Skip to main content

Advertisement

Log in

Inverse problem for a nonlinear parabolic equation with nonsmooth periodic coefficients

  • Published:
SeMA Journal Aims and scope Submit manuscript

Abstract

We prove a result of stability for an inverse problem associated to a parabolic nonlinear equation by using a Carleman inequality. To do this, existence, uniqueness, regularity results and maximum principle for our problem are established. This stability inequality concerns the reconstruction of an \(L^\infty ({\mathbb R}^n)\) coefficient, basing ourselves by taking an arbitrary open set of observation.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

References

  1. Berestycki, H., Hamel, F., Roques, L.: Analysis of the periodically fragmented environment model: 1 Influence of periodic heterogeneous environment on species persistence. J. Math. Biol. 51(1), 75 (2005)

    Article  MATH  MathSciNet  Google Scholar 

  2. Bukhgeim, A.L., Klibanov, M.V.: Uniqueness in the large of a class of multidimensional inverse problems. Sov. Math. Dokl. 24, 244–247 (1981)

    Google Scholar 

  3. Choi, J.: Inverse problem for a parabolic equation with space-periodic boundary conditions by a Carleman estimate. Inverse ILL Posed Probl. 11(2), 111–135 (2003)

    Article  MATH  MathSciNet  Google Scholar 

  4. Cristofol, M., Roques, L.: Biological invasions: deriving the regions at risk from partial measurements. Math. Biosci. 215(2), 158–166 (2008)

    Article  MATH  MathSciNet  Google Scholar 

  5. Cristofol, M., Roques, L.: An inverse problem involving two coefficients in a nonlinear reaction–diffusion equation. C. R. Math. 350(9–10), 469–473 (2012)

    Article  MATH  MathSciNet  Google Scholar 

  6. Dautray, R., Lions, J.L.: Mathematical Analysis and Numerical Methods for Science and Technology, vol. 3. Masson, Paris (1984)

  7. Engländer, J., Pinsky, R.G.: Uniqueness/nonuniqueness for nonegative solutions of second-order parabolic equations of the form \(u_t=Lu+Vu-\gamma u^p\) in \(\mathbb{R}^n\). J. Differ. Equ. 192, 396–428 (2003)

    Article  MATH  Google Scholar 

  8. Evans, C.: Partial Differential Equations. Graduate Studies in Mathematics, vol. 19, AMS, Providence (1997)

  9. Fisher, R.A.: The wave of advance of avantageous genes. Ann. Eugen. 7, 335–369 (1937)

    Google Scholar 

  10. Fursikov, A.V., Imanuvilov, O.Y.: Controllability of Evolution Equations. Lecture Notes Series, vol. 34. Seoul National University, Korea (1996)

  11. Immanuvilov, O.Y., Yamamoto, M.: Lipschitz stability in inverse parabolic problems by the Carleman estimate. Inverse Probl. 14, 1229–1245 (1998)

    Article  Google Scholar 

  12. Kolmogorov, A.N., Petrovsky, I.G., Piskunov, N.S.: Etude de l’quation de la diffusion avec croissance de la quantit de matire et son application un problme Biologique, Bull, Univ. Etat Moscou, Sr. Internationale.A, 1, pp. 1–26 (1937)

  13. Ladyzenskaja, O.A., Solonnikov, V.A., Ural’ceva, N.N.: Linear and Quasi-linear Equations on Parabolic. American Mathematical Society, Providence (1968)

  14. Lieberman, G.M.: Second Order Parabolic Differential Operators. World Scientific Publishing, Singapore (1996)

    Book  Google Scholar 

  15. Murray, J.D.: Interdisciplinary applied mathematics. In: Mathematical Biology, 3rd edn, vol. 17. Springer, New York (2002)

  16. Prilepko, A.I., Orlovsky, D.G., Vasin, I.A.: Methods for solving inverse problems in mathematical physics. In: Dekker, M. (ed.) Pure and Applied Mathematics. Marcel Dekker, Inc., New York, Basel (2000)

  17. Shigesada, N., Kawasaki, K.: Biological Invasions: Theory ans Practice. Oxford Series in Ecology and Evolution. Oxford University Press, Oxford (1997)

    Google Scholar 

  18. Skellam, J.G.: Random dispersal in theoretical populations. Biometrika 38, 196–218 (1951)

    Article  MATH  MathSciNet  Google Scholar 

  19. Turchin, P.: Quantitative Analysis of Movement: Measuring and Modeling Population Re-distribution in Animals and Plants. Sinauer Associates, Sunderland (1998)

    Google Scholar 

  20. Yamamoto, M.: Carleman estimates for parabolic equations and applications. Inverse Probl. 10, 1–75 (2009)

Download references

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to D. E. Teniou.

Additional information

This work was partially supported by the franco-algerian agreement Tassili 11 MDU 834.

Appendix

Appendix

Much of the demonstration of Theorem 2 is classic, we will give some indications.

We introduce the following weight functions:

$$\begin{aligned} \varphi (t, x) = \frac{ e^{\lambda \psi (x)}}{t(T-t)},\; \alpha (t, x) = \frac{e^{\lambda \psi } -e^{2\lambda ||\psi ||_{\infty }}}{t(T-t)}, \end{aligned}$$
(5.1)

where \(\lambda >0\); remark that \(\alpha <0\). We establish a Carleman estimate by taking into account the fact that the weight function \(\psi \) is \(L\)-periodic. Set \(\displaystyle w=e^{s\alpha }z.\) We have

$$\begin{aligned} L_1w+L_2w=f_s \; \text{ on }\; Q, \end{aligned}$$

with

$$\begin{aligned} L_1 w=-\Delta w-s^2\lambda ^2\varphi ^2|\nabla \psi |^2w-s\alpha _t w, \end{aligned}$$
(5.2)
$$\begin{aligned} L_2w= w_t+2s\lambda \varphi \nabla \psi .\nabla w+2s\lambda ^2\varphi w |\nabla \psi |^2, \end{aligned}$$
(5.3)

and

$$\begin{aligned} f_s=fe^{2s\alpha }-s\lambda \varphi w\Delta \psi +s\lambda ^2\varphi w |\nabla \psi |^2. \end{aligned}$$
(5.4)

So

$$\begin{aligned} \Vert f_s\Vert ^2_{L^2(Q)}= \Vert L_1w\Vert ^2_{L^2(Q)} + \Vert L_2w\Vert ^2_{L^2(Q)} +2(L_1w,L_2w)_{L_2(Q)}. \end{aligned}$$
(5.5)

We estimate the term \((L_1w,L_2w)_{L^2(Q)}\) which will be written in the following way

$$\begin{aligned} (L_1w,L_2w)_{L^2(Q)}=A_1+A_2+A_3, \end{aligned}$$
(5.6)

where

$$\begin{aligned} A_1&= -\int _Q\Big (\Delta w+s^2\lambda ^2\varphi ^2 |\nabla \psi |^2 w+s \alpha _t w\Big ) \Big ( w_t+2s\lambda ^2\varphi w |\nabla \psi |^2\Big )dx dt, \nonumber \\ A_2&= -\int _Q\Big (s^2\lambda ^2\varphi ^2|\nabla \psi |^2w+s\alpha _t w\Big )\Big (2s\lambda \varphi \nabla \psi .\nabla w\Big )dx dt, \nonumber \\ A_3&= -\int _Q\Big (\Delta w\Big ) \Big (2s\lambda \varphi \nabla \psi .\nabla w\Big )dx dt. \end{aligned}$$
(5.7)

Thanks to \(L\)-periodicity on \(w, \varphi \) and \(\psi \) and the fact that \(w(0,.)=w(T,.)=0\), integration by parts gives

$$\begin{aligned} A_1&= \int _Q\Big \{\frac{s^2 \lambda ^2}{2}( \varphi ^2|\nabla \psi |^2)_tw^2 -\frac{s}{2}\alpha _{tt}w^2+2s \lambda ^2\varphi |\nabla \psi |^2|\nabla w|^2\\&\quad +\,2s\lambda ^2 w \nabla w.\nabla (\varphi |\nabla \psi |^2) -2s^3\lambda ^4\varphi ^3 w^2|\nabla \psi |^4-2s^2\lambda ^2\varphi \alpha _t w^2|\nabla \psi |^2\Big \}dx dt.\\ A_2&= \int _Q\Big \{3s^3\lambda ^4\varphi ^3|\nabla \psi |^4w^2+s^3\lambda ^3\varphi ^3w^2\nabla . \Big (|\nabla \psi |^2 \nabla \psi \Big ) \nonumber \\&\quad \quad +\,s^2\lambda w^2 \nabla . \Big (\varphi \alpha _t (\nabla \psi )\Big ) \Big \}dx dt.\\ A_3&= \int _Q\Big \{2s\lambda ^2\varphi \Big (\Big (\nabla \psi .\nabla w\Big )^2+2s\lambda \varphi \Big (\nabla w. (\nabla \nabla \psi ).\nabla w\Big ) \Big ) \nonumber \\&\quad \quad -\,s\lambda ^2\varphi |\nabla \psi |^2|\nabla w|^2-s\lambda \varphi |\nabla w|^2\Delta \psi \Big \}dx dt. \end{aligned}$$

we get

$$\begin{aligned} (L_1w,L_2w)_{L^2(Q)}&= \int _Q \left\{ s^3\lambda ^4\varphi ^3|\nabla \psi |^4 w^2+s\lambda ^2\varphi |\nabla \psi |^2|\nabla w|^2\right. \nonumber \\&\quad \quad \left. +2s\lambda ^2\varphi (\nabla \psi .\nabla w)^2\right\} dx dt+X_1, \end{aligned}$$
(5.8)

where \(X_1\) is a term which can be absorbed.

Now inserting (5.8) in (5.5), we obtain

$$\begin{aligned} ||f_s||_{L^2(Q)}^2&= ||L_1w||_{L^2(Q)}^2+||L_2w||_{L^2(Q)}^2\nonumber \\&\quad +2\int _Q \{s^3\lambda ^4\varphi ^3|\nabla \psi |^4 w^2+\,s\lambda ^2\varphi |\nabla \psi |^2|\nabla w|^2\nonumber \\&\quad +2s\lambda ^2\varphi (\nabla \psi .\nabla w)^2\} dx dt+2X_1. \end{aligned}$$
(5.9)

Taking into account (5.4), we see that

$$\begin{aligned} ||f_s||_{L^2(Q)}^2\le 3\left( ||fe^{s\alpha }||_{L^2(Q)}^2+||s\lambda \varphi w\Delta \psi ||_{L^2(Q)}^2+ ||s\lambda ^2\varphi w |\nabla \psi |^2||_{L^2(Q)}^2\right) . \end{aligned}$$
(5.10)

From the Proposition 2, there exists \(\delta \) such that \(0<\delta \le |\nabla \psi |^2, x\in \overline{\Omega \backslash \omega }\). We get

$$\begin{aligned}&||L_1w||_{L^2(Q)}^2+||L_2w||_{L^2(Q)}^2+\int _Q\left( s^3\lambda ^4\varphi ^3w^2+s\lambda ^2\varphi |\nabla w|^2\right) dx dt \nonumber \\&\quad \le C\left( \int _{(0,T)\times \omega }\left( s^3\lambda ^4\varphi ^3w^2+s\lambda ^2\varphi |\nabla w|^2\right) dx dt +||{f}e^{s\alpha }||_{L^2(Q)}^2\right) . \end{aligned}$$
(5.11)

It remains to estimate the terms \(w_t\) and \(\Delta w\). We use the following relations:

$$\begin{aligned} \Delta w&= -L_1 w+s^2\lambda ^2\varphi ^2|\nabla \psi |^2w+s\alpha _t w, \\ w_t&= L_2w-2s\lambda \varphi \nabla \psi .\nabla w-2s\lambda ^2\varphi w |\nabla \psi |^2. \end{aligned}$$

To obtain the Carleman inequality (1.2) in \(z\), we use the fact that \(w=e^{s\alpha }z\) in \(Q\). \(\square \)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Kaddouri, I., Teniou, D.E. Inverse problem for a nonlinear parabolic equation with nonsmooth periodic coefficients. SeMA 66, 55–69 (2014). https://doi.org/10.1007/s40324-014-0024-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s40324-014-0024-7

Keywords

Mathematics Subject Classification

Navigation