The authors have addressed many of the key points from referee #1 and #2, and have greatly improved this paper. One critique is that the paper could be broken into two or three papers, but with additional text in the Introduction providing a road map and a rationale for the reader that explains why the authors chose to do both a quality assessment of chronologies in the pollen database and to use the database to characterize the last termination (e.g. as a proof of concept).The new title is more appropriate for the content.

The authors have addressed many of the key points from referee #1 and #2, and have greatly improved this paper. One critique is that the paper could be broken into two or three papers, but with additional text in the Introduction providing a road map and a rationale for the reader that explains why the authors chose to do both a quality assessment of chronologies in the pollen database and to use the database to characterize the last termination (e.g. as a proof of concept).The new title is more appropriate for the content.

The manuscript by Flantau et al. has been greatly improved by the reviewer’s suggestions and now includes a critically important addition within the supplementary information that specifies the original authors and manuscripts that form the bulk of this work. Unfortunately, the supplementary information was either unavailable or possibly overlooked during the review process. I suggest one final improvement to this manuscript, which is to move the tabular information for the original data sources from supplementary material to a primary table within the main manuscript. In addition to providing this information "front and center", the authors have informed the editor that all re-calibrated age models, a major effort and contribution from this paper, are now available through a data repository with detailed instructions for the various R scripts used to produce age models. There is also detailed information regarding how to implement the age uncertainty, which provides an additional significant contribution to the user community. The authors have also provided detailed information and the rationale regarding how and why age models from individual records were adjusted. This additional information will provide the original authors and those that created the data greater detail on how the originally published age models were changed to ease in future reproduction and interrogation of these records. The efforts put forth by Flantua et al. are commendable and will become a valuable contribution for future use and dissemination of these palynological records.