![Data Assimilation can mean very(!) different things to different people
Science goal / application à determines the framework
[Stammer et al., 2016]
• I have data from 20XX-YY,
will it help the model before
and after?
• I have observations from
20XX, why can’t the model
match the data?
• ECCO uses … (e.g., SST) to
“constrain” the state
estimate](https://image.slidesharecdn.com/morningkeynote-230322111809-a24c6d7f/85/morningkeynote-pdf-28-320.jpg)
![Polar Cyberinfrastructure
• The Polar Cyberinfrastructure program considers proposals that
promote effective collaboration between Polar and cyberinfrastructure
researchers (in both the Arctic & Antarctic solicitations).
• …and/or will support proposals that provide significant benefit to the
Polar research community including
o cost-effective transfer of data from remote field locations;
o long-term sustainable curatorship, standardization, management and discovery
of data and metadata;
o visualization, manipulation, and analysis, particularly for understanding
complexity; [including AI & ML]
o access and interoperability across scientific disciplines;
o promotion of effective use of High Performance Computing (HPC) for direct and
sustainable advances in current Arctic research; and
o e-learning and educational tools based on cyberinfrastructure components.](https://image.slidesharecdn.com/morningkeynote-230322111809-a24c6d7f/85/morningkeynote-pdf-75-320.jpg)
morningkeynote.pdf
- 1. Decolonizing US-Based Research in Greenland Future of Greenland Ice Sheet Science (FOGSS) Workshop 22 March 2023 Melody Brown Burkins, PhD Director, Institute of Arctic Studies UArctic Chair in Science Diplomacy & Inclusion Steering & Host Committees, International Conference on Arctic Research Planning (ICARP IV) Dartmouth
- 3. Alaska Icefield Experience (1990 https://www.youtube.com/watch?v=W 0iv7EYFiE Picture Credit: Daniel Otto; https://juneauicefield.org/what-to- expect/field-camps
- 4. Established in 1989, We Are: ● International and Transdisciplinary ● Host to Opportunities for Students in Experiential Education, Research, and Policy ● Host to Opportunities for Faculty to Coordinate, Collaborate, & Engage in Arctic Policy ● Host to Diverse Arctic Knowledge Holders (Fulbright Canada Research Chair in Arctic Studies) ● Hub for Building and Maintaining Trusted Partnerships with US & Global Arctic institutions and Arctic Indigenous Peoples’ Organizations ● Hub for Developing Informed, Inclusive, Impactful Solutions to Arctic & Global Policy Challenges, from Climate Change to Governance, Security, Health, and Human Rights
- 5. Scholarship & Solutions Informed by Arctic Knowledge The John Sloan Dickey Center for International Understanding
- 7. Arctic Indigenous Peoples’ Guidance for Ethical and Equitable Engagement ‘Nothing About Us Without Us’ • Co-Design of Research Questions with Communities • Recognize Indigenous Knowledge in its Own Right & Respect of Diverse Ways of Knowing • Commitment to Knowledge Co-Production • Commitment to Knowledge Reciprocity • Uphold Rights to Self-Determination (UNDRIP, FPIC) • Respect of Language, History, and Culture • Practice Good Governance • Exercise Accountability - Build Trust • Equitably Fund Indigenous Representation & Knowledge Summary from documents in previous slide, with specific points from the Inuit Circumpolar Council synthesis report (2022) Circumpolar Inuit Protocols for Equitable and Ethical Engagement (ICC EEEP) https://hh30e7.p3cdn1.secureserver.net/wp-content/uploads/EEE-Protocols-LR-WEB.pdf
- 8. Oct 2018 June 2021
- 9. • Knowledge co-production has been linked to cultivating trust, capacity, and knowledge flows, which can assist learning within the participating stakeholder groups, build networks, foster social capital, strengthen funding for collaborative research, inform policy formation, rally public acceptance, and develop actions that contribute to sustainability (Arnott et al. 2020; Norström et al. 2020; Reyers et al. 2015). • The processes of bridging cultural/epistemological differences force partners to openly confront histories of colonisation, reflect on their positions as researchers, and develop decolonising methods to redress those histories in pursuit of Indigenous data sovereignty, ownership, and intellectual property rights (CTKW 2014; Hill et al. 2020b; Maclean et al. 2021; Zurba et al. 2019). from Zurba et al. (2021): https://link.springer.com/article/10.1007/s11625-021-00996-x
- 10. • Inclusive • Co-Developed • Ethical & Equitable Knowledge Solutions • Inclusive • Co-Developed • Ethical & Equitable Effective & Sustainable Inclusive Climate Research
- 12. “Addressing unprecedented Arctic environmental changes requires listening to one another, aligning values, and collaborating across knowledge systems, disciplines, and sectors of society.” 2022 Arctic Report Card
- 13. 2018 2021 Bilateral collaboration between the Greenland (Kalaallit Nunaat) and United States Research Communities – from a vision to everyday practice Jennifer L. Mercer , Josephine Nymand, Lauren E. Culler , Rebecca Lynge, Sten Lund, Bo Gregersen, Brett Makens, Ross A. Virginia and Kristian G. Moore
- 14. 2022 The solution lies in climate resilient development… This why the choices made in the next few years will play a critical role in deciding our future and that of generations to come. To be effective, these choices need to be rooted in our diverse values, worldviews and knowledges, including scientific knowledge, Indigenous Knowledge and local knowledge. This approach will facilitate climate resilient development and allow locally appropriate, socially acceptable solutions.
- 16. When it comes to climate change, Greenland takes center stage, and it holds many of the answers we’re looking for. This explains why researchers from all over the world find their way here. The knowledge gained in Greenland is valuable and has the potential to drive society forward – both locally and globally. For that reason, it is important to have a comprehensive unit in place that can ensure that the knowledge gained here is guaranteed the greatest possible outreach and impact, and that Greenland is not just a backdrop for research, but an active participant.
- 18. 10 Year Plan: “Indigenizing Arctic Research”
- 19. Fourth International Conference on Arctic Research Planning ICARP IV (2022-2026) Fifth International Polar Year IPY-5 (2032-2033)
- 20. www.iasc.info
- 21. • Inclusive • Co-Developed • Ethical & Equitable Knowledge Solutions • Inclusive • Co-Developed • Ethical & Equitable Effective & Sustainable Inclusive Climate Research
- 22. What can you do? • Practice and Teach Ethical & Equitable Engagement in Arctic Research & Scholarship • If You Haven’t Yet: Share Your Research with the Arctic Hub of Greenland (https://arctichub.gl) • Co-Design Future Research with Greenland’s Researchers • Consider a FOGSS Contribution to ICARP IV (www.iasc.info) in 2023
- 24. Data-model integration, ECCO perspectives An T. Nguyen The University of Texas at Austin Future of Greenland ice Sheet Science Workshop, March 22-23, 2023
- 25. How physical oceanographers have been ingesting/assimilating data to get to a better state estimate Focus of talk https://www.ecco-group.org/docs/ss_2019_ECCO_FHL_assim_2.pdf
- 26. Resulting framework (in ECCO approach): - is data-constrained & dynamically coherent - allows observations to be viewed in fullest space-time context - allows connection & comparison of diverse observations - exposes observation impact on underlying model Why data assimilation?
- 27. It’s all about … – making optimal use of, – consistently extracting, – or combining information contained in observations and physical laws expressed through a model, and taking into account all uncertainties. What is data assimilation?
- 28. Data Assimilation can mean very(!) different things to different people Science goal / application à determines the framework [Stammer et al., 2016] • I have data from 20XX-YY, will it help the model before and after? • I have observations from 20XX, why can’t the model match the data? • ECCO uses … (e.g., SST) to “constrain” the state estimate
- 29. Data Assimilation: Smooth/adjoint vs Filter Approach Ø Strictly obeys model physics at all time Ø Bring all observations into a dynamically consistent description of the past and recent time-varying ocean circulation. Ø Bring all observations into a model for the purpose of prediction / forecasting Ø Model updates can break conservation law Ø days to months timescale Ø Initialization, operational Ø Ocean dynamics & variability, global & regional energy, heat & water budgets. Ø Decadal to multi-decadal timescale. Ø non-linear inversion, iterative Ø Update: priors ßà data-model misfits Data Assimilation (filter, sequential): Smoothed trajectory Filter trajectory Data Time State Estimation (smoother/adjoint, non-sequential):
- 30. Combine two incomplete information sources Observations (“data”): • incomplete/sparse probing of the physical system – spatial sampling – temporal sampling – incomplete state • different physical variables • heterogeneous data streams • measurement errors WHOI database (hydrography) Argo T/P, Jason GRACE WOCE
- 31. Combine two incomplete information sources Physical model: • representation of time-evolving state via equations of motion, conservation laws, theory, … • A dynamical interpolator • uncertainties/errors: – initial conditions – boundary conditions (surface, bottom, lateral) – model parameters (e.g., internal mixing coefficients) – “model errors” (formulation, discretization, …)
- 32. ECCO Misfit Function, Part 1 Model -- data misfit: y(t): data ; x(t): model T,S,U,V etc. R(t): error covariance Initial conditions: x(0): initial guess, e.g., WOA18, PIOMASS x0: optimized (e.g., data constrained) P(0): error covariance Uncertain Parameters ß different from just initial condition DA problem such as in NWP u(t): input parameter adjustments to e.g., time-dependent 2D+1D atmospheric forcing (from ECMWF for example) 3D time-mean internal mixing coefficients, horizontal eddy-stirring Q(t): error covariance ECCO: Forget el al. 2015 ASTE: Nguyen et al., 2021a
- 33. Uncertainty scale of in situ obs. model grid length-scale Nguyen et al., 2021b Fenty, 2010, Ph.D. thesis Fenty & Heimbach, JPO, 2013a,b Data error Model representation error
- 34. Uncertainty in inputs, e.g. surface forcing Collow et al., 2020 Graham et al., 2019, Evaluation of Six Atmospheric Reanalyses over Arctic Sea Ice from Winter to Early Summer Collow et al., 2020, Recent Arctic Ocean Surface Air Temperatures in Atmospheric Reanalyses and Numerical Simulations ~6ºC
- 35. Where to map this error/bias in atmospheric reanalysis to? Data Assimilation, an example Forcing from Reanalyses ∂h ∂t = S Fi + dF i source? Conservation equation for thickness h ha(t1) = hp(t1) + dh(t1) dh(t1) = dh(t1) Based on statistics? Artificial jumps (e.g., most of reanalyses) dh(t1) = f(dT(t1)) Propagated through all physics? (ECCO)
- 36. ECCO Misfit Function, Part 2: Physical consistency • State estimates integrate very diverse sets of data using models under strict constraint of physical consistency https://www.ecco-group.org/docs/ss_2019_ECCO_FHL_assim_2.pdf Model physics Part 1
- 37. Example: ASTE R1, subpolar gyre + N. Atlantic Compare to Argo data, 2002-2015, 1600-2000m qArgo – qi46 sq 0 10 -10 qArgo – qR1 sq Nguyen et al., 2021 kredi,450-500m kz,450-500m 100 104 10-3.8 10-5.2 m2/s m2/s
- 38. Example: ASTE R1, Greenland & Labrador Seas Nguyen et al., 2021
- 39. Summary State estimates: - allows observations to be viewed in fullest space-time context - data-constrained & is dynamically coherent - Long term variability can be meaningfully investigated Thank you! Questions?
- 40. Carton et al., 2019 Xie et al., 2017 Investigation of changes, caution
- 41. Water in Greenland Olga Sergienko Princeton University/GFDL
- 42. Credit: Ian Willis (SPRI) At the surface
- 43. Smith et al. (2015) Supraglacial streams and rivers July 2012 satellite imagery analysis
- 44. Surface runoff from firn areas Tedstone and Machguth (2022) satellite imagery analysis 2013-2020 1985-1992
- 45. Drainage of supraglacial lakes Maier et al. (2023) 2018 winter drainage
- 46. At the bed Credit: Tom Cowton (St. Andrews)
- 47. Chu et al. (2016) Perennial subglacial water radar reflectivity Russell Glacier and Isunnguata Sermia 2010 melt season
- 48. Harper et al. (2021) Perennial subglacial water borehole measurements Russell Glacier and Isunnguata Sermia 2011-2017 winter seasons
- 49. Forster et al. (2014) In the interior Perennial firn aquifers
- 50. Chu et al. (2018) Perennial firn aquifers Helheim Glacier 2012 IceBridge survey
- 51. Culberg et al. (2021) Ice slabs
- 52. J. Miller et al. (2022) Mapping of ice slabs and firn aquifers L-band brightness temperature imagery SMAP satellite
- 53. J. Miller et al. (2022) Mapping of ice slabs and firn aquifers 2017 IceBridge survey
- 54. O. Miller et al. (2017) Lateral water flow
- 55. Firn models Vandecrux et al. (2020) All one dimensional
- 56. Firn models Vandecrux et al. (2020) All one dimensional used in GSFCv1.2
- 57. Hight and volume change ICESat-2 laser altimetry Ben Smith (AGU 2022) GSFCv1.2
- 58. Seasonal vs secular trends ICESat-2 laser altimetry Ben Smith (AGU 2022) CW SW NE Residual to firn model (GSFCv1.2)
- 59. Regional climate models Fettweis et al. (2013) RACMO and MAR
- 60. RACMO and MAR calibration/evaluation PROMICE AWS Fausto et al. (2021) SUMup firn cores Alexander et al. (2019)
- 61. PROMICE AWS Fausto et al. (2021) SUMup firn cores Alexander et al. (2019) J. Miller et al. (2022) RACMO and MAR calibration ice slabs and firn aquifers
- 62. Water in Greenland — Non-negligible component of the mass balance — Has direct and indirect effects on • other ice-sheet components (snow, firn and ice) • climate components (atmospher/ocean/biosphere) — Present at the surface, bed, in the interior and easily moves — Affects altimetry observations
- 63. Wish list — Close the water budget — Improve understanding and quantify its effects — Estimate the residence time
- 64. Pathways — Close the water budget — Improve understanding of its effects and quantify them — Estimate the residence time Systematic monitoring (quantitative measurements) Systematic monitoring (quantitative measurements) Development of multiphase snow/firn/ice models
- 65. Supporting Open Polar Science at NSF Allen Pope, [email protected]
- 66. 2023: Federal Year of Open Science • “Open Science is the principle and practice of making research products and processes available to all, while respecting diverse cultures, maintaining security and privacy, and fostering collaborations, reproducibility, and equity.” • Multi-agency initiative to spark change and inspire open science engagement and adoption. • https://open.science.gov/ oNASA’s Transformation to OPen Science & Open Source Science Inititative oNSF FAIROS RCNs, GEO OSE, and more
- 67. “Nelson” Memo on Public Access • Goal: Ensuring Free, Immediate, and Equitable Access to Federally Funded Research • Tells agencies to, among other things, update public access policies as soon as possible (and no later than December 31st, 2025) to make publications and their supporting data resulting from federally funded research publicly accessible without an embargo on their free and public release. • NSF Public Access Repository: https://par.nsf.gov/ oCurrently publications are uploaded (with persistent identifiers) to PAR via reporting, but with a 1-year embargo
- 68. What is Open Science? • An Open Science Ecosystem • Open science aims to ensure the free availability and usability of scholarly publications, the data that result from scholarly research, and the methodologies, including code or algorithms, that were used to generate those data. This & subsequent slides draw from, “Open Science by Design: Realizing a Vision for 21st Century Research” (NASEM, 2018)
- 69. Why Open Science? • Rigor and reliability • Ability to address new questions • Faster and more inclusive dissemination of knowledge • Broader participation in research • Effective use of resources • Improved performance of research tasks • Open publication for public benefit
- 70. Limitations • Costs & infrastructure • Structure of scholarly communications • Lack of supportive culture, incentives, and training • Privacy, security, and proprietary barriers to sharing • Disciplinary differences
- 71. Limitations – that NSF is hoping to address • Costs & infrastructure • Structure of scholarly communications • Lack of supportive culture, incentives, and training • Privacy, security, and proprietary barriers to sharing • Disciplinary differences
- 72. OPP Data Policy • Leapfrogging with the polar research community • 16-055 was updated by 22-106 o “Dear Colleague Letter: Office of Polar Programs Data, Code, and Sample Management Policy” o Like the title says – all data, code, and samples o Advancing FAIR & CARE o Everything to be put in long-lived, publicly-accessible repository o Timeline: within 2 years of collection or by end of the award, whichever is first o Upload (meta)data to Arctic Data Center and/or US Antarctic Program Data Center o Reporting guidance o Complementary resources for DMPs, choosing repositories, etc. • Default to open – aligns with the statement, “as open as possible, as closed as necessary.”
- 73. Data Management Principles FAIR Principles for Data Management • Findability • Accessibility • Interoperability • Reusability • https://www.go-fair.org/ fair-principles/ CARE Principles for Indigenous Data Governance • Collective Benefit • Authority to Control • Responsibility • Ethics • https://www.gida-global.org/care
- 74. Open Polar Science Resources • Arctic Data Center - https://arcticdata.io/ o(Meta)data repository, DMP tools, trainings, portals, resources & more! • US Antarctic Program Data Center - https://www.usap-dc.org/ oIncludes former Antarctic Glaciological Data Center content o(Meta)data repository, Antarctic Treaty compliance, trainings, DMP tools, & more! • Polar Geospatial Center - http://pgc.umn.edu/ oDifferent services depending on current funding, but something for everybody oOpen data products like DEMs, commercial high-rez satellite data, outreach/training, geospatial services • Physical Repositories: Ice Core Facility, Polar Rock Repository, Marine Core Repository
- 75. Polar Cyberinfrastructure • The Polar Cyberinfrastructure program considers proposals that promote effective collaboration between Polar and cyberinfrastructure researchers (in both the Arctic & Antarctic solicitations). • …and/or will support proposals that provide significant benefit to the Polar research community including o cost-effective transfer of data from remote field locations; o long-term sustainable curatorship, standardization, management and discovery of data and metadata; o visualization, manipulation, and analysis, particularly for understanding complexity; [including AI & ML] o access and interoperability across scientific disciplines; o promotion of effective use of High Performance Computing (HPC) for direct and sustainable advances in current Arctic research; and o e-learning and educational tools based on cyberinfrastructure components.
- 76. Supporting Open Polar Research Software • Builds on Supporting Data and Sample Reuse in Polar Research (DCL 21-041) • Encouraging: o Translating / modernizing code into open languages/licenses o Upgrading, refining, and/or documenting code to enable broad use o Training, workshops, hackathons, cohort-building activities, etc. o Broadening participation and building open cyberinfrastructure skills and capacity in the polar research community across career stages and career paths. • BUT… o Make sure to align with other policies described earlier, and o Make sure it doesn’t fit in another NSF program • Both supplements and new proposals are welcome • Dear Colleague Letter 23-053 https://beta.nsf.gov/funding/opportunities/supporting-open-polar-research-software
- 77. GEO-wide & NSF-wide Collaboration • Works across GEO and with the Office of Advanced Cyberinfrastructure o https://www.nsf.gov/geo/geo-ci/index.jsp • Opportunities like: o Geosciences Open Science Ecosystem (GEO OSE; follow-on from EarthCube) o Advancing Geosciences using AI & ML (DCL, NSF 23-046) o Cyberinfrastructure for Sustained Science Innovation o CyberTraining o Strengthening the Cyberinfrastructure Professionals Ecosystem o CloudBank & PATh • This includes open, flexible, scalable, and interoperable cyberinfrastruture that will enable a broad and diverse community of geoscientists to integrate data, models, software, and knowledge. • GEO is looking to building on existing efforts by promoting an ecosystem for geoscience research that: o Links data & models, big data & small data, networking knowledge from diverse perspectives, o Builds on EarthCube experience and investments to move towards flexible, scalable workflows and tools, and o Provides exemplars for advancing NSF priority areas for open & inclusive science, FAIR data, reproducibility & replicability, CARE & TRUST principles, and BA-JEDI.
- 78. NSF Support of Open Science • Open science is a part of existing & likely upcoming solicitations – Full range of proposal types! oStandard/Collaborative, Supplements, Planning Proposals, RCNs, Conference/Workshop, CAREER, Mid-Career Advancement, Career-Life Balance, REUs, RAPID, EAGER, TCUP, RUI/ROA, FASED, HBCU-EiR, etc. • Takeaway: Do your prep, write a one-pager, and reach out to POs! Don’t be afraid to ask - NSF supports all sorts of (well-reviewed) great ideas!
- 79. Alexander Robel | 2023 FOGSS Workshop | Georgia Tech From Greenland to Georgia A Perspective on Sea Level Projections and What Ice Sheet Scientists Can Do For Coastal Communities 2015 US Hwy 80 Drone video courtesy of Sean Compton
- 81. 0 20 40 60 80 High Tide (inches above NADV88 datum) 0 5 10 15 20 25 30 Occurences per year High tides on Highway 80 (Fort Pulaski Tide Gauge) 1935-1940 2014-2019 Flooding on Highway 80 ~1 ft local sea level rise
- 82. Starting in 2019, GA DOT spent first several million $ to raise lowest parts of Highway 80 by 8 inches
- 83. 0 20 40 60 80 High Tide (inches above NADV88 datum) 0 5 10 15 20 25 30 Occurences per year High tides on Highway 80 (Fort Pulaski Tide Gauge) 1935-1940 2014-2019 Flooding on Highway 80 0 10 20 30 40 50 60 70 80 High Tide (inches above NADV88 datum) 0 5 10 15 20 25 30 Occurences per year High tides on Highway 80 (Fort Pulaski Tide Gauge) 1935-1940 2014-2019 Flooding on Highway 80 Flooding on Raised Highway 80
- 84. A Chatham County, GA SLR Projections (NOAA) •“Typical” design life for a residential building is 30 years: ~1 ft of uncertainty in SLR •Multifamily building design life closer to 50 years: 2-3 ft uncertainty in SLR •Critical infrastructure (bridges, physical plant, etc) design life 75 years: 4 ft uncertainty in SLR
- 85. A Chatham County, GA SLR Projections (NOAA) Cost of a mistake: Homeowner: •Seawall install: $1-2k/foot •Seawall cap: $100-200/foot •Chronic flooding: loss of home value Government: • Beach nourishment: ~$million/mile/year • NFIP premium subsidy: ~$billion/year
- 86. A problem: most communities use little or no information from state-of-the-art ice sheet/sea level projections when planning for sea level rise Hirschfeld et al. 2022 North American survey respondents
- 87. So…how can ice sheet scientists help facilitate effective adaptation with the best possible information?
- 88. 1. Build relationships with boundary organizations or directly with practitioners to address community needs (Ultee et al., Earth’s Future, 2018) 2. Recruit students from frontline communities into ice sheet science to internalize on-the-ground expertise that can work parallel to co-production (Robel, Ultee, Ranganathan, Nash, In Prep)
- 89. Listen to communities and practitioners tell us what they need https://www.cearhub.org/
- 90. Tell communities and practitioners what we know, what we don’t know, and what we need • Advocacy recognizing the immense financial implications tied to sea level projections • $12-71 billion/year globally ($3-4 billion/year in US) for coastal adaptation/armoring which uses sea level projections from ice sheet models (Hinkel et al. 2014, Neumann et al. 2014) • NSF funding for ice sheet modeling is less than $3 million/year, NASA/ DOE+other agencies likely less than $10 million total
- 91. Need: short-term sea level projections (<50 years) Potential users: homeowners, engineers, planning professionals Need better methods to transiently assimilate observations of recent ice sheet change to initialize future projections Aschwanden et al. 2021 V2015 − V2000 ∝ V2060 − V2015
- 92. Need: short-term sea level projections (<50 years) Potential users: homeowners, engineers, planning professionals Certain ice sheet processes are more important for short-term projections (also more and varied uncertainty quantification) Aschwanden et al. 2019
- 93. Need: long-term sea level projections (>50 years) Where do/should we build new communities in coastal areas? The most important processes are… 🤐 Not going to catch me in this trap! The reality is that more and better uncertainty quantification and optimal experimental design are needed to formally answer these questions for processes that already have parameterizations in models. Most UQ and Bayesian calibration studies have focused on Antarctica. Greenland Examples:
- 94. Need: long-term sea level projections (>50 years) Where do/should we build new communities in coastal areas? But, this isn’t enough because: (1) We don’t have well-informed priors on all the parameters we do have in models. (2) not at all processes are parameterized in any way in most models (aka structural model uncertainty) implying perfect confidence of zero importance We need structural change in our funding and educational model to: “Grow the modeler pipeline and treat models as instruments that require maintenance to continue operation.” (2022 FOGSS Report) Frequency Parameter Value
- 95. But…the information communities want and need is not always global sea level projections
- 96. Need: translating knowledge across disciplines Extending knowledge beyond ice sheets InSAR/GNSS for estimating vertical land motion Shirzaei et al. 2021
- 97. Need: translating knowledge across disciplines Extending knowledge beyond ice sheets Local inundation/ flood modeling using similar systems to those used for regional ice sheet modeling Park et al., Coastal Engineering, 2022
- 98. Need: building interdisciplinary teams Team science ensures that ice sheet knowledge isn’t siloed within our community and that we can address cross-cutting issues beyond ice sheets
- 99. • Fundamental science is important for the long-term enterprise of understanding and predicting ice sheet change • But…producing the most usable science requires: 1. Building relationships with communities that we want to use the scientific knowledge that we produce through intermediaries or by leveraging local connections to our institutions/groups 2. Asking communities what they actually want/need for effective adaptation 3. Working effectively to build capacity within our discipline to target those needs 4. Translating our skills to problems outside of ice sheet science 5. Building interdisciplinary teams when our skills or capacity fall short Takeaways