Detection of Annual Spruce Budworm Defoliation and Severity Classification Using Landsat Imagery

Spruce budworm (SBW) is the most destructive forest pest in eastern forests of North America. Mapping annual current-year SBW defoliation is challenging because of the large landscape scale of infestations, high temporal/spatial variability, and the short period of time when detection is possible. We used Landsat-5 and Landsat-MSS data to develop a method to detect and map SBW defoliation, which can be used as ancillary or alternative information for aerial sketch maps (ASMs). Results indicated that Landsat-5 data were capable of detecting and classifying SBW defoliation into three levels comparable to ASMs. For SBW defoliation classification, a combination of three vegetation indices, including normalized difference moisture index (NDMI), enhanced vegetation index (EVI), and normalized difference vegetation index (NDVI), were found to provide the highest accuracy (non-defoliated: 77%, light defoliation: 60%, moderate defoliation: 52%, and severe defoliation: 77%) compared to using only NDMI (non-defoliated: 76%, light defoliation: 40%, moderate defoliation: 43%, and severe defoliation: 67%). Detection of historical SBW defoliation was possible using Landsat-MSS NDVI data, and the produced maps were used to complement coarse-resolution aerial sketch maps of the past outbreak. The method developed for Landsat-5 data can be used for current SBW outbreak mapping in North America using Landsat-8 and Sentinel-2 imagery. Overall, the work highlights the potential of moderate resolution optical remote sensing data to detect and classify fine-scale patterns in tree defoliation.

[1]  J. Vogelmann,et al.  Monitoring forest changes in the southwestern United States using multitemporal Landsat data , 2009 .

[2]  M. Hardisky.  The Influence of Soil Salinity, Growth Form, and Leaf Moisture on-the Spectral Radiance of Spartina alterniflora Canopies , 2008 .

[3]  David W. Williams,et al.  SPATIAL SYNCHRONY OF SPRUCE BUDWORM OUTBREAKS IN EASTERN NORTH AMERICA , 2000 .

[4]  J. Irons,et al.  Detection of initial damage in Norway spruce canopies using hyperspectral airborne data , 2004 .

[5]  D. Maclean,et al.  Validation of Spruce Budworm Outbreak History Developed from Aerial Sketch Mapping of Defoliation in New Brunswick , 2008 .

[6]  Donald G. Leckie,et al.  Reflectance characteristics of cumulative defoliation of balsam fir , 1988 .

[7]  Kenton W. Ross,et al.  Assessment of MODIS NDVI time series data products for detecting forest defoliation by gypsy moth outbreaks , 2011 .

[8]  Justin Braaten,et al.  Automated cloud and cloud shadow identification in Landsat MSS imagery for temperate ecosystems , 2015 .

[9]  Johan Lindström,et al.  Near real-time monitoring of insect induced defoliation in subalpine birch forests with MODIS derived NDVI , 2016 .

[10]  A. Huete,et al.  The use of vegetation indices in forested regions: issues of linearity and saturation , 1997, IGARSS'97. 1997 IEEE International Geoscience and Remote Sensing Symposium Proceedings. Remote Sensing - A Scientific Vision for Sustainable Development.

[11]  J. Régnière.  Generalized Approach to Landscape-Wide Seasonal Forecasting with Temperature-Driven Simulation Models , 1996 .

[12]  Ronald J. Hall,et al.  Change Detection Methodology for Aspen Defoliation with Landsat Mss Digital Data , 1984 .

[13]  D. Watts,et al.  Identifying Gypsy Moth Defoliation in Ohio Using Landsat Data , 2004 .

[14]  Ronald J. Hall,et al.  4 Remotely Sensed Data in the Mapping of Insect Defoliation , 2006 .

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

[16]  W. Cohen,et al.  Aerial and satellite sensor detection and classification of western spruce budworm defoliation in a subalpine forest , 1995 .

[17]  D. Maclean,et al.  Effects of stand and site characteristics on susceptibility and vulnerability of balsam fir and spruce to spruce budworm in New Brunswick , 1997 .

[18]  Jeff Sauro,et al.  Estimating Completion Rates from Small Samples Using Binomial Confidence Intervals: Comparisons and Recommendations , 2005 .

[19]  C. D. Rullán-Silva,et al.  Remote Monitoring of Forest Insect Defoliation -A Review- , 2013 .

[20]  E. Arsenault,et al.  Mapping Eastern Spruce Budworm Cumulative Defoliation Severity from Landsat and SPOT , 2009 .

[21]  Onisimo Mutanga,et al.  Spectral Discrimination of Insect Defoliation Levels in Mopane Woodland Using Hyperspectral Data , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

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

[23]  A. Weiskittel,et al.  Evaluating the influence of varying levels of spruce budworm defoliation on annualized individual tree growth and mortality in Maine, USA and New Brunswick, Canada , 2017 .

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

[25]  Lars Eklundh,et al.  Mapping insect defoliation in Scots pine with MODIS time-series data , 2009 .

[26]  Alice Deschamps,et al.  MAPPING INSECT DEFOLIATION USING MULTI-TEMPORAL LANDSAT DATA , 2007 .

[27]  J. T. Gray,et al.  Map-guided approach for the automatic detection on Landsat TM images of forest stands damaged by the spruce budworm , 1998 .

[28]  A. Viña,et al.  Remote estimation of leaf area index and green leaf biomass in maize canopies , 2003 .

[29]  Andy Liaw,et al.  Classification and Regression by randomForest , 2007 .

[30]  Clayton C. Kingdon,et al.  A general Landsat model to predict canopy defoliation in broadleaf deciduous forests , 2012 .

[31]  Michael A. Wulder,et al.  Mountain Pine Beetle Red-Attack Forest Damage Classification Using Stratified Landsat TM Data in British Columbia, Canada , 2003 .

[32]  C. Miller.  The feeding impact of spruce budworm on balsam fir , 1977 .

[33]  P. Wężyk,et al.  Estimating defoliation of Scots pine stands using machine learning methods and vegetation indices of Sentinel-2 , 2018 .

[34]  Philip A. Townsend,et al.  Estimating the effect of gypsy moth defoliation using MODIS , 2008 .

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

[36]  D. Ostaff,et al.  Patterns of balsam fir foliar production and growth in relation to defoliation by spruce budworm , 1995 .

[37]  Steven E. Franklin,et al.  Relationship between Landsat TM and SPOT vegetation indices and cumulative spruce budworm defoliation , 2008 .

[38]  J. Régnière,et al.  Analysis and use of historical patterns of spruce budworm defoliation to forecast outbreak patterns in Quebec. , 2000 .

[39]  R. Lunetta,et al.  A change detection experiment using vegetation indices. , 1998 .

[40]  Nicholas C. Coops,et al.  Digital aerial photogrammetry for assessing cumulative spruce budworm defoliation and enhancing forest inventories at a landscape-level , 2018 .

[41]  Donald G. Leckie,et al.  Analysis of high resolution multispectral MEIS imagery for spruce budworm damage assessment on a single tree basis , 1992 .

[42]  S. Sader,et al.  Evaluating the Impact of Abrupt Changes in Forest Policy and Management Practices on Landscape Dynamics: Analysis of a Landsat Image Time Series in the Atlantic Northern Forest , 2015, PloS one.

[43]  A. Huete,et al.  Overview of the radiometric and biophysical performance of the MODIS vegetation indices , 2002 .

[44]  Donald G. Leckie,et al.  Automated detection and mapping of crown discolouration caused by jack pine budworm with 2.5 m resolution multispectral imagery , 2005 .

[45]  Robert H. Fraser,et al.  Mapping insect‐induced tree defoliation and mortality using coarse spatial resolution satellite imagery , 2005 .

[46]  P. Warner.  Ordinal logistic regression , 2008, Journal of Family Planning and Reproductive Health Care.

[47]  Volker C. Radeloff,et al.  Detecting Jack Pine Budworm Defoliation Using Spectral Mixture Analysis: Separating Effects from Determinants , 1999 .

[48]  Effects of 80 years of forest management on landscape structure and pattern in the eastern Canadian boreal forest , 2015, Landscape Ecology.

[49]  D. Gray.  The relationship between climate and outbreak characteristics of the spruce budworm in eastern Canada , 2008 .

[50]  H. Morin,et al.  Fire Regime and Old-Growth Boreal Forests in Central Quebec, Canada: An Ecosystem Management Perspective , 2011 .

[51]  David A. MacLean,et al.  Effects of Spruce Budworm Outbreaks on the Productivity and Stability of Balsam Fir Forests , 1984 .

[52]  R. GrayDavid,et al.  The influence of forest composition and climate on outbreak characteristics of the spruce budworm in eastern Canada , 2013 .

[53]  Joanne C. White,et al.  Remote sensing of forest pest damage: a review and lessons learned from a Canadian perspective* , 2016, The Canadian Entomologist.

[54]  Olga V. Brovkina,et al.  Composite indicator for monitoring of Norway spruce stand decline , 2017 .

[55]  D. Ostaff,et al.  Patterns of balsam fir mortality caused by an uncontrolled spruce budworm outbreak , 1989 .

[56]  Donald G. Leckie,et al.  Sensor band selection for detecting current defoliation caused by the spruce budworm , 1988 .

[57]  S. E. Franklin,et al.  Satellite remote sensing of spruce budworm forest defoliation in Western Newfoundland , 1994 .

[58]  B. Rock,et al.  Detection of changes in leaf water content using Near- and Middle-Infrared reflectances , 1989 .

[59]  D. Maclean,et al.  Accuracy of aerial sketch-mapping estimates of spruce budworm defoliation in New Brunswick , 1996 .