Please check the following links and PDF attachment.

http://link.springer.com/article/10.1007/s12665-013-2383-2

http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6705048&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D6705048

Good luck

Ok, for most of glaicers, X, C, L has no obviously IFG, for the glacier movement will cause decorrelation. Anyhow, you may got IFG if the glaicer moves very slowly. Or otherwise you have to use pixel track method. This is happen in different part of one glaicer as well.

So, you need some knowledge of the velocity of the glaicer and you need to compute the movement of the glaicer during the revisiting time, and at last you need to compare this values with wavelength. Like L band PALSAR, LOS displacement during time span need to small than 23.6 cm/4 (Colesanti, 2006, P181), then you got IFG, or you need to use Pixel track.

Conclusion, L C X , is one option, the others things, glaicer velocity, revisiting time.

Thank you Arvind an Liu for the response. The information is very useful.

Regards!

You also have to consider the penetration depth of L-band in dry snow. Jurgen Dall at the Danish Technical University has shown that C-band scattering centre heights on Greenland may be as much as 13 m below the glacier surface. This means for L-band, where the penetration depth is theoretically much greater, that you are not mapping the surface but somewhere at or well below the snow-firn transition.