Using Adaptive Mesh Refinement to Model Ocean...

18
Using Adaptive Mesh Refinement to Model Ocean Flows M.J. Berger, D.L. George, R.J. LeVeque, and K.T. Mandli Department of Applied Mathematics University of Washington Seattle, WA, USA IMUM, Boston, 2010-08-17 Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 1 / 18

Transcript of Using Adaptive Mesh Refinement to Model Ocean...

Page 1: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Using Adaptive Mesh Refinement to Model Ocean Flows

M.J. Berger, D.L. George, R.J. LeVeque, and K.T. Mandli

Department of Applied MathematicsUniversity of Washington

Seattle, WA, USA

IMUM, Boston, 2010-08-17

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 1 / 18

Page 2: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Outline

1 Numerical Method

2 ExamplesCircular OceanChile TsunamiStorm Surge

3 Future Work

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 2 / 18

Page 3: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Goals

Solve depth averaged equations

ht + (hu)x + (hv)y = 0

(hu)t +

(hu2 +

1

2gh2

)x

+ (huv)y = −ghbx −h

ρ(PA)x +

1

ρ[τsx + τbx ]

(hu)t + (huv)x +

(hv2 +

1

2gh2

)y

= −ghby −h

ρ(PA)y +

1

ρ[τsy + τby ]

Resolve ocean to inundation spatial and temporal scales

Handle arbitrary bathymetry and inundation

Determine grids based on both physical and numerical errorestimation

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 3 / 18

Page 4: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Numerical Method

Wave-propagation algorithm, class of high-resolution finite volumemethods using a Godunov type scheme

Evolve grid cell averages by determining waves at edges of grid cellsRiemann solution preserves non-trivial steady states and handlesdry-states and inundation

Adaptive mesh refinement:

Structured, patch based scheme with properly nested gridsTime steps chosen to satisfy CFL condition on current gridConservative fix-ups required along boundaries using ghost cells

GeoClaw

Download at: http://www.clawpack.org/Open source research code incorporating:

Solvers for shallow water systems using finite volume methodsAdaptive mesh refinement for complex topographyDry-states and inundation handlingTools for handling multiple topography specifications

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 4 / 18

Page 5: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Circular Ocean

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 5 / 18

Page 6: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Circular Ocean

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 6 / 18

Page 7: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Circular Ocean

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 7 / 18

Page 8: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Circular Ocean

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 8 / 18

Page 9: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Chile Tsunami

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 9 / 18

Page 10: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Chile Tsunami

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 10 / 18

Page 11: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Chile Tsunami

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 11 / 18

Page 12: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Chile Tsunami

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 12 / 18

Page 13: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Storm Surge Preliminary Results

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 13 / 18

Page 14: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Storm Surge Preliminary Results

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 14 / 18

Page 15: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Future Work

Multilayer shallow water equations:

∂t(ρ1h1) +

∂x(ρ1h1u1) = 0,

∂t(ρ1h1u1) +

∂x

(ρ1h1u

21 + gρ1

h21

2+ gρ2h1h2

)= ρ2gh2

∂h1

∂x− ρ1gh1b

′(x)

∂t(ρ2h2) +

∂x(ρ2h2u2) = 0,

∂t(ρ2h2u2) +

∂x

(ρ2h2u

22 + gρ2

h22

2

)= −ρ2gh2

∂h1

∂x− ρ2gh2b

′(x)

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 15 / 18

Page 16: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Future Work

Multilayer shallow water equations:

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 16 / 18

Page 17: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Future Work

Multilayer shallow water equations:

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 17 / 18

Page 18: Using Adaptive Mesh Refinement to Model Ocean …mseas.mit.edu/IMUM2010/Presentations/presentation_mandli.pdfUsing Adaptive Mesh Re nement to Model Ocean Flows M.J. Berger, D.L. George,

Special Thanks To:

Mandli (UW Applied Math) AMR applied to ocean flows IMUM 2010, 2010-08-17 18 / 18