At successful exit the **process control flag** is switched on:

subtrrefpha: 1

The output (in the **products result file**) looks like:

******************************************************************* *_Start_subtrrefpha: ******************************************************************* Data_output_file: Outdata/cint.minrefpha.raw Data_output_format: complex_real4 First_line (w.r.t. original_master): 245 Last_line (w.r.t. original_master): 14964 First_pixel (w.r.t. original_master): 7 Last_pixel (w.r.t. original_master): 3998 Multilookfactor_azimuth_direction: 40 Multilookfactor_range_direction: 8 Number of lines (multilooked): 368 Number of pixels (multilooked): 499 ******************************************************************* * End_subtr_refphase:_NORMAL *******************************************************************

We noticed that if the precise orbits are not long enough (not enough time before and after first/last line), this results in a wrong reference phase for obvious reasons. Interpolation near the end of the data points is not very good with cubic splines. This can be solved by using more orbital data points after the last line of the scene (see cards for step M_PORBITS).

Figure 25.1 shows the result of subtracting the reference phase polynomial from the interferogram (Figure 23.1). (The same scene as described in section 23.0.4, again multilooked, now by factors 4 and 4, resulting in total multilooking of 40 and 8, which agrees on the terrain with a resolution of about 160 meters square.)

This step can be mis-used to correct for residual orbital fringes (if you know what you are doing).

To do this, first count the number of fringes you want to remove from
the interferogram. Then, edit the **products result file** and create a section for
the step COMPREFPHA. In this output section, simply define a
polynomial that describes for example a linear trend in range of, say,
2.5 fringes. Then, run this step, and doris will not know that it is
not the reference phase polynomial that is subtracted from the
interferogram, but an additional correction polynomial that has been
inserted by hand.