A functional-data approach to the Argo data

The Argo data is a modern oceanography dataset that provides unprecedented global coverage of temperature and salinity measurements in the upper 2,000 meters of depth of the ocean. We study the Argo data from the perspective of functional data analysis (FDA). We develop spatio-temporal functional kriging methodology for mean and covariance estimation to predict temperature and salinity at a fixed location as a smooth function of depth. By combining tools from FDA and spatial statistics, including smoothing splines, local regression, and multivariate spatial modeling and prediction, our approach provides advantages over current methodology that consider pointwise estimation at fixed depths. Our approach naturally leverages the irregularly-sampled data in space, time, and depth to fit a space-time functional model for temperature and salinity. This model is then used to provide predictions continuously as a function of pressure. The resulting methodology provides new tools for scientific problems where the dependence on depth must be considered. We conclude with two important examples on the estimation of the ocean heat content and the mixed layer depth.

[1]  Jane-Ling Wang,et al.  From sparse to dense functional data and beyond , 2016 .

[2]  Piotr Kokoszka,et al.  Some Recent Developments in Inference for Geostatistical Functional Data , 2019, Revista Colombiana de Estadística.

[3]  J. O. Ramsay,et al.  Functional Data Analysis (Springer Series in Statistics) , 1997 .

[4]  Dean Roemmich,et al.  Unabated planetary warming and its ocean structure since 2006 , 2015 .

[5]  Mikael Kuusela,et al.  Locally stationary spatio-temporal interpolation of Argo profiling float data , 2017, Proceedings of the Royal Society A.

[7]  Pascal Monestiez,et al.  A Cokriging Method for Spatial Functional Data with Applications in Oceanology , 2008 .

[8]  D. Roemmich,et al.  135 years of global ocean warming between the Challenger expedition and the Argo Programme , 2012 .

[9]  Dean Roemmich,et al.  The 2004-2008 mean and annual cycle of temperature, salinity, and steric height in the global ocean from the Argo Program , 2009 .

[10]  G. Johnson,et al.  As El Niño builds, Pacific Warm Pool expands, ocean gains more heat , 2017 .

[11]  T. Hsing,et al.  Uniform convergence rates for nonparametric regression and principal component analysis in functional/longitudinal data , 2010, 1211.2137.

[12]  Roger Woodard,et al.  Interpolation of Spatial Data: Some Theory for Kriging , 1999, Technometrics.

[13]  H. Rue,et al.  An explicit link between Gaussian fields and Gaussian Markov random fields: the stochastic partial differential equation approach , 2011 .

[14]  S. Levitus,et al.  World ocean heat content and thermosteric sea level change (0–2000 m), 1955–2010 , 2012 .

[15]  A. Gelfand,et al.  Bayesian Nonparametric Functional Data Analysis Through Density Estimation. , 2009, Biometrika.

[16]  B. Silverman,et al.  Nonparametric regression and generalized linear models , 1994 .

[17]  Ashton Wiens,et al.  Modeling spatial data using local likelihood estimation and a Matérn to spatial autoregressive translation , 2020, Environmetrics.

[18]  T. Hsing,et al.  Theoretical foundations of functional data analysis, with an introduction to linear operators , 2015 .

[19]  M. Reimherr,et al.  A geometric approach to confidence regions and bands for functional parameters , 2016, 1607.07771.

[20]  Jianqing Fan.  Local Polynomial Modelling and Its Applications: Monographs on Statistics and Applied Probability 66 , 1996 .

[21]  David Bolin,et al.  Multivariate type G Matérn stochastic partial differential equation random fields , 2019 .

[22]  Hans-Georg Müller,et al.  Functional Data Analysis , 2016 .

[23]  Nancy Wilkins-Diehr,et al.  XSEDE: Accelerating Scientific Discovery , 2014, Computing in Science & Engineering.

[24]  H. Barnes,et al.  Descriptive Physical Oceanography , 1977 .

[25]  P. Hall,et al.  Properties of principal component methods for functional and longitudinal data analysis , 2006, math/0608022.

[26]  T. McDougall.  Potential Enthalpy: A Conservative Oceanic Variable for Evaluating Heat Content and Heat Fluxes , 2003 .

[27]  G. Johnson,et al.  Estimating Global Ocean Heat Content Changes in the Upper 1800 m since 1950 and the Influence of Climatology Choice , 2014 .

[28]  H. Niessner,et al.  On computing the inverse of a sparse matrix , 1983 .

[29]  Joseph Guinness,et al.  Gaussian process learning via Fisher scoring of Vecchia’s approximation , 2019, Statistics and Computing.

[30]  Hadley Wickham,et al.  ggplot2 - Elegant Graphics for Data Analysis (2nd Edition) , 2017 .

[31]  W. Brechner Owens,et al.  An improved calibration method for the drift of the conductivity sensor on autonomous CTD profiling floats by θ–S climatology , 2009 .

[32]  Kehui Chen,et al.  Weak Separablility for Two-way Functional Data: Concept and Test , 2017 .

[33]  Jorge Mateu,et al.  Statistics for spatial functional data: some recent contributions , 2009 .

[34]  Jorge Nocedal,et al.  A Limited Memory Algorithm for Bound Constrained Optimization , 1995, SIAM J. Sci. Comput..

[35]  F. Roquet,et al.  The Thermohaline Modes of the Global Ocean , 2019, Journal of Physical Oceanography.

[36]  S. Wood.  Low‐Rank Scale‐Invariant Tensor Product Smooths for Generalized Additive Mixed Models , 2006, Biometrics.

[37]  Dean Roemmich,et al.  An Argo mixed layer climatology and database , 2017 .

[38]  Jianqing Fan,et al.  Local polynomial modelling and its applications , 1994 .

[39]  Jerry M. Davis,et al.  A Functional Data Analysis of Spatiotemporal Trends and Variation in Fine Particulate Matter. , 2018, Atmospheric environment.

[40]  Peter Craven,et al.  Smoothing noisy data with spline functions , 1978 .

[41]  Lynne D. Talley,et al.  A New Algorithm for Finding Mixed Layer Depths with Applications to Argo Data and Subantarctic Mode Water Formation , 2009 .

[42]  Yehua Li,et al.  Unified Principal Component Analysis for Sparse and Dense Functional Data under Spatial Dependency , 2020 .