Error when passing in ee.Projection() as parameter for open_dataset() about xee HOT 2 OPEN

I took a quick look. Xee doesn't do a great job exporting data in this image's native projection. Earth Engine batch export tools are the way to go for something like that.

I think you've hit a bug. We calculate the specified region's bounds in EPSG:32605 projection and it ends up being a complex multi-polygon (which Xee isn't equipped to deal with). You could consider using a different projection for the output (EPSG:4326 for example).

from xee.

I got the same issue, so I tried to void passing a geometry.

jasper_bbox = ee.Geometry.Rectangle(-118.19, 52.00, -117.92, 53.00)
lc9 = ee.ImageCollection('LANDSAT/LC09/C02/T1_L2') \
.filterBounds(jasper_bbox)\
.filterMetadata('CLOUD_COVER', 'less_than', 1)
crs = lc9.first().select('SR_B2').projection().getInfo()['crs']
lc9 = lc9.map(lambda img: img.reproject(crs, scale = 30))

lc9_clip = lc9.map(lambda img: img.clip(jasper_bbox))

ds = xr.open_dataset(
    lc9,
    engine='ee',
    crs = crs,
    scale = 30
)

But no matter whether I pass lc9 or lc9_clip, I got the same dimension (time: 6 X: 12299 Y: 310967).
Do you know any workaround?

BTW, I also tried to convert the projection of the geometry to UTM and found it returns a polygon with 19 vertices. Got the same Value error in return.

from xee.