm regions. Given the relatively low concentration in pyrrolizidine alkaloids concentration (in the order of mg g-1) and resultant predictions, it is promising that pyrrolizidine alkaloids interact with incident light. Further studies should be considered to determine if such a non-destructive method may predict changes in PA concentration in relation to plant natural enemies. Spectroscopy may be used to study plant defenses in intact plant tissues, and may provide managers of toxic plants, food industry and multitrophic-interaction researchers with faster and larger monitoring possibilities
Using spectral information from the NIR water absorption features for the retrieval of canopy water content
Clevers, J.G.P.W. ; Kooistra, L. ; Schaepman, M.E. - \ 2008
International Journal of applied Earth Observation and Geoinformation 10 (2008)3. - ISSN 0303-2434 - p. 388 - 397.
leaf optical-properties - fuel moisture-content - imaging spectrometry data - vegetation liquid water - reflectance data - ecosystem processes - hyperspectral data - aviris data - red edge - part 1
Canopy water content (CWC) is important for mapping and monitoring the condition of the terrestrial ecosystem. Spectral information related to the water absorption features at 970 nm and 1200 nm offers possibilities for deriving information on CWC. In this study, we compare the use of derivative spectra, spectral indices and continuum removal techniques for these regions. Hyperspectral reflectance data representing a range of canopies were simulated using the combined PROSPECT + SAILH model. Best results in estimating CWC were obtained by using spectral derivatives at the slopes of the 970 nm and 1200 nm water absorption features. Real data from two different test sites were analysed. Spectral information at both test sites was obtained with an ASD FieldSpec spectrometer, whereas at the second site HyMap airborne imaging spectrometer data were also acquired. Best results were obtained for the derivative spectra. In order to avoid the potential influence of atmospheric water vapour absorption bands the derivative of the reflectance on the right slope of the canopy water absorption feature at 970 nm can best be used for estimating CWC.
Estimation of green grass/herb biomass from airborne hyperspectral imagery using spectral indices and partial least squares regression
Cho, M.A. ; Skidmore, A.K. ; Corsi, F. ; Wieren, S.E. van; Sobhan, I. - \ 2007
International Journal of applied Earth Observation and Geoinformation 9 (2007)4. - ISSN 0303-2434 - p. 414 - 424.
vegetation indexes - area index - red edge - biophysical relationships - imaging spectrometry - aboveground biomass - canopy reflectance - nitrogen status - leaves - contamination
The main objective was to determine whether partial least squares (PLS) regression improves grass/herb biomass estimation when compared with hyperspectral indices, that is normalised difference vegetation index (NDVI) and red-edge position (REP). To achieve this objective, fresh green grass/herb biomass and airborne images (HyMap) were collected in the Majella National Park, Italy in the summer of 2005. The predictive performances of hyperspectral indices and PLS regression models were then determined and compared using calibration (n = 30) and test (n = 12) data sets. The regression model derived from NDVI computed from bands at 740 and 771 nm produced a lower standard error of prediction (SEP = 264 g m¿2) on the test data compared with the standard NDVI involving bands at 665 and 801 nm (SEP = 331 g m¿2), but comparable results with REPs determined by various methods (SEP = 261 to 295 g m¿2). PLS regression models based on original, derivative and continuum-removed spectra produced lower prediction errors (SEP = 149 to 256 g m¿2) compared with NDVI and REP models. The lowest prediction error (SEP = 149 g m¿2, 19% of mean) was obtained with PLS regression involving continuum-removed bands. In conclusion, PLS regression based on airborne hyperspectral imagery provides a better alternative to univariate regression involving hyperspectral indices for grass/herb biomass estimation in the Majella National Park.
Predicting in situ pasture quality in the Kruger National Park, South Africa using continuum removed absorption features
Mutanga, O. ; Skidmore, A.K. ; Prins, H.H.T. - \ 2004
Remote Sensing of Environment 89 (2004)3. - ISSN 0034-4257 - p. 393 - 408.
infrared reflectance spectroscopy - multiple linear-regression - red edge - chlorophyll concentration - spectral discrimination - cross-validation - leaf reflectance - forest canopy - aviris data - nitrogen
The remote sensing of pasture quality as determined by nitrogen, phosphorous, potassium, calcium and magnesium concentration is critical for a better understanding of wildlife and livestock feeding patterns. Although remote sensing techniques have proved useful for assessing the concentration of foliar biochemicals under controlled laboratory conditions, more investigation is required to assess their capabilities in the field, where inconsistent results have been obtained so far. We investigated the possibility of determining the concentration of in situ biochemicals in a savanna rangeland, using the spectral reflectance of five grass species. Canopy spectral measurements were taken in the field using a GER 3700 spectroradiometer. We tested the utility of using four variables derived from continuum-removed absorption features for predicting canopy nitrogen, phosphorus, potassium, calcium and magnesium concentration: (i) continuum-removed derivative reflectance (CRDR), (ii) band depth (BD), (iii) band depth ratio (BDR) and (iv) normalised band depth index (NBDI). Stepwise linear regression was used to select wavelengths from the absorption-feature-based variables. Univariate correlation analysis was also done between the first derivative reflectance and biochemicals. Using a training data set, the variables derived from continuum-removed absorption features could predict biochemicals with R2 values ranging from 0.43 to 0.80. Results were highest using CRDR data, which yielded R2 values of 0.70, 0.80, 0.64, 0.50 and 0.68 with root mean square errors (RMSE) of 0.01, 0.004, 0.03, 0.01 and 0.004 for nitrogen, phosphorous, potassium, calcium and magnesium, respectively. Predicting biochemicals on a test data set, using regression models developed from a training data set, resulted in R2 values ranging from 0.15 to 0.70. The error of prediction (RSE) in the test data set was 0.08 (±10.25% of mean), 0.05 (±5.2% of mean), 0.02 (±11.11% of mean), 0.05 (±11.6% of mean) and 0.03 (±15% of mean) for nitrogen, potassium, phosphorous, calcium and magnesium, respectively, using CRDR. When data was partitioned into species groups, the R2 increased significantly to >0.80. With high-quality radiometric and geometric calibration of hyperspectral imagery, the techniques applied in this study (i.e. continuum removal on absorption features) may also be applied on data acquired by airborne and spaceborne imaging spectrometers to predict and ultimately to map the concentration of macronutrients in tropical rangelands
Integrating Imaging spectrometry and Neural Networks to map tropical grass quality in the Kruger National Park, South Africa
Mutanga, O. ; Skidmore, A.K. - \ 2004
Remote Sensing of Environment 90 (2004)1. - ISSN 0034-4257 - p. 104 - 115.
land-cover classification - chlorophyll content - red edge - absorption features - feature-selection - vegetation types - nitrogen - leaf - savanna - spectrometry
A new integrated approach, involving continuum-removed absorption features, the red edge position and neural networks, is developed and applied to map grass nitrogen concentration in an African savanna rangeland. Nitrogen, which largely determines the nutritional quality of grasslands, is commonly the most limiting nutrient for grazers. Therefore, the remote sensing of foliar nitrogen concentration in savanna rangelands is important for an improved understanding of the distribution and feeding patterns of wildlife. Continuum removal was applied on two absorption features located in the visible (R550-757) and the SWIR (R2015-2199) from an atmospherically corrected HYMAP MKI image. A feature selection algorithm was used to select wavelength variables from the absorption features. Selected band depths from the absorption features as well as the red edge position (REP) were input into a backpropagation neural network. The best-trained neural network was used to map nitrogen concentration over the whole study area. Results indicate that the new integrated approach could explain 60% of the variation in savanna grass nitrogen concentration on an independent test data set, with a root mean square error (rmse) of 0.13 (+/- 8.30% of the mean observed nitrogen concentration). This result is better compared to the result obtained using multiple linear regression, which yielded an R-2 of 38%, with a RMSE of 0.16 (+/- 10.30% of the mean observed nitrogen concentration) on an independent test data set. The study demonstrates the potential of airborne hyperspectral data and neural networks to estimate and ultimately to map nitrogen concentration in the mixed species environments of Southern Africa. (C) 2004 Elsevier Inc. All rights reserved.
Early detection of drought stress in grass swards with imaging spectroscopy
Schut, A.G.T. ; Ketelaars, J.J.M.H. - \ 2003
NJAS Wageningen Journal of Life Sciences 51 (2003)3. - ISSN 1573-5214 - p. 319 - 337.
grasveld - lolium perenne - droogte - stress - spectroscopie - spectrometrie - reflectiefactor - grass sward - lolium perenne - drought - stress - spectroscopy - spectrometry - reflectance - grown perennial ryegrass - leaf water status - spectral reflectance - red edge - leaves - vegetation - cotton
The potential of an experimental imaging spectroscopy system with high spatial (0.28-1.45 mm2) and spectral (5-13 nm) resolution was explored for early detection of drought stress in grass. A climate chamber experiment was conducted with nine Lolium perenne L. mini swards with drought stress treatments at two nitrogen levels. Images were recorded once every two days. Growth was monitored by changes in ground cover (GC), index of reflection intensity (IRI) and wavelength position of and gradient at inflection points, as estimated from images. Drought stress increased leaf dry matter and sugar content. Drought stress decelerated and ultimately reversed GC evolution, and kept IRI at low values. In contrast to unstressed growth, all absorption features narrowed and became shallower under drought stress. The inflection points near 1390 and 1500 nm were most sensitive to drought stress. Differences between drought stress and control swards were detected shortly before leaf water content dropped below 80%. The evolution of inflection point wavelength positions reversed under drought stress, except for the inflection point at the red edge where the shift to longer wavelengths during growth accelerated. The relation between inflection points at 705 and 1390 nm differentiated unstressed swards at an early growth stage from drought-stressed swards in a later growth stage
Imaging spectroscopy for early detection of nitrogen deficiency in grass swards
Schut, A.G.T. ; Ketelaars, J.J.M.H. - \ 2003
NJAS Wageningen Journal of Life Sciences 51 (2003)3. - ISSN 1573-5214 - p. 297 - 317.
grasveld - lolium perenne - mineraaltekorten - stikstof - spectroscopie - stress - reflectiefactor - grass sward - lolium perenne - mineral deficiencies - nitrogen - spectroscopy - stress - reflectance - red edge - canopy reflectance - thylakoid proteins - leaf reflectance - corn leaves - chlorophyll - carotenoids - spectra - growth - plants
The potential of an experimental imaging spectroscopy system with high spatial (0.16–0.28 mm²) ) and spectral resolution (5–13 nm) was explored for early detection of nitrogen (N) stress. From June through October 2000, a greenhouse experiment was conducted with 15 Lolium perenne L. mini-swards and 5 N treatments. Images were recorded twice a week. With the experimental system, spectra of grass leaves in the canopy can be obtained. Treatment effects on ground cover (GC) and changes in leaf spectral characteristics were studied separately. Leaf pixels with similar reflection intensity were grouped in intensity classes (IC). An index of reflection intensity (IRI) indicates the percentages of strongly reflecting grass pixels. Blue edge, green edge and red edge positions were calculated for each IC. Both GC and IRI increased until harvest, with largest increases for liberal N treatments. The width of the chlorophylldominated absorption band around 680 nm (CAW) increased up to a maximum of 133 nm for both liberal and limited N in the first two weeks after harvesting. CAW decreased for limited N in the second half of the growth period in contrast to liberal N. At harvest CAW explained 95% of the variation in relative dry matter (DM) yield between treatments. Principal component analyses showed an intertwined response of the principal components to both DM yield and N content. Edge positions changed strongly with IC. Possible effects of sensor characteristics, canopy geometry, leaf angle and changes in leaf characteristics with canopy position on the observed relation between IC and edge position are discussed.