One needs to read all the documentation on the mission, the levels of processing, and about the derivative datasets which incorporate various corrections from other sources. That now spans decades. The 'raw' DEMs are created from stereo views, so have massive errors like peaks and drop outs and are very noisy. Also, ASTER DEM are generally a DSM ( Digital Surface Model ), not bare earth ( DEM ).

Dear Wms Priyankara,

It is worth mentioning that ALOS 12.5m DEM data is currently mentioned on the Internet, and some are called ALOS PALSAR 12.5m DEM data, which can be downloaded from NASA ASF (Alaska Satellite Facility) ASF Data Search https://search.asf.alaska.edu/#/ . But in fact, the ALOS satellite DEM data is produced by the stereoscopic images captured by PRISM, and the onboard PALSAR (Phased Array type L-band Synthetic Aperture Radar) is mainly used to obtain orthophotos. The DEM resolution in the ALOS PALSAR Hi-Res Terrain Corrected data provided by NASA ASF is indeed 12.5m, but this DEM data is actually used to perform terrain correction on the original PALSAR data. In the ASF radiometrically terrain corrected ALOS PALSAR products product guide The source of DEM data used for terrain correction is clearly pointed out in the figure below.

You can find more at https://asf.alaska.edu/data-sets/derived-data-sets/alos-palsar-rtc/alos-palsar-radiometric-terrain-correction/

Hope this can help anyway,

Kind Regards

I strongly agree with Zhouxuan Xiao , It is important to understand the properties of each data by going through their documentations. More importantly, there is a need to validate their results with ground control points, which are useful for validation and correction of the image outputs.