Comment on "Global OH trend inferred from methylchloroform measurements" by Maarten Krol et al. - Reply

In a comment on the paper by Krol et al. [ 1998], Prinn and Huang [this iss ue] compare the estimates of the linear OH trends from the Atmospheric Life time Experiment/Global Atmospheric Gases Experiment/Advanced Global Atmosph eric Gases Experiment (ALE/GAGE/AGAGE) methylchloroform data using a recurs ive weighted least squares (Kalman) filter described by Prinn et al. [1995] with the OH trend obtained using the ensemble method by Krol et al. [1998] . We refer to these two papers as K98 and P95 hereafter. The comment interp rets the difference between the estimated trends (0.0 +/- 0.2% per year est imated by P95 and 0.46 +/- 0.6% per year estimated by K98) by a combination of different emissions and different treatments of the observations. More specifically, the comment claims that the polynomial fitting by K98 leads t o a arbitrary deweighting of (1) the CH3CCl3 data at the beginning and part icularly at the end of the time series, and (2) the CH3CCl3 data in the Sou thern Hemisphere relative to the Northern Hemisphere. However, the major di fferences between the two trend estimates are attributed to the different e mission scenarios. The comment is organized in three parts, the second and the third parts being additional comments on our first reply. For clarity r easons, we organize our reply in the same way as the comment.