Sentinel-1 dual polarized layerstack or combined with landsat 8 images

Hello to all friends and experts

I have seen in some crop mapping articles (rice maping with phenology,…) that layerstacked Sentinel-1 time series and Landsat 8 NDVI time series images and use them in classification.
Now I have a few questions to ask:

1- How is this layerstacking of NDVI time series and Sentinel-1 time series images possible?

2- Do VH / VV images really mean the ratio of VH to VV that should be written and executed in SNAP Band math? How to calculate this amount? Should it be calculated before converting to decibels or after?

3- To stacking Sentinel 1 images with Landsat 8 NDVI in ArcGIS, do we have to change the spatial resolution of Sentinel-1 images? For example, from 10 to 30 meters.
Or, for example, with the same spatial resolution of 10 meters Sentinel-1 and 30 meters Landsat 8, it is possible to map crops such as rice with the help of classification algorithms such as random forest?

I need your guidance, dear experts

You can create stacks of products with the same coordinate reference system using the Collocation module. As the Sentinel-1 geo-accuracy is higher than the one of optical sensors, you should select it as the reference. This operator performs a resampling if the other products have a different resolution, so you don’t need to do that in advance.

Calculating the ratio between VH and VV is a good way to determine the share of cross-polarization per pixel (normalized against the general backscatter intensity). I would to it before conversion to dB.

Some of your questions are also addressed here: Synergetic use of S1 (SAR) and S2 (optical) data and use of analysis tools, maybe this clarifies some points. It makes suggestions on further analysis as well.