Ninth International Geostatistics Congress, Oslo, Norway
June 11 – 15, 2012


Plenary 7

Abstract No.:



Low Parameter Representations of Anthropogenic Greenhouse Gas Flux Fields for Inverse Parameter Estimation


S.A. McKenna, Sandia National Laboratories (US)
B. van Bloemen Waanders, Sandia National Laboratories (US)
J. Ray, Sandia National Laboratories (US)


Characterization of anthropogenic fluxes of greenhouse gases (GHG?s) from the earth?s surface is necessary for understanding the role these gases play in climate change. In contrast to biogenic fluxes that vary relatively smoothly in space, anthropogenic fluxes are highly localized and define a ?spiky? surface with large regions at, or near, zero flux. Such fields are not well modeled by smooth interpolation algorithms (e.g., covariance-based interpolators). Here, we examine several approaches to representing anthropogenic flux fields with a focus on non-smooth interpolators. In particular, harmonic and bi-harmonic interpolators can be defined as solutions of second-order partial differential equations (PDE?s). Resulting fields represent sample locations as discontinuities at local maxima/minima. The goal of this work is to examine low parameter representations of anthropogenic flux fields. Our motivation for limiting the parameterization is to facilitate efficient inverse modeling of the GHG flux fields from measurements of GHG concentrations in the atmosphere. PDE-based interpolators are compared to other approximations of the flux field including Gaussian kernels and wavelets as well as kriging. A highly resolved model of GHG flux across the continental United States (VULCAN) serves as the ground truth and is used as the basis for this comparison. An advantage of the PDE-based interpolators is the recent development of ?image guided interpolation? (Hale, 2009) where local information on the orientation and aspect ratio of features from a spatially exhaustive image of a covariate can be extracted and used to guide interpolation. Here, population density, per capita energy use and nocturnal illumination (?lights at night?) serve as potential covariates for estimation of anthropogenic GHG fluxes.

Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000




    Copyright 2012 International Geostatistics Congress

Produced by X-CD Technologies Inc.