Separability of coffee leaf rust infection levels with machine learning methods at Sentinel-2 MSI spectral resolutions

Coffee leaf rust (CLR) caused by the fungus Hemileia vastarix is a devastating disease in almost all coffee producing countries and remote sensing approaches have the potential to monitor the disease. This study evaluated the potential of Sentinel-2 band settings for discriminating CLR infection levels at leaf levels. Field spectra were resampled to the band settings of the Sentinel-2, and evaluated using the random forest (RF) and partial least squares discriminant analysis (PLS-DA) algorithms with and without variable optimization. Using all variables, Sentinel-2 Multispectral Imager (MSI)-derived vegetation indices achieved higher overall accuracy of 76.2% when compared to 69.8% obtained using raw spectral bands. Using the RF out-of-bag (OOB) scores, 4 spectral bands and 7 vegetation indices were identified as important variables in CLR discrimination. Using the PLS-DA Variable Importance in Projection (VIP) score, 3 Sentinel-2 spectral bands (B4, B6 and B5) and 5 vegetation indices were found to be important variables. Use of the identified variables improved the CLR discrimination accuracies to 79.4 and 82.5% for spectral bands and indices respectively when discriminated with the RF. Discrimination accuracy slightly increased through variable optimization for PLS-DA using spectral bands (63.5%) and vegetation indices (71.4%). Overall, this study showed the potential of the Sentinel 2 MSI band settings for CLR discrimination as part of crop condition assessment. Nevertheless further studies are required under field conditions.

[1]  S. Savary,et al.  Effects of crop management patterns on coffee rust epidemics , 2004 .

[2]  F. Baret,et al.  Estimating Canopy Characteristics from Remote Sensing Observations: Review of Methods and Associated Problems , 2008 .

[3]  D. Kutywayo,et al.  Bioclimatic modelling of current and projected climatic suitability of coffee (Coffea arabica) production in Zimbabwe , 2016, Regional Environmental Change.

[4]  Y. G. Prasad,et al.  Hyperspectral remote sensing of yellow mosaic severity and associated pigment losses in Vigna mungo using multinomial logistic regression models , 2013 .

[5]  A. Gitelson,et al.  Quantitative estimation of chlorophyll-a using reflectance spectra : experiments with autumn chestnut and maple leaves , 1994 .

[6]  W. E. Larson,et al.  Coincident detection of crop water stress, nitrogen status and canopy density using ground-based multispectral data. , 2000 .

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

[8]  Michael E. Schaepman,et al.  Retrieval of foliar information about plant pigment systems from high resolution spectroscopy , 2009 .

[9]  N. R. Patel,et al.  Monitoring of water stress in wheat using multispectral indices derived from Landsat-TM , 2016 .

[10]  Giles M. Foody,et al.  Feature Selection for Classification of Hyperspectral Data by SVM , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[11]  Jan G. P. W. Clevers,et al.  Remote estimation of crop and grass chlorophyll and nitrogen content using red-edge bands on Sentinel-2 and -3 , 2013, Int. J. Appl. Earth Obs. Geoinformation.

[12]  A. Skidmore,et al.  Red edge shift and biochemical content in grass canopies , 2007 .

[13]  Charis Gresser,et al.  Mugged: Poverty in Your Coffee Cup , 2004 .

[14]  B. Koetz,et al.  Capability of the Sentinel 2 mission for tropical coral reef mapping and coral bleaching detection , 2012 .

[15]  A. Skidmore,et al.  Narrow band vegetation indices overcome the saturation problem in biomass estimation , 2004 .

[16]  Thomas R. Loveland,et al.  A review of large area monitoring of land cover change using Landsat data , 2012 .

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

[18]  James K. M. Brown,et al.  Aerial Dispersal of Pathogens on the Global and Continental Scales and Its Impact on Plant Disease , 2002, Science.

[19]  Mario Chica-Olmo,et al.  An assessment of the effectiveness of a random forest classifier for land-cover classification , 2012 .

[20]  F. Baret,et al.  Potentials and limits of vegetation indices for LAI and APAR assessment , 1991 .

[21]  Ingmar Nitze,et al.  COMPARISON OF MACHINE LEARNING ALGORITHMS RANDOM FOREST, ARTIFICIAL NEURAL NETWORK AND SUPPORT VECTOR MACHINE TO MAXIMUM LIKELIHOOD FOR SUPERVISED CROP TYPE CLASSIFICATION , 2012 .

[22]  Tiziana Simoniello,et al.  Early Identification of Land Degradation Hotspots in Complex Bio-Geographic Regions , 2015, Remote. Sens..

[23]  D. Cressey.  Coffee rust regains foothold , 2013, Nature.

[24]  N. Coops,et al.  Assessment of Dothistroma Needle Blight of Pinus radiata Using Airborne Hyperspectral Imagery. , 2003, Phytopathology.

[25]  Ronei J. Poppi,et al.  Discrimination between authentic and counterfeit banknotes using Raman spectroscopy and PLS-DA with uncertainty estimation , 2013 .

[26]  Jayme Garcia Arnal Barbedo,et al.  Digital image processing techniques for detecting, quantifying and classifying plant diseases. , 2013 .

[27]  L. Maffia,et al.  Biological control of coffee rust by antagonistic bacteria under field conditions in Brazil , 2009 .

[28]  Michael E. Schaepman,et al.  Experimental Evaluation of Sentinel-2 Spectral Response Functions for NDVI Time-Series Continuity , 2013, IEEE Transactions on Geoscience and Remote Sensing.

[29]  Markus Reichstein,et al.  Similarities in ground- and satellite-based NDVI time series and their relationship to physiological activity of a Scots pine forest in Finland , 2004 .

[30]  P. Mecocci,et al.  Random Forest ensembles for detection and prediction of Alzheimer's disease with a good between-cohort robustness , 2014, NeuroImage: Clinical.

[31]  Lie Deng,et al.  Identification of pummelo cultivars by using Vis/NIR spectra and pattern recognition methods , 2015, Precision Agriculture.

[32]  A. Gitelson,et al.  Use of a green channel in remote sensing of global vegetation from EOS- MODIS , 1996 .

[33]  Jacques Avelino,et al.  Landscape context and scale differentially impact coffee leaf rust, coffee berry borer, and coffee root-knot nematodes. , 2012, Ecological applications : a publication of the Ecological Society of America.

[34]  J. Eitel,et al.  Suitability of existing and novel spectral indices to remotely detect water stress in Populus spp. , 2006 .

[35]  Nicholas C. Coops,et al.  Spectral reflectance characteristics of eucalypt foliage damaged by insects , 2001 .

[36]  S. Simpfendorfer,et al.  Effect of stripe rust on the yield response of wheat to nitrogen , 2014 .

[37]  J. A. Schell,et al.  Monitoring vegetation systems in the great plains with ERTS , 1973 .

[38]  S. G. Nelson,et al.  Relationship Between Remotely-sensed Vegetation Indices, Canopy Attributes and Plant Physiological Processes: What Vegetation Indices Can and Cannot Tell Us About the Landscape , 2008, Sensors.

[39]  Onisimo Mutanga,et al.  Developing detailed age-specific thematic maps for coffee (Coffea arabica L.) in heterogeneous agricultural landscapes using random forests applied on Landsat 8 multispectral sensor , 2017 .

[40]  Clement Atzberger,et al.  Using Low Resolution Satellite Imagery for Yield Prediction and Yield Anomaly Detection , 2013, Remote. Sens..

[41]  Alan A. Ager,et al.  Broadband, red-edge information from satellites improves early stress detection in a New Mexico conifer woodland , 2011 .

[42]  B. Bertrand,et al.  Coffee resistance to the main diseases: leaf rust and coffee berry disease , 2006 .

[43]  M. Shivanna,et al.  IDENTIFICATION OF RAPD (RANDOM AMPLIFIED POLYMORPHIC DNA) MARKERS FOR ETHIOPIAN WILD COFFEA ARABICA L. GENETIC RESOURCES CONSERVED IN INDIA , 2011 .

[44]  G. Carter,et al.  Leaf optical properties in higher plants: linking spectral characteristics to stress and chlorophyll concentration. , 2001, American journal of botany.

[45]  Z. Niu,et al.  Identification of yellow rust in wheat using in-situ spectral reflectance measurements and airborne hyperspectral imaging , 2007, Precision Agriculture.

[46]  M. Hill.  Vegetation index suites as indicators of vegetation state in grassland and savanna: An analysis with simulated SENTINEL 2 data for a North American transect , 2013 .

[47]  Gary R. Watmough,et al.  Evaluating the capabilities of Sentinel-2 for quantitative estimation of biophysical variables in vegetation , 2013 .

[48]  Onisimo Mutanga,et al.  High density biomass estimation for wetland vegetation using WorldView-2 imagery and random forest regression algorithm , 2012, Int. J. Appl. Earth Obs. Geoinformation.

[49]  Johannes R. Sveinsson,et al.  Random Forest classification of multisource remote sensing and geographic data , 2004, IGARSS 2004. 2004 IEEE International Geoscience and Remote Sensing Symposium.

[50]  O. Mutanga,et al.  Evaluating the utility of the medium-spatial resolution Landsat 8 multispectral sensor in quantifying aboveground biomass in uMgeni catchment, South Africa , 2015 .

[51]  Bardan Ghimire,et al.  An Evaluation of Bagging, Boosting, and Random Forests for Land-Cover Classification in Cape Cod, Massachusetts, USA , 2012 .

[52]  Reza Ehsani,et al.  Review: A review of advanced techniques for detecting plant diseases , 2010 .

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

[54]  Stefano Amaducci,et al.  Empirical Estimation of Leaf Chlorophyll Density in Winter Wheat Canopies Using Sentinel-2 Spectral Resolution , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[55]  Janet Franklin,et al.  Mapping land-cover modifications over large areas: A comparison of machine learning algorithms , 2008 .

[56]  Rainer Laudien,et al.  COMPARISON OF REMOTE SENSING BASED ANALYSIS OF CROP DISEASES BY USING HIGH RESOLUTION MULTISPECTRAL AND HYPERSPECTRAL DATA - CASE STUDY: RHIZOCTONIA SOLANI IN SUGAR BEET - , 2004 .

[57]  L. Zambolim,et al.  Photosynthetic and antioxidative alterations in coffee leaves caused by epoxiconazole and pyraclostrobin sprays and Hemileia vastatrix infection. , 2015, Pesticide biochemistry and physiology.

[58]  Li He,et al.  Improved remote sensing detection of wheat powdery mildew using dual-green vegetation indices , 2016, Precision Agriculture.

[59]  L. Plümer,et al.  Development of spectral indices for detecting and identifying plant diseases , 2013 .

[60]  Zhengwei Yang,et al.  Phenology-Based Assessment of Perennial Energy Crops in North American Tallgrass Prairie , 2011 .

[61]  S. Ram,et al.  COFFEE LEAF RUST ( CLR ) AND DISEASE TRIANGLE : A CASE STUDY , 2012 .

[62]  A. Viña,et al.  Remote estimation of canopy chlorophyll content in crops , 2005 .

[63]  Raquel Ghini,et al.  Diseases in tropical and plantation crops as affected by climate changes: current knowledge and perspectives , 2011 .

[64]  Alfred Stein,et al.  Characterising and quantifying vegetative drought in East Africa using fuzzy modelling and NDVI data , 2012 .

[65]  Pedro Larrañaga,et al.  A review of feature selection techniques in bioinformatics , 2007, Bioinform..

[66]  D. Kutywayo,et al.  The Impact of Climate Change on the Potential Distribution of Agricultural Pests: The Case of the Coffee White Stem Borer (Monochamus leuconotus P.) in Zimbabwe , 2013, PloS one.

[67]  J. Steinier,et al.  Smoothing and differentiation of data by simplified least square procedure. , 1972, Analytical chemistry.

[68]  G. Foody.  Thematic map comparison: Evaluating the statistical significance of differences in classification accuracy , 2004 .

[69]  A. Savitzky,et al.  Smoothing and Differentiation of Data by Simplified Least Squares Procedures. , 1964 .