Some questions about ANC interpolation

[Lyusijie]

Hello Dr. Foster
I am a Ph.D. student from Shanghai Jiao Tong University, China. Recently, I read your paper 'An Evaluation of Interpolation Techniques for Reconstructing Ionospheric TEC Maps', IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, VOL. 46, NO. 7, JULY 2008. Thanks for your excellent study and Matlab toolkit open source.

I have some questions as follows and hope for your help:

1. I run your code and test the interpolation performance compare to the paper. In your paper, the rmse of biharmonic spline for simulated data (Fig.5&Fig.6) is the largest. However, the rmse of ANC is usually the largest, ANC>BSI>rbf=kringing. I get the biharmonic spline interpolation result from griddata(xi, yi, zi, xx, yy,'v4'). I wonder the result in the paper is a universal phenomenon or an occasional phenomenon?

2)It has been mentioned in the paper that the function of Cubic interpolation is available in MATLAB 2007a's griddata. I tried this function and found that the interpolation result is nan at the boundary. It seems that this function cannot interpolate boundary values. So how to assess the rmse of boundary values?