Optimizing spectral resolutions for the classification of C 3 and C 4 grass species, using wavelengths of known absorption features

Abstract. Hyperspectral remote-sensing approaches are suitable for detection of the differences in 3-carbon ( C 3 ) and four carbon ( C 4 ) grass species phenology and composition. However, the application of hyperspectral sensors to vegetation has been hampered by high-dimensionality, spectral redundancy, and multicollinearity problems. In this experiment, resampling of hyperspectral data to wider wavelength intervals, around a few band-centers, sensitive to the biophysical and biochemical properties of C 3 or C 4 grass species is proposed. The approach accounts for an inherent property of vegetation spectral response: the asymmetrical nature of the inter-band correlations between a waveband and its shorter- and longer-wavelength neighbors. It involves constructing a curve of weighting threshold of correlation (Pearson’s r ) between a chosen band-center and its neighbors, as a function of wavelength. In addition, data were resampled to some multispectral sensors—ASTER, GeoEye-1, IKONOS, QuickBird, RapidEye, SPOT 5, and WorldView-2 satellites—for comparative purposes, with the proposed method. The resulting datasets were analyzed, using the random forest algorithm. The proposed resampling method achieved improved classification accuracy ( κ = 0.82 ), compared to the resampled multispectral datasets ( κ = 0.78 , 0.65, 0.62, 0.59, 0.65, 0.62, 0.76, respectively). Overall, results from this study demonstrated that spectral resolutions for C 3 and C 4 grasses can be optimized and controlled for high dimensionality and multicollinearity problems, yet yielding high classification accuracies. The findings also provide a sound basis for programming wavebands for future sensors.

[1]  Achim Zeileis,et al.  Conditional variable importance for random forests , 2008, BMC Bioinformatics.

[2]  G. Henebry,et al.  A technique for monitoring ecological disturbance in tallgrass prairie using seasonal NDVI trajectories and a discriminant function mixture model , 1997 .

[3]  W. Bond,et al.  Will global change improve grazing quality of grasslands? A call for a deeper understanding of the effects of shifts from C4 to C3 grasses for large herbivores , 2010 .

[4]  J. C. Price.  How unique are spectral signatures , 1994 .

[5]  D. Killick.  An account of the plant ecology of the Cathedral Peak area of the Natal Drakensberg. , 1963 .

[6]  F. Csillag,et al.  The Influence of Vegetation Index and Spatial Resolution on a Two-Date Remote Sensing-Derived Relation to C4 Species Coverage , 2001 .

[7]  A. Skidmore,et al.  Spectral discrimination of vegetation types in a coastal wetland , 2003 .

[8]  Onisimo Mutanga,et al.  Spectral resampling based on user-defined inter-band correlation filter: C3 and C4 grass species classification , 2013, Int. J. Appl. Earth Obs. Geoinformation.

[9]  D. Sims,et al.  Relationships between leaf pigment content and spectral reflectance across a wide range of species, leaf structures and developmental stages , 2002 .

[10]  Onisimo Mutanga,et al.  Examining the utility of random forest and AISA Eagle hyperspectral image data to predict Pinus patula age in KwaZulu-Natal, South Africa , 2011 .

[11]  G. A. Blackburn,et al.  Quantifying Chlorophylls and Caroteniods at Leaf and Canopy Scales: An Evaluation of Some Hyperspectral Approaches , 1998 .

[12]  S. Tarantola,et al.  Detecting vegetation leaf water content using reflectance in the optical domain , 2001 .

[13]  Rasmus Fensholt,et al.  Detecting Canopy Water Status Using Shortwave Infrared Reflectance Data From Polar Orbiting and Geostationary Platforms , 2010, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[14]  Nasir D. Memon,et al.  Context-based lossless interband compression-extending CALIC , 2000, IEEE Trans. Image Process..

[15]  P. M. Hansena,et al.  Reflectance measurement of canopy biomass and nitrogen status in wheat crops using normalized difference vegetation indices and partial least squares regression , 2003 .

[16]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[17]  Onisimo Mutanga,et al.  Field spectrometry of papyrus vegetation (Cyperus papyrus L.) in swamp wetlands of St Lucia, South Africa , 2009, 2009 IEEE International Geoscience and Remote Sensing Symposium.

[18]  Achim Zeileis,et al.  Bias in random forest variable importance measures: Illustrations, sources and a solution , 2007, BMC Bioinformatics.

[19]  A. Smith,et al.  Weed–Crop Discrimination Using Remote Sensing: A Detached Leaf Experiment1 , 2003, Weed Technology.

[20]  Jennifer A. Miller,et al.  Contextual land-cover classification: incorporating spatial dependence in land-cover classification models using random forests and the Getis statistic , 2010 .

[21]  Coleen Vogel,et al.  Climate Change Impacts on African Rangelands , 2008 .

[22]  P. Donnelly,et al.  Quantitative Leaf Anatomy of C3 and C4 Grasses (Poaceae): Bundle Sheath and Mesophyll Surface Area Relationships , 1994 .

[23]  Mahesh Pal,et al.  Random forest classifier for remote sensing classification , 2005 .

[24]  Joydeep Ghosh,et al.  Random forests of binary hierarchical classifiers for analysis of hyperspectral data , 2003, IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003.

[25]  Kiona Ogle,et al.  Implications of interveinal distance for quantum yield in C4 grasses: a modeling and meta-analysis , 2003, Oecologia.

[26]  G. Carter.  Ratios of leaf reflectances in narrow wavebands as indicators of plant stress , 1994 .

[27]  B. Wylie,et al.  NDVI, C3 AND C4 PRODUCTION, AND DISTRIBUTIONS IN GREAT PLAINS GRASSLAND LAND COVER CLASSES , 1997 .

[28]  Liangyun Liu,et al.  Mapping C3 and C4 plant functional types using separated solar-induced chlorophyll fluorescence from hyperspectral data , 2011 .

[29]  Roberta E. Martin,et al.  Predicting tropical plant physiology from leaf and canopy spectroscopy , 2009, Oecologia.

[30]  R. Sage,et al.  The Nitrogen Use Efficiency of C(3) and C(4) Plants: II. Leaf Nitrogen Effects on the Gas Exchange Characteristics of Chenopodium album (L.) and Amaranthus retroflexus (L.). , 1987, Plant physiology.

[31]  M. Schlossberg,et al.  An evaluation of multi-spectral responses on selected turfgrass species , 2000 .

[32]  R. Schulze,et al.  Incoming solar radiation patterns and vegetation response: Examples from the natal drakensberg , 1977, Vegetatio.

[33]  J. Gamon,et al.  The photochemical reflectance index: an optical indicator of photosynthetic radiation use efficiency across species, functional types, and nutrient levels , 1997, Oecologia.

[34]  A. Prasad,et al.  Newer Classification and Regression Tree Techniques: Bagging and Random Forests for Ecological Prediction , 2006, Ecosystems.

[35]  Giles M. Foody,et al.  Discriminating and mapping the C3 and C4 composition of grasslands in the northern Great Plains, USA , 2007, Ecol. Informatics.

[36]  Joseph M. Craine,et al.  Grazing and landscape controls on nitrogen availability across 330 South African savanna sites , 2009 .

[37]  M. Ashton,et al.  Accuracy assessments of hyperspectral waveband performance for vegetation analysis applications , 2004 .

[38]  P. Curran.  Remote sensing of foliar chemistry , 1989 .

[39]  P. Thenkabail,et al.  Hyperspectral Vegetation Indices and Their Relationships with Agricultural Crop Characteristics , 2000 .

[40]  A. Skidmore,et al.  Discriminating tropical grass (Cenchrus ciliaris) canopies grown under different nitrogen treatments using spectroradiometry , 2003 .

[41]  Jonathan Cheung-Wai Chan,et al.  Evaluation of random forest and adaboost tree-based ensemble classification and spectral band selection for ecotope mapping using airborne hyperspectral imagery , 2008 .

[42]  G. Asner.  Biophysical and Biochemical Sources of Variability in Canopy Reflectance , 1998 .

[43]  G. F. Hughes,et al.  On the mean accuracy of statistical pattern recognizers , 1968, IEEE Trans. Inf. Theory.

[44]  Lorenzo Bruzzone,et al.  Semisupervised Classification of Hyperspectral Images by SVMs Optimized in the Primal , 2007, IEEE Transactions on Geoscience and Remote Sensing.

[45]  D. Horler,et al.  The red edge of plant leaf reflectance , 1983 .

[46]  J. Kattge,et al.  Improving assessment and modelling of climate change impacts on global terrestrial biodiversity. , 2011, Trends in ecology & evolution.

[47]  Lorenzo Bruzzone,et al.  On the role of spectral resolution and classifier complexity in the analysis of hyperspectral images of forest areas , 2007, SPIE Remote Sensing.

[48]  E. Hunt,et al.  Estimating near-infrared leaf reflectance from leaf structural characteristics. , 2001, American journal of botany.

[49]  Lorenzo Bruzzone,et al.  Classification of hyperspectral remote sensing images with support vector machines , 2004, IEEE Transactions on Geoscience and Remote Sensing.

[50]  Lalit Kumar,et al.  Optimal band selection from hyperspectral data for Lantana camara discrimination , 2012 .

[51]  José M. Paruelo,et al.  Grass species differentiation through canopy hyperspectral reflectance , 2009 .

[52]  Trevor R. Hill,et al.  Description, classification and ordination of the dominant vegetation communities, Cathedral Peak, KwaZulu-Natal Drakensberg , 1996 .

[53]  A. Skidmore,et al.  Nitrogen detection with hyperspectral normalized ratio indices across multiple plant species , 2005 .

[54]  T. Faurtyot.  Vegetation water and dry matter contents estimated from top-of-the-atmosphere reflectance data: A simulation study , 1997 .

[55]  Lalit Kumar,et al.  Mapping Coastal Vegetation Using an Expert System and Hyperspectral Imagery , 2004 .

[56]  José M. Paruelo,et al.  Trait differences between grass species along a climatic gradient in South and North America , 2008 .

[57]  Guizhong Liu,et al.  An Efficient Compression Algorithm for Hyperspectral Images Based on Correlation Coefficients Adaptive Three Dimensional Wavelet Zerotree Coding , 2007, 2007 IEEE International Conference on Image Processing.

[58]  Gregory Asner,et al.  Improving Discrimination of Savanna Tree Species Through a Multiple-Endmember Spectral Angle Mapper Approach: Canopy-Level Analysis , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[59]  Carolin Strobl,et al.  Unbiased split selection for classification trees based on the Gini Index , 2007, Comput. Stat. Data Anal..

[60]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[61]  Andrew K. Skidmore,et al.  Hyperspectral predictors for monitoring biomass production in Mediterranean mountain grasslands: Majella National Park, Italy , 2009 .

[62]  Rick L. Lawrence,et al.  Mapping invasive plants using hyperspectral imagery and Breiman Cutler classifications (RandomForest) , 2006 .

[63]  J. De Baerdemaeker,et al.  Weed Detection Using Canopy Reflection , 2002, Precision Agriculture.

[64]  O. Ghannoum,et al.  C4 photosynthesis and water stress. , 2008, Annals of botany.

[65]  Onisimo Mutanga,et al.  Discriminating the early stages of Sirex noctilio infestation using classification tree ensembles and shortwave infrared bands , 2011 .