Change detection in SNAP

I have downloaded two images of before disaster and after a disaster cyclone images keeping threshold as +2 and +1 in change detection option in SNAP
Performed following steps

  1. calibrate
    2.speckle filter
    3)Terrain Correction
    4)change to db
    5)stacking of image
    the ratio output shown in figure does not show many pixels.
    Kindly Suggest how to select threshold so that a disaster map could be drawn or its roughly based on trial?.Thanks


To my experience, there is no universal threshold which magically separates the chances from the constant surface, mostly because the different land cover types differently respond to change.

A decrease of 2 db might well represent the transition from non-flooded to flooded in urban areas, but falsely erestimate flooding on agriculture.

One way to solve this is to visually check and compare the image statistics at a location of known change, for example with the pin manager, with profile plots and histograms of digitized ROIs. Your then select the threshold based on these findings.
A nice demonstration is given here:

Also in this tutorial, the mentioned analysis tools are explained: Synergetic use of S1 (SAR) and S2 (optical) data and use of analysis tools

ok @ABraun Thanks for the links.