Search Menu
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information
Search Scholarly Publications

Title: SICOPOLIS-AD v1: an open-source adjoint modeling framework for ice sheet simulation enabled by the algorithmic differentiation tool OpenAD

Abstract

We present a new capability of the ice sheet model SICOPOLIS that enables flexible adjoint code generation via source transformation using the open-source algorithmic differentiation (AD) tool OpenAD.The adjoint code enables efficient calculation of the sensitivities of a scalar-valued objective function or quantity of interest (QoI) to a range of important, often spatially varying and uncertain model input variables, including initial and boundary conditions, as well as model parameters.Compared to earlier work on the adjoint code generation of SICOPOLIS, our work makes several important advances:(i) it is embedded within the up-to-date trunk of the SICOPOLIS repository – accounting for 1.5 decades of code development and improvements – and is readily available to the wider community;(ii) the AD tool used, OpenAD, is an open-source tool;(iii) the adjoint code developed is applicable to both Greenland and Antarctica, including grounded ice as well as floating ice shelves, with an extended choice of thermodynamical representations.A number of code refactorization steps were required. They are discussed in detail in an Appendix as they hold lessons for the application of AD to legacy codes at large.As an example application, we examine the sensitivity of the total Antarctic Ice Sheet volume to changes in initial ice thickness, austral summer precipitation, andmore » basal and surface temperatures across the ice sheet.Simulations of Antarctica with floating ice shelves show that over 100 years of simulation the sensitivity of total ice sheet volume to the initial ice thickness and precipitation is almost uniformly positive, while the sensitivities to surface and basal temperature are almost uniformly negative. Sensitivity to austral summer precipitation is largest on floating ice shelves from Queen Maud to Queen Mary Land. The largest sensitivity to initial ice thickness is at outlet glaciers around Antarctica. Comparison between total ice sheet volume sensitivities to surface and basal temperature shows that surface temperature sensitivities are higher broadly across the floating ice shelves, while basal temperature sensitivities are highest at the grounding lines of floating ice shelves and outlet glaciers. A uniformly perturbed region of East Antarctica reveals that, among the four control variables tested here, total ice sheet volume is the most sensitive to variations in austral summer precipitation as formulated in SICOPOLIS.Comparison between adjoint- and finite-difference-derived sensitivities shows good agreement, lending confidence that the AD tool is producing correct adjoint code.The new modeling infrastructure is freely available at http://www.sicopolis.net (last access: 2 April 2020) under the development trunk.« less

Authors:
 [1]; ORCiD logo [2]; ORCiD logo [3]; ORCiD logo [1]
+ Show Author Affiliations
  1. Univ. of Texas, Austin, TX (United States)
  2. Argonne National Lab. (ANL), Argonne, IL (United States)
  3. Hokkaido Univ., Sapporo (Japan)
Publication Date:
Research Org.:
Argonne National Laboratory (ANL), Argonne, IL (United States)
Sponsoring Org.:
USDOE Office of Science (SC); National Science Foundation (NSF); Japan Society for the Promotion of Science (JSPS); Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan
OSTI Identifier:
1632866
Grant/Contract Number:  
AC02-06CH11357; SC0008060; 1750035; JP16H02224; JP17H06104; JP17H06323
Resource Type:
Accepted Manuscript
Journal Name:
Geoscientific Model Development (Online)
Additional Journal Information:
Journal Name: Geoscientific Model Development (Online); Journal Volume: 13; Journal Issue: 4; Journal ID: ISSN 1991-9603
Publisher:
European Geosciences Union
Country of Publication:
United States
Language:
English
Subject:
58 GEOSCIENCES

Citation Formats

Logan, Liz C., Narayanan, Sri Hari Krishna, Greve, Ralf, and Heimbach, Patrick. SICOPOLIS-AD v1: an open-source adjoint modeling framework for ice sheet simulation enabled by the algorithmic differentiation tool OpenAD. United States: N. p., 2020. Web. doi:10.5194/gmd-13-1845-2020.
Copy to clipboard
Logan, Liz C., Narayanan, Sri Hari Krishna, Greve, Ralf, & Heimbach, Patrick. SICOPOLIS-AD v1: an open-source adjoint modeling framework for ice sheet simulation enabled by the algorithmic differentiation tool OpenAD. United States. https://doi.org/10.5194/gmd-13-1845-2020
Copy to clipboard
Logan, Liz C., Narayanan, Sri Hari Krishna, Greve, Ralf, and Heimbach, Patrick. Thu . "SICOPOLIS-AD v1: an open-source adjoint modeling framework for ice sheet simulation enabled by the algorithmic differentiation tool OpenAD". United States. https://doi.org/10.5194/gmd-13-1845-2020. https://www.osti.gov/servlets/purl/1632866.
Copy to clipboard
@article{osti_1632866,
title = {SICOPOLIS-AD v1: an open-source adjoint modeling framework for ice sheet simulation enabled by the algorithmic differentiation tool OpenAD},
author = {Logan, Liz C. and Narayanan, Sri Hari Krishna and Greve, Ralf and Heimbach, Patrick},
abstractNote = {We present a new capability of the ice sheet model SICOPOLIS that enables flexible adjoint code generation via source transformation using the open-source algorithmic differentiation (AD) tool OpenAD.The adjoint code enables efficient calculation of the sensitivities of a scalar-valued objective function or quantity of interest (QoI) to a range of important, often spatially varying and uncertain model input variables, including initial and boundary conditions, as well as model parameters.Compared to earlier work on the adjoint code generation of SICOPOLIS, our work makes several important advances:(i) it is embedded within the up-to-date trunk of the SICOPOLIS repository – accounting for 1.5 decades of code development and improvements – and is readily available to the wider community;(ii) the AD tool used, OpenAD, is an open-source tool;(iii) the adjoint code developed is applicable to both Greenland and Antarctica, including grounded ice as well as floating ice shelves, with an extended choice of thermodynamical representations.A number of code refactorization steps were required. They are discussed in detail in an Appendix as they hold lessons for the application of AD to legacy codes at large.As an example application, we examine the sensitivity of the total Antarctic Ice Sheet volume to changes in initial ice thickness, austral summer precipitation, and basal and surface temperatures across the ice sheet.Simulations of Antarctica with floating ice shelves show that over 100 years of simulation the sensitivity of total ice sheet volume to the initial ice thickness and precipitation is almost uniformly positive, while the sensitivities to surface and basal temperature are almost uniformly negative. Sensitivity to austral summer precipitation is largest on floating ice shelves from Queen Maud to Queen Mary Land. The largest sensitivity to initial ice thickness is at outlet glaciers around Antarctica. Comparison between total ice sheet volume sensitivities to surface and basal temperature shows that surface temperature sensitivities are higher broadly across the floating ice shelves, while basal temperature sensitivities are highest at the grounding lines of floating ice shelves and outlet glaciers. A uniformly perturbed region of East Antarctica reveals that, among the four control variables tested here, total ice sheet volume is the most sensitive to variations in austral summer precipitation as formulated in SICOPOLIS.Comparison between adjoint- and finite-difference-derived sensitivities shows good agreement, lending confidence that the AD tool is producing correct adjoint code.The new modeling infrastructure is freely available at http://www.sicopolis.net (last access: 2 April 2020) under the development trunk.},
doi = {10.5194/gmd-13-1845-2020},
journal = {Geoscientific Model Development (Online)},
number = 4,
volume = 13,
place = {United States},
year = {Thu Apr 09 00:00:00 EDT 2020},
month = {Thu Apr 09 00:00:00 EDT 2020}
}
Copy to clipboard
Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record
https://doi.org/10.5194/gmd-13-1845-2020
Copyright Statement
Other availability
Search WorldCat Search WorldCat to find libraries that may hold this journal

Figures / Tables:

Figure 1 Figure 1: Schematic of AD applied to a simple function, $\mathcal{F}$(a,b,c) = y = sin(a · b)c (see top figure). Intermediate results of the function evaluation are stored in the intermediate variables t1 and t2. We wish to determine the gradient of $\mathcal{F}$ with respect to all inputs a,b,c, symbolicallymore » expressed as $∇\mathcal{F}$ = [ad, bd, cd]. In SICOPOLIS-AD, the entire forward code is composed of many lines of simple functions like $\mathcal{F}$ , in sequence (blue downward-pointing array on the left). OpenAD provides $∇\mathcal{F}$ by relating the partials of t1 and t2 to intrinsically differentiable functions, like sin(), here the red text, p1 (black upward-pointing array on the right). The partial derivatives of $\mathcal{F}$ are computed via the writing to memory of intermediate partial quantities, like dy (the adjoint or dual of y) , dt1, and dt2. Thus, the sought sensitivity of a QoI, $\mathcal{F}$ , is related to input parameters, a, b, and c in this algorithmic (albeit much simplified) manner.« less
All figures and tables (6 total)
Save / Share:
Export Metadata
Save to My Library
You must Sign In or Create an Account in order to save documents to your library.

Works referenced in this record:

Bedmap2: improved ice bed, surface and thickness datasets for Antarctica
journal, January 2013

  • Fretwell, P.; Pritchard, H. D.; Vaughan, D. G.
  • The Cryosphere, Vol. 7, Issue 1
  • DOI: 10.5194/tc-7-375-2013

The creep of polycrystalline ice
journal, March 1955

  • Glen, J. W.
  • Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences, Vol. 228, Issue 1175, p. 519-538
  • DOI: 10.1098/rspa.1955.0066

Evaluating Derivatives
book, January 2008

Generating efficient derivative code with TAF
journal, October 2005

The Theory of Glacier Sliding
journal, January 1964

Greenland ice-sheet volume sensitivity to basal, surface and initial conditions derived from an adjoint model
journal, January 2009

Climate Model Intercomparisons: Preparing for the Next Phase
journal, March 2014

  • Meehl, Gerald A.; Moss, Richard; Taylor, Karl E.
  • Eos, Transactions American Geophysical Union, Vol. 95, Issue 9
  • DOI: 10.1002/2014EO090001

Practical global oceanic state estimation
journal, June 2007

Comparison of Numerical Schemes for the Solution of the Ice-Thickness Equation in a Dynamic/Thermodynamic Ice-Sheet Model
journal, July 2002

  • Greve, Ralf; Calov, Reinhard
  • Journal of Computational Physics, Vol. 179, Issue 2
  • DOI: 10.1006/jcph.2002.7081

Comparison of hybrid schemes for the combination of shallow approximations in numerical simulations of the Antarctic Ice Sheet
journal, January 2017

A Three-Dimensional Time-Dependent Model of the West Antarctic Ice Sheet
journal, January 1984

Theory of shallow ice shelves
journal, February 1999

  • Weis, M.; Greve, R.; Hutter, K.
  • Continuum Mechanics and Thermodynamics, Vol. 11, Issue 1
  • DOI: 10.1007/s001610050102

Comparison of adjoint and nudging methods to initialise ice sheet model basal conditions
journal, January 2016

  • Mosbeux, Cyrille; Gillet-Chaulet, Fabien; Gagliardini, Olivier
  • Geoscientific Model Development, Vol. 9, Issue 7
  • DOI: 10.5194/gmd-9-2549-2016

initMIP-Antarctica: an ice sheet model initialization experiment of ISMIP6
journal, January 2019

A structural glaciological analysis of the 2002 Larsen B ice-shelf collapse
journal, January 2008

Greenland ice sheet contribution to sea-level rise from a new-generation ice-sheet model
journal, January 2012

Solution of Nonlinear Stokes Equations Discretized By High-Order Finite Elements on Nonconforming and Anisotropic Meshes, with Application to Ice Sheet Dynamics
journal, January 2015

  • Isaac, Tobin; Stadler, Georg; Ghattas, Omar
  • SIAM Journal on Scientific Computing, Vol. 37, Issue 6
  • DOI: 10.1137/140974407

ECCO version 4: an integrated framework for non-linear inverse modeling and global ocean state estimation
journal, January 2015

  • Forget, G.; Campin, J. -M.; Heimbach, P.
  • Geoscientific Model Development, Vol. 8, Issue 10
  • DOI: 10.5194/gmd-8-3071-2015

An efficient regional energy-moisture balance model for simulation of the Greenland Ice Sheet response to climate change
journal, January 2010

An efficient exact adjoint of the parallel MIT General Circulation Model, generated via automatic differentiation
journal, October 2005

Inference of accumulation-rate patterns from deep layers in glaciers and ice sheets
journal, January 2007

  • Waddington, Edwin D.; Neumann, Thomas A.; Koutnik, Michelle R.
  • Journal of Glaciology, Vol. 53, Issue 183
  • DOI: 10.3189/002214307784409351

Data assimilation using a hybrid ice flow model
journal, January 2011

Algorithm 799: revolve: an implementation of checkpointing for the reverse or adjoint mode of computational differentiation
journal, March 2000

  • Griewank, Andreas; Walther, Andrea
  • ACM Transactions on Mathematical Software, Vol. 26, Issue 1
  • DOI: 10.1145/347837.347846

An inexact Gauss-Newton method for inversion of basal sliding and rheology parameters in a nonlinear Stokes ice sheet model
journal, January 2012

  • Petrat, Noemi; Zhu, Hongyu; Stadler, Georg
  • Journal of Glaciology, Vol. 58, Issue 211
  • DOI: 10.3189/2012JoG11J182

Differentiable Programming Tensor Networks
journal, September 2019

An optimized treatment for algorithmic differentiation of an important glaciological fixed-point problem
journal, January 2016

  • Goldberg, Daniel N.; Narayanan, Sri Hari Krishna; Hascoet, Laurent
  • Geoscientific Model Development, Vol. 9, Issue 5
  • DOI: 10.5194/gmd-9-1891-2016

Parameter and state estimation with a time-dependent adjoint marine ice sheet model
journal, January 2013

Optimal initial conditions for coupling ice sheet models to Earth system models: PEREGO ET AL.
journal, September 2014

  • Perego, Mauro; Price, Stephen; Stadler, Georg
  • Journal of Geophysical Research: Earth Surface, Vol. 119, Issue 9
  • DOI: 10.1002/2014JF003181

Source-to-source adjoint Algorithmic Differentiation of an ice sheet model written in C
journal, November 2017

Variational Assimilation of Meteorological Observations With the Adjoint Vorticity Equation. I: Theory
journal, October 1987

  • Talagrand, Olivier; Courtier, Philippe
  • Quarterly Journal of the Royal Meteorological Society, Vol. 113, Issue 478
  • DOI: 10.1002/qj.49711347812

Programming language features, usage patterns, and the efficiency of generated adjoint code
journal, February 2016

Fitting dynamics to data
journal, January 1988

  • Thacker, William Carlisle; Long, Robert Bryan
  • Journal of Geophysical Research, Vol. 93, Issue C2
  • DOI: 10.1029/JC093iC02p01227

Various isotropic and anisotropic ices found in glaciers and polar ice caps and their corresponding rheologies
journal, December 1985

  • Lliboutry, L.; Duval, P.
  • International Journal of Rock Mechanics and Mining Sciences & Geomechanics Abstracts, Vol. 22, Issue 6
  • DOI: 10.1016/0148-9062(85)90267-0

Construction of the adjoint MIT ocean general circulation model and application to Atlantic heat transport sensitivity
journal, December 1999

  • Marotzke, Jochem; Giering, Ralf; Zhang, Kate Q.
  • Journal of Geophysical Research: Oceans, Vol. 104, Issue C12
  • DOI: 10.1029/1999JC900236

Paleothermometry by control methods
journal, January 1991

  • MacAyeal, Douglas R.; Firestone, John; Waddington, Edwin
  • Journal of Glaciology, Vol. 37, Issue 127
  • DOI: 10.3189/S0022143000005761

The ERA-40 re-analysis
journal, October 2005

  • Uppala, S. M.; KÅllberg, P. W.; Simmons, A. J.
  • Quarterly Journal of the Royal Meteorological Society, Vol. 131, Issue 612
  • DOI: 10.1256/qj.04.176

Comparison of thermodynamics solvers in the polythermal ice sheet model SICOPOLIS
journal, March 2016

Basal conditions for Pine Island and Thwaites Glaciers, West Antarctica, determined using satellite and airborne data
journal, January 2009

Inversion of basal friction in Antarctica using exact and incomplete adjoints of a higher-order model: ANTARCTIC BASAL FRICTION INVERSION
journal, September 2013

  • Morlighem, M.; Seroussi, H.; Larour, E.
  • Journal of Geophysical Research: Earth Surface, Vol. 118, Issue 3
  • DOI: 10.1002/jgrf.20125

Thermomechanical balances of ice sheet flows
journal, May 1984

An improved Antarctic dataset for high resolution numerical ice sheet models (ALBMAP v1)
journal, January 2010

Application of control methods for modelling the flow of Pine Island Glacier, West Antarctica
journal, January 2003

A new bed elevation dataset for Greenland
journal, January 2013

  • Bamber, J. L.; Griggs, J. A.; Hurkmans, R. T. W. L.
  • The Cryosphere, Vol. 7, Issue 2
  • DOI: 10.5194/tc-7-499-2013

Committed retreat of Smith, Pope, and Kohler Glaciers over the next 30 years inferred by transient model calibration
journal, January 2015

A continuum–mechanical formulation for shallow polythermal ice sheets
journal, May 1997

  • Greve, Ralf
  • Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences, Vol. 355, Issue 1726
  • DOI: 10.1098/rsta.1997.0050

Design and results of the ice sheet model initialisation experiments initMIP-Greenland: an ISMIP6 intercomparison
journal, January 2018

A new ice thickness and bed data set for the Greenland ice sheet: 1. Measurement, data reduction, and errors
journal, December 2001

  • Bamber, J. L.; Layberry, R. L.; Gogineni, S. P.
  • Journal of Geophysical Research: Atmospheres, Vol. 106, Issue D24
  • DOI: 10.1029/2001JD900054

Antarctic snow accumulation mapped using polarization of 4.3-cm wavelength microwave emission
journal, January 2006

  • Arthern, Robert J.; Winebrenner, Dale P.; Vaughan, David G.
  • Journal of Geophysical Research, Vol. 111, Issue D6
  • DOI: 10.1029/2004JD005667

MPAS-Albany Land Ice (MALI): a variable-resolution ice sheet model for Earth system modeling using Voronoi grids
journal, January 2018

  • Hoffman, Matthew J.; Perego, Mauro; Price, Stephen F.
  • Geoscientific Model Development, Vol. 11, Issue 9
  • DOI: 10.5194/gmd-11-3747-2018

OpenAD/F: A Modular Open-Source Tool for Automatic Differentiation of Fortran Codes
journal, July 2008

  • Utke, Jean; Naumann, Uwe; Fagan, Mike
  • ACM Transactions on Mathematical Software, Vol. 34, Issue 4
  • DOI: 10.1145/1377596.1377598

Parameterization of the Annual Surface Temperature and Mass Balance of Antarctica
journal, January 1990

An enthalpy formulation for glaciers and ice sheets
journal, January 2012

  • Aschwanden, Andy; Bueler, Ed; Khroulev, Constantine
  • Journal of Glaciology, Vol. 58, Issue 209
  • DOI: 10.3189/2012JoG11J088

The Theory of Glacier Sliding
journal, January 1964

Parameterization of the Annual Surface Temperature and Mass Balance of Antarctica
journal, January 1990

An improved Antarctic dataset for high resolution numerical ice sheet models (ALBMAP v1)
journal, January 2010

  • Le Brocq, A. M.; Payne, A. J.; Vieli, A.
  • Earth System Science Data Discussions, Vol. 3, Issue 1
  • DOI: 10.5194/essdd-3-195-2010

Greenland Ice Sheet contribution to sea-level rise from a new-generation ice-sheet model
journal, January 2012

  • Gillet-Chaulet, F.; Gagliardini, O.; Seddik, H.
  • The Cryosphere Discussions, Vol. 6, Issue 4
  • DOI: 10.5194/tcd-6-2789-2012

Application of control methods for modelling the flow of Pine Island Glacier, West Antarctica
text, January 2003

  • Vieli, Andreas; Payne, Antony J.
  • International Glaciological Society
  • DOI: 10.5167/uzh-136283

An improved Antarctic dataset for high resolution numerical ice sheet models (ALBMAP v1)
text, January 2010

  • Le Brocq, A. M.; Payne, Antony J.; Vieli, Andreas
  • Copernicus Publications
  • DOI: 10.5167/uzh-136285

Paleothermometry by control methods
journal, January 1991

  • MacAyeal, Douglas R.; Firestone, John; Waddington, Edwin
  • Journal of Glaciology, Vol. 37, Issue 127
  • DOI: 10.1017/s0022143000005761

A Three-Dimensional Time-Dependent Model of the West Antarctic Ice Sheet
journal, January 1984

Larsen B Ice Shelf rheology preceding its disintegration inferred by a control method
journal, January 2007

  • Khazendar, A.; Rignot, E.; Larour, E.
  • Geophysical Research Letters, Vol. 34, Issue 19
  • DOI: 10.1029/2007gl030980

Programming language features, usage patterns, and the efficiency of generated adjoint code
journal, February 2016

Algorithm 799: revolve: an implementation of checkpointing for the reverse or adjoint mode of computational differentiation
journal, March 2000

  • Griewank, Andreas; Walther, Andrea
  • ACM Transactions on Mathematical Software, Vol. 26, Issue 1
  • DOI: 10.1145/347837.347846

Sensitivity Analysis Using an Adjoint of the PSU-NCAR Mesoseale Model
journal, August 1992

Application of control methods for modelling the flow of Pine Island Glacier, West Antarctica
journal, January 2003

Greenland ice-sheet volume sensitivity to basal, surface and initial conditions derived from an adjoint model
journal, January 2009

Comparison of adjoint and nudging methods to initialise ice sheet model basal conditions
journal, January 2016

  • Mosbeux, Cyrille; Gillet-Chaulet, Fabien; Gagliardini, Olivier
  • Geoscientific Model Development, Vol. 9, Issue 7
  • DOI: 10.5194/gmd-9-2549-2016
    Sort by:
    [ × clear filter / sort ]

    Figures / Tables found in this record:

    Figure 1 (p. 5)figure
    Table 1 (p. 8)table
    Figure 2 (p. 9)figure
    Figure 3 (p. 10)figure
    Figure A1 (p. 14)figure
    Table A1 (p. 14)table
      [ × clear filter / sort ]
      Figures/Tables have been extracted from DOE-funded journal article accepted manuscripts.

      Similar Records in DOE PAGES and OSTI.GOV collections:

      Other research related to this record: