How to Facilitate Coordination of Modeling with Field Campaigns Within the ACP Framework: note from a breakout session at the February 23-25, 1999 Annual ACP Science Meeting

On Coordination of Modeling with Field Campaigns within the ACP Framework
Breakout Session Summary DOE Atmospheric Chemistry Program Annual Meeting, February 1999
Reported by Carl Berkowitz Pacific Northwest National Laboratory

A plenary session was held during the 1999 Annual Science Team to consider what could be done to enhance participation of the ACP modeling community in the planning, execution, and analysis of results from, the ACP field campaigns. This session was motivated by the observation that there has been little input from modelers in any of these activities.

Two categories of suggestions were put forward. The first category consisted of specific ways to enhance communications between modelers and field scientists. It was suggested that communications could be enhanced by a) holding special workshops to focus on specific scientific processes and/or b) the development and maintenance of three "web sites" describing laboratory, field and modeling work within ACP. The later activity would be assigned to the ACP Science Director, who would also be charged with keeping these sites up to date and in notifying ACP participants of changes to these sites as such changes are implemented. The second set of suggestions were of an administrative nature, and are summarized in the final section of these notes.

It should be noted that many of the issues discussed at the 1999 meeting had been brought up at the 1998 ACP meeting. Although field scientists have posed the driving questions in the past, there is no reason why any scientific questions would be barred from consideration in planning. Another issue brought up in 1998, and again in 1999, was how to make data sets more useful. To this end, it was suggested at both meetings that attention be given to ease of accessibility. Many of the suggestions made during the 1999 plenary session addressed this specific issue. A specific request made by modelers during the 1998 meeting was for repetitive field measurements (e.g., daily measurements made over the course of a month) at sites that to be revisited during multiple years.

Perhaps the best summary of how these efforts could be coordinated to come out of the 1998 meeting was for field scientists to "invite a modeler to lunch." We would only modify it in 1999 to suggest that modelers should also feel free to "invite a field scientist to lunch."

Laboratory scientists should also join us for lunch. Although the charge of this year's breakout session was to facilitate interaction between modelers and field scientists, it was also thought appropriate to bring laboratory scientists into the discussion. In at least one institution, such coordination is already underway, and proving very fruitful, with updated accommodation coefficients being evaluated in an Eulerian model that is being used to evaluate measurements taken during the NARSTO '95 field campaign. We think this organizational model should be extended to other participants within ACP.

Two suggestions received most of the attention for how communication could be enhanced. The first suggestion was for ACP to have special working meetings of modelers, laboratory investigators, and field scientists. Each meeting would focus on a specific process of interest to these three groups of scientists, e.g., a special workshop to consider our understanding of accommodation coefficients of nitrogen species on aerosols. Results from the workshop would serve to focus research among and between participants. Follow-up workshops some would allow for the synthesis of results from the field, laboratory, and modeling studies. It was noted that having a special workshop separate from the more general ACP meeting would have fewer distractions and provide more focus.

The plenary group almost unanimously requested that three "web sites" be established to serve as a media to facilitate the exchange of information between modelers, field scientists and laboratory investigators, each group having a site. All web sites would have three common features:

The modeling web page would have a wish list in which laboratory people could request a particular sensitivity study involving some laboratory related value (e.g., an uptake rate).
The field page could have a wish list in which modelers could ask that a particular measurement be made during the next field campaign.
And the laboratory page could have a wish list in which modelers could ask for guidance regarding such things as sensitivity of uptake rates to various trace gas concentrations

Finally, each web site should contain relevant planning documents for future field, laboratory or modeling activities in which other ACP scientists could participate. Specific items suggested for inclusion on these web sites were as follows:

Field Campaign Website:

1. Upcoming/potential field campaigns

Who to contact
Posting of planning documents that state the objective(s) of field campaign

Flight plans
Instrumentation expected to be onboard the G-1
Expected dates

2. Summary of past field campaigns

Existing aircraft observations, or who to obtain such observations from
Available surface observations collected during the campaign
Available meteorology and emissions data
Mission summaries (instrument status, ancillary observations, etc.)

3. Wish list: Suggestions for field campaign people from:

Modelers
Laboratory scientists

4. E-mail notification of website updates, sent to scientists in ACP.

Modeler's Website

1. Ongoing or planned model analysis relevant to ACP

Who to contact
Posting, or linking to, of abstracts or "gray literature" progress reports (e.g., AMS extended abstracts) so others can familiarize themselves with modeling work.

2. A catalog of ACP models:

Short descriptions posted on ACP web page

An overview of the model
On-going applications
Past analysis available for use by others.

Links to more detailed descriptions.

3. Linkages of small-scale (e.g. STEM) models to field studies (e.g. NARSTO) for evaluation of process descriptions (e.g., vertical exchange processes) and subsequent inclusion in large-scale models (MOZART). And the use of large-scale models to provide initial and boundary conditions to small scale models.

It is expected that testing these descriptions (e.g., cloud venting, NOx/NOy ratios, etc.) would be done within the smaller scale models. The grid resolution in these codes is more amenable to comparison with typical ACP field program results. Following testing in these codes, they could be incorporated into the larger-scale models.
Availability of large-scale models to provide boundary conditions, initial conditions for the smaller-scale regional-models.

4. Wish list: Suggestions for modelers from:

Field scientists:
Laboratory scientists

5. E-mail notification of website updates, sent to scientists in ACP.

A similar site for laboratory studies was also proposed, although no details were suggested. It was strongly requested that ACP management serve as the focal point for implementing these web sites, keeping them current, and in notifying interested scientists of updates and changes.

Three administrative procedures were discussed to increase participation, but none received support from the plenary participants. We note them here for the sake of completeness.

The first suggestion was to require that ACP participants be formally asked to change their research efforts to focus on the analysis of results from field campaigns. Such a request from DOE would result in potentially wasting 2 years of work on programs with a lifetime of only 3 years. Further, investigators may not have either the tools or desire to follow-up on such a change. The second suggestion involved having field campaign staff design their studies as "blind tests" of model performance. However, this would likely be done at the cost of programs designed to elucidate more pressing scientific issues, which would have a much different design than those of such a test. Finally, it was also suggested that the next round of ACP proposals require that all modeling programs be linked to some field effort. Plenary participants noted in response to this suggestion that some, if not most, focused efforts, would not naturally fit into such a model, and while this should be encouraged, it could greatly harm the program if it were laid down as a firm rule for future participation in ACP.