Foot-only routes in public-transport isochrones

Hello,

I need some advice concerning pt-isochrones.

Problem: We want to calculate pt-isochrones from multiple points in Germany at nighttime (for example 30-minutes-isochrones) to investigate how accessible some points of interest are using public transportation.
We set up a custom graphhopper server including GTFS-data for calculating routes and isochrones per public transport. This works well so far.
However, now we need some help interpreting the results, concerning the following:
At night, there is often no public transport available, especially in rural areas. If we understand correctly, Graphhopper will suggest a walking route if no public transport is available during the specified time period.
However, when looking at individual routes, we have noticed that some routes are only-foot routes, even though public transport would theoretically be available. Does Graphhopper automatically chooses the shorter route here?

In short, this leads to these two questions:

  1. Is it true, that the resulting pt-isochrones include pt-routes as well as only-foot routes? Is there a possibility to only include public transport routes?
  2. When does Graphhopper use an only-foot route instead of a public transport route to calculate the isochrone? Are there fixed thresholds or parameters that determine the decision which mean of transport is used?

We appreciate any information or hints.
Thank you!

One follow-up question:
We have analysed our data again in more detail. In doing so, we have found that the isochrones of the routes on foot are sometimes larger than the pt-isochrones. This is particularly the case when shorter time intervals are selected, e.g. 15 minutes. This contradicts the assumption that Graphhopper always selects the shorter route on foot or by pt. Does anyone have an idea how this can be explained?