Convolutional Neural Network Approach for Mapping Arctic Vegetation Using Multi-Sensor Remote Sensing Fusion

Accurate and high-resolution maps of vegetation are critical for projects seeking to understand the terrestrial ecosystem processes and land-atmosphere interactions in Arctic ecosystems, such as U.S. Department of Energy's Next Generation Ecosystem Experiment (NGEE) Arctic. However, most existing Arctic vegetation maps are at a coarse resolution and with a varying degree of detail and accuracy. Remote sensing-based approaches for mapping vegetation, while promising, are challenging in high latitude environments due to frequent cloud cover, polar darkness, and limited availability of high-resolution remote sensing datasets (e.g., ∼ 5 m). This study proposes a new remote sensing based multi-sensor data fusion approach for developing high-resolution maps of vegetation in the Seward Peninsula, Alaska. We focus detailed analysis and validation study around the Kougarok river, located in the central Seward Peninsula of Alaska. We seek to evaluate the integration of hyper-spectral, multi-spectral, radar, and terrain datasets using unsupervised and supervised classification techniques over a ∼343.72 km 2 area for generating vegetation classifications at a variety of resolutions (5 m and 12.5 m). We fist applied a quantitative goodness-of-fit method, called Mapcurves, that shows the degree of spatial concordance between the public coarse resolution maps and k-means clustering values and relabels the k values based on the best overlap. We develop a convolutional neural network (CNN) approach for developing high resolution vegetation maps for our study region in Arctic. We compare two CNN approaches: (1) breaking up the images into small patches (e.g., 6 x 6) and predict the vegetation class for entire patch and (2) semantic segmentation and predict the vegetation class for every pixel. We also perform accuracy assessments of the developed data products and evaluate varying CNN architectures. The fusion of hyperspectral and optical datasets performed the best, with accuracy values increased from 0.64 to 0.96-0.97 when using a training map produced by unsupervised clustering and Mapcurves labeling for both CNN models.

[1]  Yuzhen Ye.  Clustering Algorithms , 2021, Wireless RF Energy Transfer in the Massive IoT Era.

[2]  Peerapon Vateekul,et al.  Road Segmentation of Remotely-Sensed Images Using Deep Convolutional Neural Networks with Landscape Metrics and Conditional Random Fields , 2017, Remote. Sens..

[3]  Yong Dou,et al.  Airport Detection on Optical Satellite Images Using Deep Convolutional Neural Networks , 2017, IEEE Geoscience and Remote Sensing Letters.

[4]  Rob Heylen,et al.  Spatial Resolution Enhancement of Hyperspectral Images Using Spectral Unmixing and Bayesian Sparse Representation , 2017, Remote. Sens..

[5]  Gang Fu,et al.  Classification for High Resolution Remote Sensing Imagery Using a Fully Convolutional Network , 2017, Remote. Sens..

[6]  Hao Wu,et al.  Convolutional Recurrent Neural Networks forHyperspectral Data Classification , 2017, Remote. Sens..

[7]  Donatella Zona,et al.  Numerical Terradynamic Simulation Group 11-2016 Mapping Arctic Tundra Vegetation Communities Using Field Spectroscopy and Multispectral Satellite Data in North Alaska , USA , 2017 .

[8]  Nathan Collier,et al.  Modeling the spatiotemporal variability in subsurface thermal regimes across a low-relief polygonal tundra landscape , 2016 .

[9]  Jitendra Kumar,et al.  Mapping Arctic Plant Functional Type Distributions in the Barrow Environmental Observatory Using WorldView-2 and LiDAR Datasets , 2016, Remote. Sens..

[10]  Monique Bernier,et al.  Land Cover Classification in SubArctic Regions Using Fully Polarimetric RADARSAT-2 Data , 2016, Remote. Sens..

[11]  Bo Du,et al.  Deep Learning for Remote Sensing Data: A Technical Tutorial on the State of the Art , 2016, IEEE Geoscience and Remote Sensing Magazine.

[12]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[13]  Pernille Bronken.  The regional species richness and genetic diversity of Arctic vegetation reflect both past glaciations and current climate , 2016 .

[14]  Howard E. Epstein,et al.  Differentiating among Four Arctic Tundra Plant Communities at Ivotuk, Alaska Using Field Spectroscopy , 2016, Remote. Sens..

[15]  Radomír Mech,et al.  Deep Multi-patch Aggregation Network for Image Style, Aesthetics, and Quality Estimation , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[16]  Sergey Ioffe,et al.  Rethinking the Inception Architecture for Computer Vision , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[17]  Jitendra Kumar,et al.  Characterization and Classification of Vegetation Canopy Structure and Distribution within the Great Smoky Mountains National Park Using LiDAR , 2015, 2015 IEEE International Conference on Data Mining Workshop (ICDMW).

[18]  Claudio Cusano,et al.  Single and Multiple Illuminant Estimation Using Convolutional Neural Networks , 2015, IEEE Transactions on Image Processing.

[19]  Masami Fukuda,et al.  Vegetation and Permafrost Thaw Depth 10 Years after a Tundra Fire in 2002, Seward Peninsula, Alaska , 2015 .

[20]  Thomas Brox,et al.  U-Net: Convolutional Networks for Biomedical Image Segmentation , 2015, MICCAI.

[21]  Lesley Evans Ogden,et al.  Plants Duke It Out in a Warming Arctic , 2015 .

[22]  Xiaoou Tang,et al.  Image Super-Resolution Using Deep Convolutional Networks , 2014, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[23]  D. M. Lawrence,et al.  Climate change and the permafrost carbon feedback , 2014, Nature.

[24]  Jitendra Malik,et al.  Hypercolumns for object segmentation and fine-grained localization , 2014, 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[25]  Gang Wang,et al.  Deep Learning-Based Classification of Hyperspectral Data , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[26]  John E. Walsh,et al.  Future Arctic climate changes: Adaptation and mitigation time scales , 2014 .

[27]  Rob Fergus,et al.  Visualizing and Understanding Convolutional Networks , 2013, ECCV.

[28]  Benjamin Smith,et al.  Tundra shrubification and tree-line advance amplify arctic climate warming: results from an individual-based dynamic vegetation model , 2013 .

[29]  Richard T. Mills,et al.  Representativeness-based sampling network design for the State of Alaska , 2013, Landscape Ecology.

[30]  Miriam C. Jones,et al.  Lateglacial and Holocene climate, disturbance and permafrost peatland dynamics on the Seward Peninsula, western Alaska , 2013 .

[31]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[32]  Benjamin Smith,et al.  Modelling Tundra Vegetation Response to Recent Arctic Warming , 2012, AMBIO.

[33]  R. Knutti,et al.  September Arctic sea ice predicted to disappear near 2°C global warming above present , 2011 .

[34]  J. G. Lyon,et al.  Hyperspectral Remote Sensing of Vegetation , 2011 .

[35]  E. Pottier,et al.  Polarimetric Radar Imaging: From Basics to Applications , 2009 .

[36]  Paul F. Hessburg,et al.  Mapcurves: a quantitative method for comparing categorical maps , 2006, J. Geogr. Syst..

[37]  F. Chapin,et al.  Evidence and Implications of Recent Climate Change in Northern Alaska and Other Arctic Regions , 2004 .

[38]  C. Racine,et al.  Tundra Fire and Vegetation Change along a Hillslope on the Seward Peninsula, Alaska, U.S.A , 2004 .

[39]  A. McGuire,et al.  Land Cover Change on the Seward Peninsula: The Use of Remote Sensing to Evaluate the Potential Influences of Climate Warming on Historical Vegetation Dynamics , 2001 .

[40]  William W. Hargrove,et al.  Multivariate Geographic Cluster Using a Beowulf-style Parallel Computer , 1999, PDPTA.

[41]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[42]  L. D. Hinzman,et al.  Hydrological variations among watersheds with varying degrees of permafrost , 2002 .