Time-Varying Modeling of Land Cover Change Dynamics Due to Forest Fires

Seasonal variations in land cover are commonly represented using a constant frequency cosine model with time-varying parameters. As frequency represents the constant annual vegetation growth cycle, the model is not adequate to represent dynamics such as sudden changes in land cover and subsequent regrowth. In this paper, we present a new model to capture time-varying changes in the vegetation growth cycle and detect abrupt changes in land cover due to forest fires. We also design a sequential Monte Carlo estimation approach of the time-varying frequency in the proposed nonlinear model using the particle filter (PF). We further propose a binary hypothesis land cover change detector that is based on a dissimilarity measure between windowed time-series observed during the same month of consecutive years. Experiments show that the PF estimation can detect change with lower delay than the existing approaches. Unsupervised mapping of the fire severity from the model parameter estimates is also developed.

[1]  D. Legates,et al.  Crop identification using harmonic analysis of time-series AVHRR NDVI data , 2002 .

[2]  Tom Fawcett,et al.  Robust Classification for Imprecise Environments , 2000, Machine Learning.

[3]  Charles Elkan,et al.  The Foundations of Cost-Sensitive Learning , 2001, IJCAI.

[4]  Tomás Martínez-Marín,et al.  Particle filter approach for crop phenological stage estimation using time series of NDVI images , 2015, 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[5]  Chao Chen,et al.  Using Random Forest to Learn Imbalanced Data , 2004 .

[6]  Jagannath Aryal,et al.  Near-Real Time Detection of Beetle Infestation in Pine Forests Using MODIS Data , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[7]  Waldo Kleynhans,et al.  Meta-Optimization of the Extended Kalman Filter’s Parameters Through the Use of the Bias Variance Equilibrium Point Criterion , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[8]  Waldo Kleynhans,et al.  Land Cover Change Detection Using the Internal Covariance Matrix of the Extended Kalman Filter Over Multiple Spectral Bands , 2013, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[9]  Hugo Carrão,et al.  A Nonlinear Harmonic Model for Fitting Satellite Image Time Series: Analysis and Prediction of Land Cover Dynamics , 2010, IEEE Transactions on Geoscience and Remote Sensing.

[10]  Neil J. Gordon,et al.  A tutorial on particle filters for online nonlinear/non-Gaussian Bayesian tracking , 2002, IEEE Trans. Signal Process..

[11]  Jan Verbesselt,et al.  Hierarchical image segmentation based on similarity of NDVI time series , 2008 .

[12]  Waldo Kleynhans,et al.  Detecting Land Cover Change Using an Extended Kalman Filter on MODIS NDVI Time-Series Data , 2011, IEEE Geoscience and Remote Sensing Letters.

[13]  Mark Noonan,et al.  The post-fire measurement of fire severity and intensity in the Christmas 2001 Sydney wildfires , 2004 .

[14]  F. Hlawatsch,et al.  Linear and quadratic time-frequency signal representations , 1992, IEEE Signal Processing Magazine.

[15]  B. Quayle,et al.  A Project for Monitoring Trends in Burn Severity , 2007 .

[16]  J. Keeley.  Fire intensity, fire severity and burn severity: a brief review and suggested usage , 2009 .

[17]  Waldo Kleynhans,et al.  Improving Land Cover Class Separation Using an Extended Kalman Filter on MODIS NDVI Time-Series Data , 2010, IEEE Geoscience and Remote Sensing Letters.

[18]  W. Verstraeten,et al.  A comparison of time series similarity measures for classification and change detection of ecosystem dynamics , 2011 .

[19]  Stan Matwin,et al.  Learning When Negative Examples Abound , 1997, ECML.

[20]  Waldo Kleynhans,et al.  Using Page's Cumulative Sum Test on MODIS Time Series to Detect Land-Cover Changes , 2013, IEEE Geoscience and Remote Sensing Letters.

[21]  Yves Grandvalet,et al.  Optimizing F-Measures by Cost-Sensitive Classification , 2014, NIPS.

[22]  Sergei Vassilvitskii,et al.  k-means++: the advantages of careful seeding , 2007, SODA '07.

[23]  N. C. Strugnell,et al.  First operational BRDF, albedo nadir reflectance products from MODIS , 2002 .

[24]  Vipin Kumar,et al.  Time series change detection: algorithms for land cover change , 2010 .

[25]  Sandeep K. S. Gupta,et al.  Estimation of dynamic parameters of MODIS NDVI time series nonlinear model using particle filtering , 2017, 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[26]  R. Lunetta,et al.  Land-cover change detection using multi-temporal MODIS NDVI data , 2006 .

[27]  Jagannath Aryal,et al.  A Statistical Framework for Near-Real Time Detection of Beetle Infestation in Pine Forests Using MODIS Data , 2014, IEEE Geoscience and Remote Sensing Letters.

[28]  Malgorzata M. O'Reilly,et al.  A Relative Density Ratio-Based Framework for Detection of Land Cover Changes in MODIS NDVI Time Series , 2016, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.