Nonlinear PCA for Spatio-Temporal Analysis of Earth Observation Data

Remote sensing observations, products, and simulations are fundamental sources of information to monitor our planet and its climate variability. Uncovering the main modes of spatial and temporal variability in Earth data is essential to analyze and understand the underlying physical dynamics and processes driving the Earth System. Dimensionality reduction methods can work with spatio-temporal data sets and decompose the information efficiently. Principal component analysis (PCA), also known as empirical orthogonal functions (EOFs) in geophysics, has been traditionally used to analyze climatic data. However, when nonlinear feature relations are present, PCA/EOF fails. In this article, we propose a nonlinear PCA method to deal with spatio-temporal Earth system data. The proposed method, called rotated complex kernel PCA (ROCK-PCA for short), works in reproducing kernel Hilbert spaces to account for nonlinear processes, operates in the complex kernel domain to account for both space and time features, and adds an extra rotation for improved flexibility. The result is an explicitly resolved spatio-temporal decomposition of the Earth data cube. The method is unsupervised and computationally very efficient. We illustrate its ability to uncover spatio-temporal patterns using synthetic experiments and real data. Results of the decomposition of three essential climate variables are shown: satellite-based global gross primary productivity (GPP), soil moisture (SM), and reanalysis sea surface temperature (SST) data. The ROCK-PCA method allows identifying their annual and seasonal oscillations, as well as their nonseasonal trends and spatial variability patterns. The main modes of variability of GPP and SM match expected distributions of land-cover and eco-hydrological zones, respectively; the interannual component of SM is shown to be highly correlated with El Niño Southern Oscillation (ENSO) phenomenon; and the SST annual oscillation is perfectly uncoupled in magnitude and phase from the global warming trend and ENSO anomalies, as well as from their mutual interactions. We provide the working source code of the presented method for the interested reader in https://github.com/DiegoBueso/ROCK-PCA.

[1]  P. O. White,et al.  PROMAX: A QUICK METHOD FOR ROTATION TO OBLIQUE SIMPLE STRUCTURE , 1964 .

[2]  Bernhard Schölkopf,et al.  Nonlinear Component Analysis as a Kernel Eigenvalue Problem , 1998, Neural Computation.

[3]  Gustau Camps-Valls,et al.  Uncertainty analysis of gross primary production upscaling using Random Forests, remote sensing and eddy covariance data , 2015 .

[4]  Haijun Yang,et al.  ENSO amplitude change in observation and coupled models , 2008 .

[5]  Rose Yu,et al.  Accelerated Online Low Rank Tensor Learning for Multivariate Spatiotemporal Streams , 2015, ICML.

[6]  N. Guttman,et al.  Statistical Descriptors of Climate , 1989 .

[7]  Marc Castella,et al.  A new method for kurtosis maximization and source separation , 2010, 2010 IEEE International Conference on Acoustics, Speech and Signal Processing.

[8]  F. Woodward,et al.  Terrestrial Gross Carbon Dioxide Uptake: Global Distribution and Covariation with Climate , 2010, Science.

[9]  J. Astola,et al.  HILBERT TRANSFORM OF DISCRETE DATA : A BRIEF REVIEW , 2022 .

[10]  C. Frankenberg,et al.  Satellite observation of tropical forest seasonality: spatial patterns of carbon exchange in Amazonia , 2015 .

[11]  Dylan B. A. Jones,et al.  Evaluating GPP and Respiration Estimates Over Northern Midlatitude Ecosystems Using Solar‐Induced Fluorescence and Atmospheric CO2 Measurements , 2018, Journal of Geophysical Research: Biogeosciences.

[12]  F. Prieto,et al.  Cluster Identification Using Projections , 2001 .

[13]  T. McMahon,et al.  Updated world map of the Köppen-Geiger climate classification , 2007 .

[14]  S. Brands Which ENSO teleconnections are robust to internal atmospheric variability? , 2017 .

[15]  Susan A. Murphy,et al.  Monographs on statistics and applied probability , 1990 .

[16]  Nello Cristianini,et al.  Support Vector Machines and Kernel Methods: The New Generation of Learning Machines , 2002, AI Mag..

[17]  J. Larson Can we define climate using information theory , 2010 .

[18]  Michael G. Sideris,et al.  Assessment of the capabilities of the temporal and spatiotemporal ICA method for geophysical signal separation in GRACE data , 2014 .

[19]  Kuldip K. Paliwal,et al.  Fast principal component analysis using fixed-point algorithm , 2007, Pattern Recognit. Lett..

[20]  Jürgen Kurths,et al.  Identifying causal gateways and mediators in complex spatio-temporal systems , 2015, Nature Communications.

[21]  Yehuda Bock,et al.  Spatiotemporal filtering using principal component analysis and Karhunen-Loeve expansion approaches for regional GPS network analysis , 2006 .

[22]  Jean-Francois Cardoso,et al.  Source separation using higher order moments , 1989, International Conference on Acoustics, Speech, and Signal Processing,.

[23]  H. Kaiser The varimax criterion for analytic rotation in factor analysis , 1958 .

[24]  Ian T. Jolliffe,et al.  Empirical orthogonal functions and related techniques in atmospheric science: A review , 2007 .

[25]  Lifeng Luo,et al.  The Second Phase of the Global Land–Atmosphere Coupling Experiment: Soil Moisture Contributions to Subseasonal Forecast Skill , 2011 .

[26]  Soledad Le Clainche Martínez,et al.  Spatio-Temporal Koopman Decomposition , 2018, J. Nonlinear Sci..

[27]  Sang-Wook Yeh,et al.  ENSO Atmospheric Teleconnections and Their Response to Greenhouse Gas Forcing , 2018 .

[28]  C. Anagnostopoulou,et al.  Representing teleconnection patterns over Europe: A comparison of SOM and PCA methods , 2015 .

[29]  J. Horel Complex Principal Component Analysis: Theory and Examples , 1984 .

[30]  Tao Lian,et al.  An Evaluation of Rotated EOF Analysis and Its Application to Tropical Pacific SST Variability , 2012 .

[31]  K. Nordhausen,et al.  Fourth Moments and Independent Component Analysis , 2014, 1406.4765.

[32]  Elizabeth C. Kent,et al.  Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century , 2003 .

[33]  G. Compo,et al.  Removing ENSO-Related Variations from the Climate Record , 2010 .

[34]  K. Trenberth,et al.  Irregularity and decadal variation in ENSO: a simplified model based on Principal Oscillation Patterns , 2014, Climate Dynamics.

[35]  Danilo P. Mandic,et al.  Empirical Mode Decomposition-Based Time-Frequency Analysis of Multivariate Signals: The Power of Adaptive Data Analysis , 2013, IEEE Signal Processing Magazine.

[36]  Ivor W. Tsang,et al.  The pre-image problem in kernel methods , 2003, IEEE Transactions on Neural Networks.

[37]  Surajit Ray,et al.  Functional factor analysis for periodic remote sensing data , 2012, 1206.6962.

[38]  Bernhard Schölkopf,et al.  Learning to Find Pre-Images , 2003, NIPS.

[39]  Mirko van der Baan,et al.  Nonstationary phase estimation using regularized local kurtosis maximization , 2009 .

[40]  A.R. Messina,et al.  Complex Empirical Orthogonal Function analysis of wide-area system dynamics , 2008, 2008 IEEE Power and Energy Society General Meeting - Conversion and Delivery of Electrical Energy in the 21st Century.

[41]  A. Arneth,et al.  Global patterns of land-atmosphere fluxes of carbon dioxide, latent heat, and sensible heat derived from eddy covariance, satellite, and meteorological observations , 2011 .

[42]  Joaquim Ballabrera-Poy,et al.  Dominant Features of Global Surface Soil Moisture Variability Observed by the SMOS Satellite , 2019, Remote. Sens..

[43]  Gustau Camps-Valls,et al.  Nonlinear Complex PCA for Spatio-Temporal Analysis of Global Soil Moisture , 2018, IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium.

[44]  Sergios Theodoridis,et al.  Ieee Transactions on Signal Processing Extension of Wirtinger's Calculus to Reproducing Kernel Hilbert Spaces and the Complex Kernel Lms , 2022 .

[45]  D. Volkov Do the North Atlantic winds drive the nonseasonal variability of the Arctic Ocean sea level? , 2014 .

[46]  Francisco J. Prieto,et al.  Multivariate Outlier Detection and Robust Covariance Matrix Estimation , 2001, Technometrics.

[47]  Alexander Ilin,et al.  Variational Gaussian-process factor analysis for modeling spatio-temporal data , 2009, NIPS.

[48]  Guonian Lü,et al.  ENSO signals and their spatial-temporal variation characteristics recorded by the sea-level changes in the northwest Pacific margin during 1965–2005 , 2009 .

[49]  Asko Noormets,et al.  Phenology of Ecosystem Processes , 2009 .

[50]  Feature extraction of climate variability, seasonality, and long‐term change signals in persistent organic pollutants over the Arctic and the Great Lakes , 2017 .

[51]  Bin Wang,et al.  How to Measure the Strength of the East Asian Summer Monsoon , 2008 .

[52]  José Luis Rojo-Álvarez,et al.  Digital Signal Processing with Kernel Methods , 2018 .

[53]  Bingni W. Brunton,et al.  Data-driven spatiotemporal modal decomposition for time frequency analysis , 2018, Applied and Computational Harmonic Analysis.

[54]  Christian Berger,et al.  An Image Transform Based on Temporal Decomposition , 2018, IEEE Geoscience and Remote Sensing Letters.

[55]  Jon Atli Benediktsson,et al.  Introduction to the Issue on Advances in Remote Sensing Image Processing , 2011, IEEE J. Sel. Top. Signal Process..

[56]  Stefan Rahmstorf,et al.  Global temperature evolution 1979–2010 , 2011 .

[57]  John M. Wallace,et al.  A large discontinuity in the mid-twentieth century in observed global-mean surface temperature , 2008, Nature.

[58]  Andrey Gavrilov,et al.  Principal nonlinear dynamical modes of climate variability , 2015, Scientific Reports.

[59]  N. Verhoest,et al.  El Niño-La Niña cycle and recent trends in continental evaporation , 2014 .

[60]  Kaare Brandt Petersen,et al.  Kernel Multivariate Analysis Framework for Supervised Subspace Learning: A Tutorial on Linear and Kernel Multivariate Methods , 2013, IEEE Signal Processing Magazine.

[61]  Jehn-Yih Juang,et al.  On the spectrum of soil moisture from hourly to interannual scales , 2007 .

[62]  S. Sparnocchia,et al.  Multivariate Empirical Orthogonal Function analysis of the upper thermocline structure of the Mediterranean Sea from observations and model simulations , 2003 .

[63]  Markus Reichstein,et al.  Predicting carbon dioxide and energy fluxes across global FLUXNET sites with regression algorithms , 2016 .

[64]  Damien Sulla-Menashe,et al.  MODIS Collection 5 global land cover: Algorithm refinements and characterization of new datasets , 2010 .

[65]  Lorenzo Bruzzone,et al.  Kernel methods for remote sensing data analysis , 2009 .

[66]  L. T. DeCarlo On the meaning and use of kurtosis. , 1997 .

[67]  Nadine Gobron,et al.  Identifying multiple spatiotemporal patterns: A refined view on terrestrial photosynthetic activity , 2010, Pattern Recognit. Lett..

[68]  Andrew J. Majda,et al.  Extraction and predictability of coherent intraseasonal signals in infrared brightness temperature data , 2016, Climate Dynamics.

[69]  Atul K. Jain,et al.  Compensatory water effects link yearly global land CO2 sink changes to temperature , 2017, Nature.

[70]  H. Dijkstra Nonlinear Climate Dynamics , 2013 .

[71]  Joseph L. Awange,et al.  Quantifying the impacts of ENSO and IOD on rain gauge and remotely sensed precipitation products over Australia , 2016 .

[72]  Wouter Dorigo,et al.  How Oceanic Oscillation Drives Soil Moisture Variations over Mainland Australia: An Analysis of 32 Years of Satellite Observations* , 2013 .

[73]  Henry F. Diaz,et al.  Potential Errors in the Application of Principal Component (Eigenvector) Analysis to Geophysical Data , 1982 .

[74]  Kazuhito Ichii,et al.  Ranking drivers of global carbon and energy fluxes over land , 2015, 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS).

[75]  T. Wigley,et al.  Global patterns of ENSO‐induced precipitation , 2000 .

[76]  Ranz,et al.  World Map of the Köppen-Geiger climate classification updated — Source link , 2006 .

[77]  Radford M. Neal Pattern Recognition and Machine Learning , 2007, Technometrics.

[78]  Bradfield Lyon,et al.  ENSO and the Spatial Extent of Interannual Precipitation Extremes in Tropical Land Areas , 2005 .

[79]  S. Seneviratne,et al.  Investigating soil moisture-climate interactions in a changing climate: A review , 2010 .

[80]  John M. Wallace,et al.  Pairwise-Rotated EOFs of Global SST , 2017 .