Clawpack: building an open source ecosystem for solving hyperbolic PDEs

Abstract

Clawpack is a software package designed to solve nonlinear hyperbolic partial differential equations using high-resolution finite volume methods based on Riemann solvers and limiters. The package includes a number of variants aimed at different applications and user communities. Clawpack has been actively developed as an open source project for over 20 years. The latest major release, Clawpack 5, introduces a number of new features and changes to the code base and a new development model based on GitHub and Git submodules. This article provides a summary of the most significant changes, the rationale behind some of these changes, and a description of our current development model.

Interseismic Coupling‐Based Earthquake and Tsunami Scenarios for the Nankai Trough

“…The Nankai Trough is currently coupled along its entire length (Figure 1b), so we scale slip to generate an event comparable to the most recent full-margin Nankai rupture in 1707. We perform tsunami inundation simulations using version 5.3.0 of the open source tsunami model GeoClaw Clawpack Development Team, 2015;Mandli et al, 2016), which is a subset of Clawpack. We perform tsunami inundation simulations using version 5.3.0 of the open source tsunami model GeoClaw Clawpack Development Team, 2015;Mandli et al, 2016), which is a subset of Clawpack.…”

“…There are no measurements of the 1707 earthquake's magnitude, but historical records document the resultant tsunami's height (Hatori, 1974(Hatori, , 1985Murakami et al, 1995); thus, we simulate tsunami inundation with a range of peak slip magnitudes for each of the four coupling-based slip distributions and choose the value that produces the best match between modeled and observed 1707 tsunami heights. We perform tsunami inundation simulations using version 5.3.0 of the open source tsunami model GeoClaw Clawpack Development Team, 2015;Mandli et al, 2016), which is a subset of Clawpack. Earthquake sources for tsunami simulations, GeoClaw model parameters, development of a topographic model, and methodology for comparing modeled and historical tsunami inundation heights are all described in detail in supporting information Texts S2-S4 and Figure S1.…”

Accelerating an Adaptive Mesh Refinement Code for Depth‐Averaged Flows Using GPUs

“…The current implementation contributes to the Clawpack ecosystem (Mandli et al, ), which uses an AoS layout since Fortran arrays are dimensioned so that q ( m , i , j ) is the m th component (depth or momenta) in the ( i , j ) grid cell. However, many applications within the Clawpack ecosystem will be affected if the AoS data layout is changed.…”

New methodology for computing tsunami generation by subaerial landslides: Application to the 2015 Tyndall Glacier landslide, Alaska

“…In the limit of vanishing solids fraction, D-Claw's model equations reduce to the nonlinear shallow-water equations that are commonly used for tsunami modeling. In fact, D-Claw provides a generalization and extension of the tsunami modeling software GeoClaw [George, 2006;Berger et al, 2011;LeVeque et al, 2011;Mandli et al, 2016], and it reproduces GeoClaw solutions if solids are absent. Like GeoClaw, D-Claw employs adaptive mesh refinement (AMR), finite-volume shock-capturing algorithms [LeVeque, 2002], inundation-front resolution, and well-balanced preservation of static and flowing steady states [George, 2008[George, , 2011George and Iverson, 2014].…”