Timely publication is something authors and readers should expect from AGU journals. Currently, the journals are falling short in this area because of the challenges of the transition from the old, author-produced copy to full electronic production and delivery. 

We believe that the entire membership should be informed about this problem and about what is being done to return to timely publication.

Current Situation

The inventory of accepted but not-yet-published articles is too high. The current timeliness targets from acceptance to publication are 9 weeks for GRL, and 16 weeks for all other journals. To be publishing at these rates, the inventory of not-yet-published articles should not exceed 189 for GRL, and 1008 for the others. At "steady state," GRL would be publishing 21 articles per week, and the other journals, 63 per week.

The inventory for GRL as of 4 May was 428 unpublished articles, 239 above our goal. For the other journals, the combined inventory was 1776 articles, 768 above our goal. In the last month, the GRL inventory decreased by 153, and the combined backlog for all other journals grew by 6.

How Did We Get into this Situation? 

Significant concentrations of energy were placed on getting the author-produced, paper-based articles into the 2001 issues. This emphasis took away time that, in hindsight, should have been spent in helping staff members learn the new processes for electronic production.

Staff underestimated the resources that would be needed in the early stages of this transition.

Processing electronic text and figure files took more time than originally expected. Especially with the early articles, there were many questions and problems that took iterations with authors and the vendor to resolve. Because AGU jumped ahead of most publishers in the way we are processing material, our staff have not had the opportunity to learn from colleagues, and have had to do a fair amount of experimentation and rethinking of workflow and process.

In addition, management information systems were not tuned to the new ways of doing work, and therefore, did not show the pressure points clearly. Consequently, the problem was not detected when it should have been.

What Is Being Done to Reduce the Backlog and Publication Time? 

Additional resources have been engaged to augment the work of staff. Bottlenecks in the process are being identified and eliminated. As an emergency measure, staff and the temporary workers have been organized into teams that concentrate on specific functions rather than on a particular journal.

Priority 1 was to use these techniques to significantly increase the number of published GRL articles. The accompanying graph shows the success that has been achieved. In 4 weeks, 193 GRL contributions were published. At this point, we project being on schedule for GRL by the end of June.

For the other journals, staff have been working during the past month to get to a steady state on a daily basis for all of the major steps of the production process. Staff are almost at that point, and the output of published articles will very soon start to show the results of this effort.

The Publications Committee, the Editors, and the staff are disappointed that the transition has led to these delays. The staff are committed to reducing the backlog, and progress will be reported in Eos and on the AGU Web site. In the meantime, members can find out about progress by going to http://www.agu.org/pubs/stats/, where the inventory is reported, along with the age of the oldest 90% and 50% of unpublished articles.

Authors

George Hornberger, Publications Committee Chair

Judy C. Holoviak, Director of Publications

© 2002 AGU