InSAR coherence problem using ERS data


I am using two paris of ERS2 SLC stripmap images in order to get one coherence image for each pair.
Pair 1 : Bperp = 273m and Btemp=70 days.
Pair 2 : Bperp = 81.9m and Btemp = 140 days.

When I process the InSAR coherence chain (Apply orbit file => Corregistration => InSAR coherence), I get a cohrence image that is not clear (displayed in the image bellow). Can I consider it as a correct coherence image that contains information even if it is noisy and not very clear?
I even tried different window sizes for the coherence calculation but I still get the same output.
Is there something (filtering, multilooking…) that I can do to get a better coherence image?
Thank you.

try reducing the Coherence Range Window Size, maybe to 5 (instead of the default 10). This generally leads to smaller coherent areas but their pattern might be less blurred.

Thank you for the reply,
I reduced the Range Window Size and I am not getting a remarkable improvement. Do you think that this can be related to an issue with the orbit file?
Do you suggest doing multilooking or filtering before or after processing the coherence product?


I’m afraid nothing will give you better coherence as the low values are caused by temporal decorrelation. 70 and 140 area way too long for a C-band radar because there are too much scatterers on the ground changing during that time.
Is choosing an image pair with smaller temporal baseline an option?

Unfortunately, I can’t find other ERS or ENVISAT images for the same date in 2003 and with the same track.