Modeling still matters: a surprising instance of catastrophic floating point errors in mathematical biology and numerical methods for ODEs

We guide the reader on a journey through mathematical modeling and numerical analysis, emphasizing the crucial interplay of both disciplines. Targeting undergraduate students with basic knowledge in dynamical systems and numerical methods for ordinary differential equations, we explore a model from mathematical biology where numerical methods fail badly due to catastrophic floating point errors. We analyze the reasons for this behavior by studying the steady states of the model and use the theory of invariants to develop an alternative model that is suited for numerical simulations. Our story intends to motivate combining analytical and numerical knowledge, even in cases where the world looks fine at first sight. We have set up an online repository containing an interactive notebook with all numerical experiments to make this study fully reproducible and useful for classroom teaching.

[1]  P. Birken,et al.  Resolving entropy growth from iterative methods , 2023, BIT Numerical Mathematics.

[2]  Benjamin K Tapley,et al.  On the preservation of second integrals by Runge-Kutta methods , 2021, Journal of Computational Dynamics.

[3]  P. Birken,et al.  Conservation Properties of Iterative Methods for Implicit Discretizations of Conservation Laws , 2022, Journal of Scientific Computing.

[4]  P. Birken,et al.  Locally conservative and flux consistent iterative methods , 2022, SIAM Journal on Scientific Computing.

[5]  D. Ketcheson,et al.  A Broad Class of Conservative Numerical Methods for Dispersive Wave Equations , 2020, Communications in Computational Physics.

[6]  Lisandro Dalcin,et al.  Relaxation Runge-Kutta Methods: Fully Discrete Explicit Entropy-Stable Schemes for the Compressible Euler and Navier-Stokes Equations , 2019, SIAM J. Sci. Comput..

[7]  Benjamin Allen,et al.  A mathematical formalism for natural selection with arbitrary spatial and genetic structure , 2018, Journal of Mathematical Biology.

[8]  Herbert Egger,et al.  Structure preserving approximation of dissipative evolution problems , 2018, Numerische Mathematik.

[9]  Otto Richter,et al.  Multi-gene-loci inheritance in resistance modeling. , 2013, Mathematical biosciences.

[10]  Charalampos Tsitouras,et al.  Runge-Kutta pairs of order 5(4) satisfying only the first column simplifying assumption , 2011, Comput. Math. Appl..

[11]  James H. Verner,et al.  Numerically optimal Runge–Kutta pairs with interpolants , 2010, Numerical Algorithms.

[12]  E. Hairer,et al.  Geometric Numerical Integration: Structure Preserving Algorithms for Ordinary Differential Equations , 2004 .

[13]  Joseph D. Skufca,et al.  Analysis Still Matters: A Surprising Instance of Failure of Runge-Kutta-Felberg ODE Solvers , 2004, SIAM Rev..

[14]  P. Maini.  Essential Mathematical Biology , 2003 .

[15]  Eitan Tadmor,et al.  Entropy stability theory for difference approximations of nonlinear conservation laws and related time-dependent problems , 2003, Acta Numerica.

[16]  A. Goriely.  Integrability and Nonintegrability of Dynamical Systems , 2001 .

[17]  G. Iooss,et al.  Topics in bifurcation theory and applications , 1999 .

[18]  Lawrence F. Shampine,et al.  The MATLAB ODE Suite , 1997, SIAM J. Sci. Comput..

[19]  R. Seydel.  Practical Bifurcation and Stability Analysis , 1994 .

[20]  L. Perko.  Differential Equations and Dynamical Systems , 1991 .

[21]  E. Haug,et al.  Generalized Coordinate Partitioning for Dimension Reduction in Analysis of Constrained Dynamic Systems , 1982 .

[22]  J. Dormand,et al.  A family of embedded Runge-Kutta formulae , 1980 .

[23]  G. Sell,et al.  The Hopf Bifurcation and Its Applications , 1976 .

[24]  Neil Fenichel.  Persistence and Smoothness of Invariant Manifolds for Flows , 1971 .