Using geemap.download_ee_image

[jcaceres85]

Hi everyone,

I'm having troubles using the geemap.download_ee_image method, as nothing gets downloaded, the progress gets stuck at 0.0%, showing this message:

Consider adjusting region, scale and/or dtype to reduce the NDVI_NoClouds.tif download size (raw: 1.23 GB).
NDVI_NoClouds.tif: | | 0.00/1.23G (raw) [ 0.0%] in 00:00 (eta: ?)

This is the code that I'm using: https://gist.github.com/jcaceres85/630138065940358d40fefc2b94082238

Am I doing something wrong? I know the image is large, but my understanding is that the geemap.download_ee_image should be able to take care of it by it's description: "Download an Earth Engine Image as a GeoTIFF. Images larger than the `Earth Engine size limit are split and downloaded as separate tiles, then re-assembled into a single GeoTIFF"

[giswqs]

The image is not a ready to download image. You are running a lot of computation in the script. It is going to take a while to make the image ready for download. Try a much large scale value to download a small image before using the native resolution.

[jcaceres85]

Thank you so much for your answer Dr. Wu, I'm just getting familiar with geemap.

I'm not sure if is allowed but just a followed up, using geemap.ee_export_image_to_drive with the same script it works, it get sent to GEE and I can downloaded after the task is finished (about one hour later). With download_ee_image I've been waiting longer and nothing has happened.

Is this the expected behaviour? Should it take longer for download_ee_image?

Thanks for the great work you do!

[giswqs]

download_ee_image is better suited for downloading original data rather than computed data. As you mentioned, it took hours to download the data to Google Drive. If you study area is small, simply filtering the original image collection, sort by cloud coverage, and select the best cloud-free image. Then export it with either functions. It should work. The more images and more steps involved in the computation, the less suitable for downloading them directly to a local computer.

[jcaceres85]

Got it, I understand it better know. Again thank you very much!

[eatasalad]

I have a similar issue when I try to use "geemap.download_ee_image()" to download the image after resample the data but it doesn't work either when I tried to export it the driver. When I use geemap.download_ee_image(), it says: OSError: Error downloading tile: Number of pixels requested from Image.gt exceeds the maximum allowed (2^31). How should I solve this problem? Below is my main code
downloadBA_modis.txt
. Thank you so much!
`year = 2017
month =1

Define time

stAllTime = ee.Date.fromYMD(year, 1, 1)
edAllTime = ee.Date.fromYMD(year, 12, 31)

Create a burned area ImageCollection for the whole year

burnedArea = (ee.ImageCollection('MODIS/061/MCD64A1')
.filterDate(stAllTime, edAllTime)
.select('BurnDate'))
burnedArea_month = ee.Image(burnedArea.toList(12).get(month - 1)) # monthly image

set the None to 0

burnedArea_month = burnedArea_month.updateMask(burnedArea_month.gt(0)).unmask(0)

load a dataset with 0.25 resolution

era5 = ee.Image(ee.ImageCollection('ECMWF/ERA5/MONTHLY').first()).select('mean_2m_air_temperature')

Get information about the ERA projection.

era5Projection = era5.projection()

Initialize a map object.

m = geemap.Map(center=[12, 15], zoom=6)

Load and display burned area data at 500 meters resolution.

m.addLayer(burnedArea_month, {'min':0, 'max':31}, 'burned area at 500 m');

display(m)

Get the burned data at ERA5 scale and projection 27830 meters

burnedArea_monthArea = burnedArea_month.gt(0).reduceResolution(
reducer=ee.Reducer.mean(),
maxPixels=65536
).multiply(ee.Image.pixelArea()
).reproject(
crs=era5Projection
).rename('area')

Display the aggregated, reprojected data.

m.addLayer(burnedArea_monthArea ,{'min':0, 'max':27830 *27830 ,'palette':['#b2182b', '#67001f']}, 'Area at ERA5 scale')
m

out_dir = "C:\Data\BA025"
out_tif_area = os.path.join(out_dir, f"BA_area_025_{year}_{month}.tif")
out_tif_area
region = ee.Geometry.BBox(-180, -90, 180, 90)

geemap.download_ee_image(
image=burnedArea_monthArea,
filename=out_tif_area,
region=region.getInfo()['coordinates'],
crs="EPSG:4326",
scale=27830,
dtype='float32', # Set data type to int16
)

#!!!! OSError: Error downloading tile: Number of pixels requested from Image.gt exceeds the maximum allowed (2^31).

geemap.ee_export_image(

burnedArea_monthArea, filename=out_tif_area, region=region, scale=27830

) # this does not work either

`