INRIA logo Ecole des Ponts ParisTech logo CEA logo CNRS logo UJF logo UJF logo Cemagref logo LSCE logo LJK logo CEREA logo INRIA Rocquencourt logo INRIA Rhône-Alpes logo INRIA Rennes logo

Constraining surface emissions of air pollutants using inverse modeling: method intercomparison and a new two-step two-scale regularization approach

Task 3c, principal investigator: CEREA/CLIME.

When constraining surface emissions of air pollutants using inverse modeling one often encounters spurious corrections to the inventory at places where emissions and observations are co-located, referred to as the co-localization problem. Several approaches have been used to deal with this problem: coarsening the spatial resolution of emissions; adding spatial correlations to the covariance matrices; adding constraints on the spatial derivatives into the functional being minimized; and multiplying the emission error covariance matrix by weighting factors. Intercomparison of methods for a carbon monoxide inversion over a city shows that even though all methods diminish the co-localization problem and produce similar general patterns, detailed information can greatly change according to the method used ranging from smooth, isotropic and short range modifications to not so smooth, non-isotropic and long range modifications. Poisson (non-Gaussian) and Gaussian assumptions both show these patterns, but for the Poisson case the emissions are naturally restricted to be positive and changes are given by means of multiplicative correction factors, producing results closer to the true nature of emission errors. Finally, we have proposed and tested a new two-step, two-scale, fully Bayesian approach that deals with the co-localization problem and can be implemented for any prior density distribution.

Time average difference between the a priori (official) carbon monoxide inventory and the optimized inventories after assimilating real observations using Gaussian estimates: grid-cell values and smoothed contours. (a) Using no modifications. (b) Using correlations with a radius of influence r=2.5 cells. (c) Adding a constraint on the derivative in the functional. (d) Using weighting factors over B . (e) Performing a two step multiscale inversion on a coarse grid of 3X3 for one of the possible grid positions and (f) the same as (e) after averaging the nine possible placements of the 3X3 grid. Note that positive values represent a decrease over the background emissions and negative values represent an increase over the background emissions. See the text for details. Units are in micro grams per square meter per second.