Setting a Course for Antarctic Integrated and Systems Science (AISS): Lessons from ARCSS A personal perspective Joshua Schimel, University of California Santa Barbara, ARCSS Committee Chair On behalf of the ARCSS Committee and ARCSS Science Management Office This is a presentation that I put together with a lot of input and information, but my experience with ARCSS has been limited (part of LAII initiatives but only on the AC for a year) and so it really is my perspective.

Ph.D. in Soil Microbiology My Background: B.A. in Chemistry Ph.D. in Soil Microbiology My research focuses on the links between soil microorganisms and ecosystem function. Background- I’m not what you’d normally consider a “System Scientist”. The bulk of my work is smaller-scale mechanistic but links to vegetation and then on to climate. In many ways I’m more of an ANS-type person, but believe in the ARCSS mission.

ARCSS History Accomplishments Limitations Management

ARCSS Program History Milestones 1989 - Began as one of 22 Global Change Research programs at NSF 1996 All-Hands Workshop - “Toward An Arctic System Synthesis: Results and Recommendations” 2002 All-Hands Workshop - Transition from “domain” groupings to integrated program 2003/2004 - Program reorganized to advance system-science, synthesis approaches

Even in the early documents a focus on global change (title) and linkages among components (artwork) were central.

A system view of the Arctic ARCSS focuses on the connections, highlighting the importance of interactions between ARCSS, ANS, and other more ‘disciplinary’ programs: you can’t do good linkage science if you don’t have solid work being done on the components you are linking. From Overpeck et al. 2005. Arctic System on Trajectory to New, Seasonally, Ice-Free State. EOS. 86: 309, 312-313.

Arctic System Complexity The vision of the Arctic System has become increasingly complex and interconnected. We now see all these elements as being important Components of the Arctic System and as interacting with each other to produce the full suite of change that we are observing.

ARCSS Program ARCSS focuses on understanding the connections and feedbacks between components of the Arctic System. THC Sea Ice Reiteration: traditional programs focus on the discrete elements. ARCSS focuses on the linkage. Neither ignores the other, however, ARCSS does some domain-specific science, ANS does some linkage science.

ARCSS Research Past Components GISP2 - Greenland Ice Sheet Project Two (GISP2) PALE - Paleoclimates from Arctic Lakes and Estuaries OAII - Ocean-Atmosphere-Ice Interactions NWP, Arctic Ocean Section, SHEBA, SBI LAII - Land-Atmosphere-Ice Interactions Flux, ATLAS, ITEX PARCS - Paleoenvironmental Arctic Sciences (incorporated GISP2 and PALE) RAISE/LSI - Russian-American Initiative on Shelf-Land Environments in the Arctic Phase 1: Incorporation of existing programs into ARCSS- GISP/PALE Phase 2: Creating integrated programs OAII/LAII, moving into PARCS and RAISE. These are still programs that are focused, concrete, and work by building units up toward the top, rather than starting with an integrated vision of “The Arctic System” and working down.

Current ARCSS Research ARCSS Program currently funds over 160 projects FWI - Freshwater Integration Study Group 1: year 5; Group 2: year 4 SBI - Western Arctic Shelf-Basin Interactions 10 years of field data and analysis Entering Phase III - Synthesis SNACS - Study of the Northern Alaska Coastal System 6 projects, In 3rd year SASS - Synthesis of Arctic System Science 9 “SASS I” projects (funded 2005) 8 “SASS II” projects (funded 2006) HARC - Human Dimensions of the Arctic System Since 1997 New generation of projects are more conceptual in their structure, more grounded in a whole-Arctic-System perspective. As we’ve become more sophisticated in our understanding of the arctic system, we’ve been able to become more sophisticated in the programs we develop.

ARCSS Research A few Results & Highlights GISP2 341 papers 34 in Science and Nature SHEBA 145 papers LAII (partial, through 2004) 115 papers total 4 in Science and Nature 5 in BioScience 25 synthesis papers 2 TV Documentaries Pick 3 initiatives to highlight the amount and high-profile outlets of the publication coming from ARCSS. Every program produces tons of great science by our normal measures. Also- though, many of these papers and products would not likely have happened in more traditional programs.

First Case Study: Freshwater Integration study (FWI) A thematic program Pick two specific case studies to illustrate how ARCSS works when its working well to produce science that is more than the sum of its parts. The first case study is an integrated research program- FWI.

Goals Arctic-CHAMP/ASOF/SEARCH Freshwater Initiative (FWI) Are Fundamentally Synthetic Q1: Is the Arctic FW Cycle Intensifying? • Quantify Stocks and Fluxes • Document Changes to the Arctic Hydrologic Cycle Q2: If So, Why? • Understand the Source of the Change: Attribution Q3: What Are the Implications • Develop Predictive Simulations of Feedbacks to the Earth and Human Systems Arctic-CHAMP= Community-wide Hydrologic Analysis and Monitoring Program ASOF = Int’l Arctic-Sub-Arctic Ocean Flux study The core questions: These may seem clear, straightforward, and obvious now, but they weren’t at the very beginning. Everyone had their own questions and it took a lot of time and work to filter down to a small set that everyone could contribute to and that would work to pull researchers together.

Schematic of the FW system and of where different components of the FWI fit in. Different groups of people with different tools and interests all linking together to work on a program that targeted a major component of the Arctic System as a whole.

Example of FWI Synthesis: Budgeteers Group The Arctic Freshwater Budget • Well-focused target • System-wide view • Many perspectives • Directions for future work Serreze et al., 2006, JGR-Oceans One activity: the budget. This provided a single concrete challenge that every group could contribute to.

Evidence for an accelerating FW cycle Multi-model mean changes in Arctic Ocean FW Budgets 1950-2050 Increasing net precipitation over land and ocean Increasing ice melt, resulting in reduced ice transport Increasing liquid FW transport to the Atlantic ocean Small increase in Bering Strait FW inflow Sign convention is such that a positive number is an increasing source (or decreasing sink) of freshwater for the Arctic. This single condensed figure represents the contributions of many workers but provides a suite of data sets all of which contribute to answering one of the core questions. No one of these data sets would be compelling by itself- the parallelism of all of them add up to something more and more compelling. Positive means net flux into Arctic Holland et al., 2007

CHANGES AND ATTRIBUTION Working Group Feedbacks & implications on major subsystems White et al. JGR, Biogeosciences (submitted) Francis et al., (in prep.) Document basic character of  Focus on interactions: this provides a synthesis of what we think we know, but also helps identify the gaps, and so, new directions for research that might be ARCSS or might be ANS or other programs.

FWI PROGRESS THROUGH 2007 >100 peer-reviewed publications >100 presentations at National and Int’l forums: ACIA, ARCSS Synthesis Retreats, AGU, EGU, ASLO, etc. >24 Graduate and Undergraduate students Outreach efforts: AGU Press Conference, CNN, NY Times documentary, Discovery Channel, Canadian Broadcasting Co., NPR’s ‘All Things Considered’ Overall: enormous success by any reasonable measure. High rate of production. High impact of production.

Second Case Study: Role of land-surface changes in Arctic summer warming Chapin et al. 2005 Science 310:657-659 An “accidental” success? Second case study: a specific synthesis activity that grew out of a larger initiative. However, this specific paper had not been planned at the outset.

Summer warming so far is associated with longer season Key findings: Summer warming so far is associated with longer season Increasing woody vegetation exacerbates warming Shrub expansion is driven by internal positive feedbacks, involving nutrient cycling A “climate surprise” identified before it has happened. Important paper, published in Science, with a very long list of authors. All of us contributed in different ways to the overall product.

Integrated conceptual model of linkages and feedbacks driving warming Key component of the overall synthesis was a feedback diagram outlining how all the components interact with each other, including terrestrial and marine, physical and biological, etc. The integrated system perspective allowed us to fit in many specific components.

Synthesis grew from network of ATLAS collaborations Paper resulted from a network of connections that developed as a result of ATLAS- this wasn’t the “target” result of ATLAS but because ATLAS was well focused and integrated a paper such as this one developed naturally. A “planned accident” This diagram is incomplete, particularly on the climate side. The central group was Terry and Amanda, plant ecology and climatology. They had had separate projects during the Flux study in LAII and through their interaction in that study, developed to the point where they could submit a single interdisciplinary proposal for ATLAS. Schimel, Welker, and others were working on soil biological processes that controlled nutrient supply to shrubs. That connected with the shrub distribution people to start understanding the belowground feedback loops that accelerate shrub expansion. That connected with large-scale vegetation and treeline dynamics projects to understand the pan-Arctic scope and time-line of vegetation change, and connected with large scale biogeochemical models and climate models to infer how vegetation change will drive climate in the future. It was having all the right people in the right place interested in developing synthesis that allowed it to happen. Being able to do that is partly a result of the learning process we had all gone through that allowed us to talk to each other meaningfully. It did not happen overnight. It also required leadership to pull the effort together.

ATLAS created opportunity for new synthesis So: this synthesis was NOT “accidental” at all. ATLAS targeted land surface - climate interactions. Created a framework to pull in “peripheral” projects. Developed a research community interested in collaborating on the larger synthesis. So, this paper was a natural result of a well thought out program that created the opportunity. Many ARCSS (and non-ARCSS) programs have probably had this kind of success. ATLAS created opportunity for new synthesis

Science that would not have grown from disciplinary programs Successes: Tons of important papers Integrated understanding Societally relevant science Science that would not have grown from disciplinary programs New scholars: undergrad, grad, postdoc Outreach

Challenges and Limitations of System Science Community building and support Requires active support & development Interaction with disciplinary research Depends on healthy disciplinary research and generates new questions Planning is hard Time & energy from busy PIs and Program officers The successes of ARCSS programs don’t happen easily and without investment in community and planning.

1. Challenge 1: Community NSF always works through a feedback loop with the community. At the lowest level, an investigator submits a proposal and NSF funds it, further driving questions and needs. At higher level, this works via programs. But, NSF must invest in maintaining the feedback loop.

Challenge 1: Community Need to support the links between communities Create a single, larger community Biology Oceanography Hydrology There is no “Arctic Community” as a single integrated entity. Rather, it is a “meta-community” that is made up of separate sub-communities that interact to varying degrees. Building the arctic community is like building a conservation reserve. There are core areas that are connected by corridors. To make the overall work, you must focus on the corridors to ensure adequate flow of people and ideas.

Community Planning Structure Arctic Researchers National Science Foundation - NSF ARCSS ARCSS Committee (AC) - takes lead on behalf of the research community in developing the ARCSS Program Science Management Office (SMO, currently at ARCUS) - work with NSF, AC, and community on priorities and strategies Project Offices (Arctic-CHAMP/FWI, SNACS synthesis coordinator, HARC Core Office) ARCSS does many things to tie together the elements- researchers and the foundation

Community Planning Activities Engage community to define priorities, initiatives, and implementation strategies Face-to-Face Workshops and Meetings Web Conferences and eTown Meetings Communities of Practice (Co-oPs) Communication Tools (email listserve, website, online surveys, etc.) These things all take effort, time, and engagement. Those activities are essential to building a community that can do the kind of sophisticated interdisciplinary work we demand of the community.

Challenge 2: Interdisciplinarity

Overly simple straw man: traditional disciplines advanced, each in their own silo.

Occasionally groups would amalgamate with their neighbors in “narrowly” interdisciplinary efforts, sometimes creating new disciplines. E.g. Biochemistry.

ARCSS changes that dynamic: focusing on linkages among disciplines to advance our understanding of the integrated system.

Arctic System Complexity That is particularly necessary when the system you are considering is this complex.

Challenge 2: Interdisciplinarity Program mutualism vs. competition Real at a programmatic level, less so at an investigator level Investigator issues: some don’t like interdisciplinary work: May feel program is a threat May feel it is an opportunity, but submit weak proposals May review good interdisciplinary proposals critically Program level: money in one program may not go to another, but: 1. The better programs cooperate, the more likely they are to generate more funding for the whole area. 2. PIs- most investigators will submit different proposals to the different programs and so new programs increase the number of targets they get to shoot at. However, all new initiatives will get some pushback from People who feel threatened by them. The only way to eliminate such criticism is via the “3-F’s rule”: Fully Fund my proposals the First time I submit them. Initiative-based programs may create important science that can’t be created any other way, and so have strong positives. But, they also have strong negatives. There will always be criticisms.

Challenge: Changing definition of “success”: ARCSS started with “domain” programs: GISP2 and PALE, then developed into LAII and OAII These programs were concrete and built the ability to integrate further: From the new perspective, some have criticized the earlier programs for not being more integrative. The more conceptual and synthetic the questions, the harder they can be to sell to the “outside” This is one of the issues that can produce criticisms. The involved community moves ahead of their peers who are not similarly involved in more synthetic science. In fact, the entire scientific community is becoming better at this, but it is easier to sell more concrete programs.

Challenge 3: Planning Good initiatives hit the “sweet spot”: Broad enough to draw a diverse community Narrow enough to have focus and coherence How do you decide who is NOT invited to the party? Hitting the sweet spot is HARD. The tendency is to make things more integrative and more inclusive. Its hard to dis-invite people to the party and when you do, you have built opponents to the initiative and maybe to the program. But, programs that get too big lose focus and collapse. We struggle with this.

This is a real challenge: Planning, community building, and community maintenance requires a lot of work and energy Leadership needs to be altruistic: Working for the program as a whole NOT representing a “constituency” A limited pool of talent that must be grown This is a hard and produces some of the negative impressions associated with the program. People outside the planning may feel left out and assume that the “insiders” are arranging things to best suit themselves. Speaking from my 1-year’s experience on the ARCSS committee, I have been extraordinarily impressed with the dedication, altruism, and community vision the members of the ARCSS committee have shown. The focus is on understanding the Arctic System as a whole, and that requires people who are willing to see Arctic System Science as their community that they are serving, rather than terrestrial ecologists or oceanographers, for example as the community they are serving. I don’t think that the community fully appreciates this.

Final Synopsis ARCSS has been enormously successful Key Elements in a successful Initiative: Core: Set of questions that pulls researchers together Scale: Hit the sweet spot Structure: Build and maintain community Final product: science that is important, exciting, and fun Overall: the science that comes out of ARCSS has been really important. But- getting there is HARD. It requires a lot of community development, interaction, planning, and coordination to develop the programs that produce that science.