Odd time_interval behavior

EO Products / Scripts
1

I am trying to download HLS images using a time_interval. This works:
time_interval = (“2025-01-15”, “2025-02-16”)
and this works:
time_interval = (“2024-02-15”, “2024-02-16”)
but this does not:
time_interval = (“2025-02-15”, “2025-02-16”)

This is the error: 400 Client Error: Bad Request for url: Sentinel Hub
Server response: “{“status”: 400, “reason”: “Bad Request”, “message”: “Invalid format”, “code”: “COMMON_BAD_PAYLOAD”, “errors”: {“parameter”: “input->data[0]->dataFilter->timeRange->from”, “invalidValue”: “…”, “description”: “ISO-8601 time representing start of search interval, e.g. 2019-01-31T14:00:00+01:00”}}”

Is there an explanation for that? I know that HLS tiles from 2025-02-15 over my region of interest were added to the SentinelHub archive in the last 24 hours and ideally I’d like to specify a single day for the search but my understanding is that EOWorkflow.execute only accepts a date range so I extended the search to two days. If it is possible to search for a single day please let me know how to do that.

2

Hi Ned,

I’m not sure exactly what your workflow is but assuming you are using EO Learn? If so, it is not really designed for extracting time intervals as short as one day. However, I think you can solve your issue by adding the time to your time interval inputs like this: 2025-02-18T00:00:00Z, 2025-02-18T23:59:59Z

This should mean that your request works correctly.

3

Thank you @william.ray. I tried adding the time component, but it would only work if I increased the range of days to 3 or more. This didn’t work:
time_interval = (“2025-02-15T00:00:00Z”, “2025-02-15T23:59:59Z”)
But this did:
time_interval = (“2025-02-15T00:00:00Z”, “2025-02-18T23:59:59Z”)

This is how I use time_interval:
workflow_nodes = linearly_connect_tasks(add_ms_bands, add_quality_bands, save, output_task)
workflow = EOWorkflow(workflow_nodes)

result = workflow.execute(
    {
        workflow_nodes[0]: {"bbox": roi_bbox, "time_interval": time_interval},
        workflow_nodes[-2]: {"eopatch_folder": farm},
    }
)

I’m trying to adapt a workflow that downloads several months of historic imagery over several ROIs. Now, I want to check every 24 hours to know when new image tiles over my ROIs are added to the archive. When new image tiles are found, imagery over the relevant ROIs should be downloaded and processed. Up to now, I’ve been only using eolearn for this, but I’ll look into other options. If you have a suggestion for the “best” tool to download imagery for specific dates, please let me know, and I’ll focus on learning to create a new workflow.

4

I think I know why I’m getting the error. When I run the program, I try to download images for all of my ROIs for a single date. The error occurs when I try to access an image over an ROI that doesn’t have a corresponding image tile for that day. I expect I’ll need to assign a tile ID to each ROI and ensure I’m only accessing ROI with an image available for the day.