Emergence of a tracer source from air concentration measurements, a new strategy for linear assimilation

[摘要] The measurement of atmospheric concentrations by a monitoring network is apromising tool for the identification of the widespread sources of tracespecies. The paper addresses the case of the species scattered linearly by aknown meteorology. The question is classical: what can be said about thesource from a set of measurements? Is it possible to guess from the valuesobserved by the measurements that the source is spread close to thedetectors, or that the tracer comes from a remote region? And, if the sourcewas a point source, would it be possible to understand it by just consideringthese values? A part of the answers is a matter of practical sense: theresolution with which an emission can be retrieved will always be limited andprobably lower for a remote region, even if the detectors and dispersionmodel are error free. The paper proposes a linear strategy of inference: toany set of values taken by the observed concentrations is associated linearlyan estimate of the source. Doubled values lead to a doubled estimate. Themethod, based on adjoint techniques, is intended to optimise the resolutionby quantifying, with the concept of illumination, which regions are well,poorly or not seen at all. The illumination tied to ordinary adjointfunctions becomes excessive close to the detectors thus leading to inversionartefacts. This may be corrected by attributing each point of the space timedomain a geometric and statistical weight. The adjoint functions aretransformed. The choice of this renormalising function is constrained by anunambiguous entropic criterion preventing any overestimation of the availableinformation that would lead to artefacts. It amounts to evenly distribute theinformation between the points organised with their weights as a "knowndomain". The theory is illustrated by calculations performed with theexperimental source ETEX1.

[学科分类] 大气科学