Converting list to feature collection in Google Earth Engine (python api) failing

Question

I am trying to convert a list of images in GEE to an image collection, then get the mean and export the results. However, the code I have doesn't seem to work. The error when exporting the image is Image.clip: The geometry for image clipping must be bounded. However, I can both view an individual image from the list and the mean image.

Here is the code.

import ee
import folium
import time
import webbrowser

ee.Initialize()

region = "Romania"
start_date = '2023-01-01' # initial date of interest (inclusive)
end_date = '2023-12-31' # final data of interest (exclusive)
band = "total_precipitation_sum"
folder = "Romania"
outname = "Romania_total_precip"

# Get a feature collection of administrative boundaries and filter by country
countries = ee.FeatureCollection('FAO/GAUL/2015/level0').select('ADM0_NAME')
region = countries.filter(ee.Filter.eq('ADM0_NAME', 'Romania'))
region_list = region.toList(region.size()) # klugey
region = ee.Image(region_list.get(0))

### get weather image collection
weatherData = ee.ImageCollection("ECMWF/ERA5_LAND/DAILY_AGGR").filter(ee.Filter.date("2023-01-01", "2023-12-31")).map(lambda image: image.clip(region.geometry()))
precip = weatherData.select(band)

precip_list = precip.toList(precip.size())

image_from_list = ee.Image(precip_list.get(0)).multiply(1000)

task = ee.batch.Export.image.toDrive(
    image = image_from_list,
    description = "RomanianPrecip1",
    folder = folder,
    region = image_from_list.geometry()
    )    
task.start()
status = None
while status != 'COMPLETED':
    status = task.status()['state']
    time.sleep(30) ## wait 60 seconds if not completed
    print(task.status())

Precip_coll = ee.ImageCollection.fromImages(precip_list)
Precip_mean = Precip_coll.mean().multiply(1000)

   
task = ee.batch.Export.image.toDrive(
    image = Precip_mean,
    description = "RomanianPrecip_mean",
    folder = folder,
    region = Precip_mean.geometry()
    )    
task.start()
status = None
while status != 'COMPLETED':
    status = task.status()['state']
    time.sleep(30) ## wait 60 seconds if not completed
    print(task.status())

Both images can be displayed in folium, so not sure what the problem is.

Answer 1

---

The answer seems to be that when the list of images is converted back to an image collection, it doesn't inherit the geometry of the original feature collection.

image_from_list.geometry().getInfo()
Precip_mean.geometry().getInfo()

yield two different geometries. I didn't expect this, but should have checked first. If I type Precip_mean = Precip_coll.mean().multiply(1000).clip(region), it works.