Unexpected Isochrone Behavior in Multimodal Mode with Docker Valhalla

[zi-thi]

Hello,
While using Docker Valhalla to generate 15min, 30min, 45min, and 60min isochrones in multimodal mode, I noticed an unusual behavior: some isochrone polygons are significantly distant from the starting point (as shown in the images below).

Images:
Both starting points are located in Bavaria, Germany, and the transit data has been properly configured. In the first image, you can see that for the starting point at lat: 48.906337, lon: 11.389950, the 30min and 60min isochrone polygons are far from the origin. A similar issue occurs in the second image.

questions:
Since multimodal mode includes walking, even if the entire trip were on foot, I would expect a significant coverage area surrounding the origin. Therefore, I would like to ask:

1. What could be the possible reasons for this behavior? (I have attached my request-code below.)
2. How exactly does Valhalla define an isochrone? Does it represent the farthest reachable distance within the given time (e.g., 60min), an average reachability range, or something else?

code:

def query_isochrone(lat, lon):
    url = "http://localhost:8002/isochrone"
    headers = {'Content-Type': 'application/json'}

    data = {
        "locations": [{"lat": lat, "lon": lon}],  
        "costing": "multimodal",  
        "costing_options": {
            "transit": {"use_bus": 1.0, "use_rail": 1.0, "use_transfers": 1.0},
            "pedestrian": {"walking_speed": 1.4}  
        },
        "contours": [
            {"time": 15},
            {"time": 30},
            {"time": 45},
            {"time": 60}
        ],  
        "date_time": {
            "type": 1,  
            "value": "2025-02-25T07:00"
        },
        "polygons": True  
    }

    response = requests.post(url, headers=headers, data=json.dumps(data))
    response_json = response.json()
    
    return response_json 

Looking forward to your reply! This would greatly help me better understand Valhalla.
Thank you!

[welcome]

Thanks for opening your first issue in this project! If you haven't already, you can join our slack and join the #gtfs-validators channel to meet our awesome community. Come say hi 👋!

Welcome to the community and thank you for your engagement in open source! 🎉