1: \begin{abstract}

2: Data remap between non-matching meshes is a critical step in multiphysics

3: coupling using a partitioned approach. The data fields being transferred

4: often have jumps in function values or derivatives. It is important

5: but very challenging to avoid spurious oscillations (a.k.a. the \emph{Gibbs

6: Phenomenon}) near discontinuities and at the same time to achieve

7: high-order accuracy away from discontinuities. In this work, we introduce

8: a new approach, called \emph{WLS-ENOR}, or \emph{Weighted-Least-Squares-based

9: Essentially Non-Oscillatory Remap}, to address this challenge. Based

10: on the WLS-ENO reconstruction technique proposed by Liu and Jiao (\emph{J.

11: Comput. Phys.} vol 314, pp 749-{}-773, 2016),\emph{ WLS-ENOR} differs

12: from WLS-ENO and other WENO schemes in that it resolves not only the

13: $\mathcal{O}(1)$ oscillations due to $C^{0}$ discontinuities, but

14: also the accumulated effect of $\mathcal{O}(h)$ oscillations due

15: to $C^{1}$ discontinuities. To this end, WLS-ENOR introduces a robust

16: detector of discontinuities and a new weighting scheme for WLS-ENO

17: near discontinuities. We also optimize the weights at smooth regions

18: to achieve superconvergence. As a result, WLS-ENOR is more than fifth-order

19: accurate and highly conservative in smooth regions, while being non-oscillatory

20: and minimally diffusive near discontinuities. We also compare WLS-ENOR

21: with some commonly used methods based on $L^{2}$ projection, moving

22: least squares, and radial basis functions.

23: \end{abstract}

24: