Cornell University / Institute for Computation Sustainability
Carla P. Gomes, Professor and Director
Institute for Computational Sustainability - Faculty of Computing and Information Science
Department of Computer Science and Department of Information Science and
Dyson School of Applied Economics and Management
5133 Upson Hall
Ithaca, NY 14853
Tel: (607) 255-9189
Fax: (607) 255-4428
Ole M. Amundsen III
Strategic Conservation Program Manager
The Conservation Fund
376 Turkey Hill Road
Ithaca, NY 14850
Office Phone: (607)-277-0999
The Institute for Computational Sustainability (ICS) based at Cornell University is designed to serve as the nexus of foundational science advancements and practical applications in sustainability. In particular, the ICS team has strong focus on large landscape conservation. Preserving and restoring habitat connectivity has been identified as a key conservation priority for government agencies and conservation organizations. Several recent studies have shown that, in order to design conservation strategies that are efficient and practical, it is crucial to incorporate economic, ecological, and biodiversity considerations from the outset of the planning process. Yet, computational tools that allow conservation planners to systematically study the tradeoff between cost and ecological benefit as well as obtain conservation plans with optimality guarantees are largely lacking. The Cornell team has achieved several computational advances in conservation planning by bringing in techniques from discrete optimization and network design.
University researchers are partnering with The Conservation Fund, a national nonprofit land conservation organization, to improve computation methods for producing computer models for wildlife conservation. The first phase of the effort developed computational methods for designing cost-effective conservation strategies for functional connectivity. Metapopulation models are widely used in ecology to describe the stochastic population dynamics of species. Hence, they are extremely relevant when designing conservation plans with the goal of increasing the expected species population in a study area. In particular, the collaboration between Cornell and TCF focused on locating sites for facilitating the movement of the Red Cockaded Woodpecker (RCW), a federally endangered species that occurs in at least 10 states: VA, NC, SC, GA, FL, AL, MS, LA, TX, and AR. The Conservation Fund (TCF) currently owns a 10,000 acre RCW mitigation bank in Tyrrell County, NC called the Palmetto Peartree Preserve (P3). This partnership resulted in a computational model that couples optimization and metapopulation dispersal models to help find the most effective locations for creating new RCW habitat among existing or future recovery units while satisfying a strict budget constraint [Sheldon2010]. The goal of the collaboration between TCF and the Cornell team is utilizing computational methods to identify strategies for cost effective expansion at P3 and identifying opportunities to host birds from other recovery units in North Carolina that are more susceptible to habitat fragmentation over time.
The ICS team has worked for several years on developing scalable approaches for wildlife corridor design --- in particular, for designing a connected wildlife corridor network of maximum habitat utility within a specified budget constraint [Conrad2012, Conrad2007, Gomes2008, Dilkina2010] The techniques were analyzed on a case study involving three core habitat areas (Yellowstone, Salmon-Selway Ecosystem and the Northern Continental Divide) for grizzly bears across a large-scale fragmented landscape spanning Montana, Idaho and Wyoming. In a different effort, the ICS team studied the problem of selecting which parcels of land to put under conservation protection in the face of possible future land use changes and habitat degradation. The conservation strategy here is one of maximizing the future landscape connectivity between pairs of core areas again subject to a budget constraint [Dilkina2011]. The ICS is currently collaborating with Michael K. Schwartz, Kevin S. McKelvey and David E. Calkin from the USDA Forest Service, Rocky Mountain Research Station, as well as Claire Montgomery from Oregon State University to apply this approach to multi-species conservation planning in Montana concerning wolverines and grizzly bears [Lai2011, Dilkina2011]. Most recently, the ICS team has concentrated on conservation strategies for wildlife corridors that incorporate resilience and reliability considerations.
The Cornell team is partnering with TCF to apply and further refine several of these conservation planning techniques in the context of a statewide plan for the West Virginia Division of Natural Resources. TCF developed core habitat and landscape resistance models for the Allegheny Woodrat, Eastern Cricket Frog, the Timber Rattle Snake and the West Virginia Flying Squirrel, as well as umbrella landscape resistance models for forest, aquatic and wetland species. These models were provided to the Cornell Team for use in developing corridor conservation plans. Most statewide green infrastructure plans are solely focused on ecological or species parameters. The Cornell team is working on providing models that take into account land values, restoration option and cost factors to make a more realist corridor plan. One challenge to designing cost-effective conservation strategies for West Virginia is the lack of parcel and land value data for the state. The Conservation Fund and the Cornell Team are addressing this by developing a predictive model of cost per acre based on proximity to urban areas.
The Cornell team provides advance optimization modeling on policy trade off in the implementation of corridor systems and conservation planning. Federal and state budgets are under increased pressure making the prospect of additional funding for the preservation and stewardship of rare and endangered species less likely. As the saying goes, government agencies “have to do more with less”. However the use of optimization techniques holds the promise of making that expression a reality. Optimization provides decision makers with information to benefit a species such as the RCW at a landscape scale, crystallizing trade-offs between regions, making resource allocations based on the actual needs of the species while taking into account economic costs.
The Institute for Computational Sustainability seeks to inject computational thinking into sustainability, and therefore its members strongly champion interdisciplinary collaborations. The Institute’s key academic discipline is computer science. However, its members include representatives of disciplines ranging from mathematics and information science, to economics, ecology, atmospheric sciences, forestry, ornithology, conservation planning, natural resources, and more.
12 Departments at Cornell University, Bowdoin College, the Conservation Fund, Howard University, Oregon State University, and the Pacific Northwest National Laboratory.
The Institute for Computational Sustainability was originally envisioned by Professor Carla Gomes (Cornell), who brought together the partner organizations listed above in 2007/2008 to propose this computational sustainability initiative to the National Science Foundation (NSF). The NSF awarded a 5-year grant to initiate this work in August of 2008, and it has been a continued champion of these efforts, along with the Atkinson Center for a Sustainable Future at Cornell University. In June 2009, ICS held the 1st International Conference on Computational Sustainability, with over 200 participants from the U.S. and abroad. Since this time, the ICS has seen growing interest in this research through new research collaborations, development of workshops and special tracks on computational sustainability at conferences, and the continued success of the International Conferences on Computational Sustainability. In July 2012, the International Conference on Computational Sustainability was held in Copenhagen, Denmark. This marks the first time this conference was held outside of the U.S., and highlights the growing interest around the world in this new research area.
The Institute for Computational Sustainability was founded to bring the analytical insight of the field of computer science to bear on a wide range of issues in sustainability. The complex nature of questions posed via sustainability promise to enrich and transform computer science through interdisciplinary cross pollination and real world application.
In terms of on-ground efforts, TCF has utilized the modeling work conducted by Cornell to identify and assess key parcels for the expansion of P3. Significant financial investments in sizable acreage holdings supported in part from scientifically supported research that the land acquisition will enhance the recovery of the RCW. In general, the work of the ICS team has the potential of significantly contributing to conservation planning by providing a critically useful tool for systematically studying not only tradeoffs between economic costs and ecological benefits, but also tradeoffs between different species.
Through the collaboration with partners such as the Conservation Fund and the USDA Rocky Mountain Research Station, the Cornell research team identifies general conservation planning computational challenges underlying the particular on-the-ground conservation initiatives. The conservation planning research at the Institute for Computational Sustainability is targeted at developing and advancing computational methods that applicable to broad conservation settings. Hence, the approaches and tools are highly transferrable to conservation initiatives at other institutions and places. In particular, the decision-support tools for landscape connectivity planning developed by ICS can inform other conservation planning efforts about important tradeoffs between economic costs and ecological benefits in a systematic and defendable way. A key lesson learned is the importance of highly interdisciplinary collaborations that involve biologists, ecologists and conservationists but also computer scientists that can contribute an essential set of techniques and insights and can significantly advance the field of conservation. Interestingly, the computational challenges brought by the direct real world applications in conservation planning have also advanced the field of computer science as the underlying problems and their computational solutions go beyond conservation and are in fact isomorphic to other important problems that arise in transportation, sensor networks, telecommunications, and computer circuit design.
This initiative and its impacts are likely to endure on a long-term basis. A critical piece of the ICS mission is to ensure the growth of a research community committed to pursuing and sustaining computational sustainability as a research area. The increasing interest in holding workshops, special conference tracks, and other meetings related to computational sustainability indicates that this community is growing far beyond the original members of this initiative. The tremendous interest in the area by students an early-career scientists also holds promise for the continued pursuit of research in this area. In terms of impacts, this initiative aims to study and provide solutions to computational problems for balancing environmental, economic, and societal needs for a sustainable future while advancing the state of the art of computing and information science. The intention is for these impacts to endure over generations.
Part of the mission and outreach of the Institute is to establish a vibrant and diverse research community in the area of computational sustainability, drawing new students into the field from all backgrounds including students from underrepresented groups via summer research experiences and other such proactive activities. The complexity of the real world problems posed by TCF is a welcomed challenge for undergraduates and graduate students as well as faculty who are hungry to make a difference in the world. By brining students and faculty into contact with real world environmental problems, researchers face problems of scale, complex interactions between model elements and public policy implementation factors that are often overlooked by academic teams. This complexity is in itself rigorous, helping expand students’ ability to diagnose a problem, assess options and improve the accuracy of proposed solutions, all critical to the educational experience.