NCEAS Working Groups
Public domain ADMB project
Project Description
AD Model Builder (ADMB) is a tool for developing integrated statistical models of complex systems. The principle advantages of the ADMB software suite over other approaches are rapid model development, numerical stability, computational speed, precision of model estimates, and the capacity to accommodate relatively large numbers of parameters and data points.
The ADMB software has earned acceptance by researchers working on all aspects of resource management. Population models based on the ADMB software are used to monitor more than 150 different sensitive endangered species and commercially valuable fish stocks around the world. The populations modeled using ADMB include such diverse species as whales, dolphins, sea lions, penguins, albatross, abalone, lobsters, tunas, billfish, sharks, rays, anchovy, and pollock.
ADMB applications extend beyond stock assessment. ADMB-based software is used for applications critical to the development of place-based management policies. ADMB is an essential building block of the methods used to reconstruct movements of many species of animals tracked with electronic tags. Spatially resolved populations models treat movement as an integral component of population dynamics and depend on ADMB for estimating movement parameters from data. -
ADMB applications are critical to the missions of fishery management agencies in the United States and abroad. Stock assessments for commercially important fish stocks and ecologically sensitive protected species around the world depend on ADMB. In the United States, every NOAA Fisheries Science Center uses ADMB in some fashion, and many • commercially important and sustainably managed fisheries depend on ADMB-based stock assessments. These fisheries include, for example, the Gulf of Alaska pollock fishery, which is widely hailed as sustainably managed. The value i of the fisheries dependent on ADMB-based assessments is enormous. The combined landed value of tropical Pacific tunas and the Gulf of Alaska and Bering Sea ground fish alone exceeds US$10 billion.
ADMB is also used at Universities and other academic and research institutions. Due to ADMBs wide use in- fisheries assessment and management, it is now taught in courses at 'several universities. The research organizations, government departments, and companies using ADMB, the types of applications being used, and a list of ADMB-based publications are attached to this proposal. (Up-to-date lists can be found on-line at http://admbproject.org) to users than software whose source code remains proprietary. Involvement of the user community in software development harnesses the power of distributed peer review and assures that the software will evolve in directions most useful to its users. Providing the ADMB software at no cost enables users with limited means, such as students and scientists in developing countries, to take advantage of sophisticated state-of-the-art computing algorithms. In addition, the community built around the open source software provides an ideal resource for assisting developing countries in- adopting the technology.
Principal Investigator(s)
John R. Sibert, Mark N. Maunder
Project Dates
Start: December 7, 2007
End: January 7, 2009
completed
Participants
- Teresa A'mar
- National Oceanic and Atmospheric Administration (NOAA)
- Johnoel Ancheta
- University of Hawaii, Mānoa
- Carol A. Blanchette
- University of California, Santa Barbara
- Sylvain Bonhommeau
- University of California, Santa Barbara
- Greg A. Breed
- University of California, Santa Cruz
- Mollie Brooks
- McMaster University
- Jennifer E. Caselle
- University of California, Santa Barbara
- Frank Davenport
- University of California, Santa Barbara
- Trevor D. Davies
- Dalhousie University
- Kelly L. Decker
- NASA Ames Research Center
- David A. Fournier
- Otter Research Ltd.
- Chris Grandin
- Fisheries and Oceans Canada
- Stephanie E. Hampton
- University of California, Santa Barbara
- Carrie V. Kappel
- University of California, Santa Barbara
- Brian P. Kinlan
- University of California, Santa Barbara
- Tin Klanjscek
- University of California, Santa Barbara
- Robert Leaf
- Virginia Polytechnic Institute and State University
- Brian Linton
- NOAA, National Marine Fisheries Service (NMFS)
- Steven Y. Litvin
- Stanford University
- Weihai Liu
- Michigan State University
- Alexander Lowe
- University of California, Santa Barbara
- Arni Magnusson
- Marine Research Institute
- Steve J.D. Martell
- University of British Columbia
- Mark N. Maunder
- InterAmerican Tropical Tuna Commission
- Duncan N. Menge
- University of California, Santa Barbara
- Erik B. Muller
- Anders Nielsen
- University of Hawaii
- Mary I. O'Connor
- University of California, Santa Barbara
- Laure Pecquerie
- University of California, Santa Barbara
- Christopher R. Perle
- Stanford University
- Christine Petersen
- University of California, Santa Barbara
- Jai Ranganathan
- University of California, Santa Barbara
- James Regetz
- University of California, Santa Barbara
- Mark P. Schildhauer
- University of California, Santa Barbara
- Derek Seiple
- University of Hawaii
- John R. Sibert
- University of Hawaii, Mānoa
- Tim Sippel
- University of Hawaii, Mānoa
- Hans J. Skaug
- University of Bergen
- Ulrich K. Steiner
- Stanford University
- Crow White
- University of California, Santa Barbara
- Casper Willestofte Berg
- Technical University of Denmark
- Arliss Winiship
- Dalhousie University
Products
-
Journal Article / 2012
AD Model Builder: Using automatic differentiation for statistical inference of highly parameterized complex nonlinear models