Issues Regarding the Assimilation of Cloud and Precipitation Data

This is the authors' response to a set of criticisms regarding a previously published work. It briefly addresses the main criticisms. In particular, it explains why some papers identified as having some fundamental flaws were referenced in the original work without detailed exposition of those flaws. It also explains why parts of the conclusion criticized as being contradictory are, in fact, not. It further highlights the need for more publishing of scientific criticisms. In the December 2007, special issue of the Journal of Atmospheric Sciences devoted to the Workshop on Assimilation of Satellite Cloud and Precipitation Observations, the authors published an article summarizing the many critical issues that render observations of cloud and precipitation difficult to analyze. Essentially, these include the inaccuracies of both current instruments and the relationships between what is actually observed (infrared or microwave energy detected at the altitude of the satellite) to what is desired (e.g., estimates of cloud drop sizes or rain rates) and the chaotic nature of atmospheric behavior and the complex mathematics describing it. The paper also included recommendations for future research and brief descriptions of many previous works concerning the subject. One reader is now attempting to publish a criticism of that paper. Her three complaints are that there was insufficient explanation of the identification of some cited works as being fundamentally flawed, that as a review the paper should have referenced some works additional to those it did, and that two recommendations were contradictory. Each of these complaints is addressed briefly in this response. First we explain why a brief list of works cited in our paper were identified as "flawed" with only a brief explanation. The design and conduct of the experiments reported in those papers violate well-established fundamentals such that, once the errors are recognized, their interpretations are no longer supported. Unfortunately, over the years, no researchers have bothered to publish criticisms of those papers, such that there are now too many to address in any single paper not devoted to that purpose. Yet, those papers are so often cited that we could not simply ignore them. Furthermore, if we had cited them without warning our readers regarding their flaws, we would have perpetrated a great disservice. In our response, however, we do offer further explanation of why some details, neglected in these papers, are critical to proper scientific evaluation. Neither did we offer insufficient references. Although we intentionally did not claim to be a "review' paper, we did cite 100 papers. That number is approximately 5 times the usual amount cited in journal articles. Although we only referenced few papers published after 2005, that was because our manuscript was submitted in January 2006, with its final, editorially-reviewed form in June 2006. We therefore could not reference papers published after this date. The problem here is that our paper was "in press" for 18 months. Finally, we explain that a careful reading of our paper reveals that our recommendations are not contradictory. Essentially, although we recommend 2 very distinct research approaches, these are complimentary and either alone is insufficient to accelerate progress. In conclusion, we recommend that the scientific community expends greater effort in publishing careful scientific criticisms so that others do not face the same dilemma we did. Likely this requires some reward system for doing so.