Robust extraction of baseline signal of atmospheric trace species using local regression

Abstract

The identification of atmospheric trace speciesmeasurements that are representative of well-mixed back-ground air masses is required for monitoring atmosphericcomposition change at background sites. We present a sta-tistical method based on robust local regression that is wellsuited for the selection of background measurements and theestimation of associated baseline curves. The bootstrap tech-nique is applied to calculate the uncertainty in the result-ing baseline curve. The non-parametric nature of the pro-posed approach makes it a very flexible data filtering method.Application to carbon monoxide (CO) measured from 1996to 2009 at the high-alpine site Jungfraujoch (Switzerland,3580 m a.s.l.), and to measurements of 1,1-difluoroethane(HFC-152a) from Jungfraujoch (2000 to 2009) and MaceHead (Ireland, 1995 to 2009) demonstrates the feasibility andusefulness of the proposed approach.The determined average annual change of CO at Jungfrau-joch for the 1996 to 2009 period as estimated from filteredannual mean CO concentrations is−2.2±1.1 ppb yr−1. Forcomparison, the linear trend of unfiltered CO measurementsat Jungfraujoch for this time period is−2.9±1.3 ppb yr−1.