Converting Earth Engine dataset into XArray

[yakubhariana70]

Hi guys,
I am a newbie in Python and starting to learn geemap. Almost two weeks ago I got a task to calculate the phenology value using FuseTS. As far as I know, the Fuse TS input data must be in the XArray dataset. In this case, I want to use Sentinel-2 imagery from the Earth Engine dataset and then convert it using geemap.ee_to_xarray command.
This is my notebook: Earth Engine to XArray

I got an error message like this:
EEException: Geometry.bounds: Unable to perform this geometry operation. Please specify a non-zero error margin.

I already searched the solutions on the internet or used Github Copilot to solve this issue, but the problem is still there.
Do you have any solution? Very thankful if you have some solution

Note:

1. The xarray conversion works well with the source of sentinel-2 imagery (sentinel2_collection)
2. But the problem shows when the conversion applied to the result of NDVI imagery

[giswqs]

You need to specify the geometry parameter, like this:

[yakubhariana70]

Wow! Dr. Wu my online guru, really appreciates your help.
Terimakasih :D

[jevegabriones]

Hi, I hope is possible to follow up on this question. I tried to implement MODIS EVI but the clipping doesn't work for the ROI, giving back the whole world. Is possible that ee_to_xarray() is not working properly for that dataset?

I tried with two MODIS collections and is the same result. Here the code:

`def calculate_MODIS_EVI1(image):
evi = image.expression(
'2.5 * ((NIR - RED) / (NIR + 2.4 * RED + 1))', {
'NIR': image.select('Nadir_Reflectance_Band2'),
'RED': image.select('Nadir_Reflectance_Band1')
}).rename('EVI')

return image.multiply(0.0001).addBands(evi).copyProperties(image, ['system:time_start'])

def calculate_MODIS_EVI2(image):
evi = image.expression(
'2.5 * ((NIR - RED) / (NIR + 6 * RED - 7.5 * BLUE + 1))', {
'NIR': image.select('sur_refl_b02'),
'RED': image.select('sur_refl_b01'),
'BLUE': image.select('sur_refl_b03')
}).rename('EVI')
return image.multiply(0.0001).addBands(evi).copyProperties(image, ['system:time_start'])

modis_collection1 = (ee.ImageCollection('MODIS/006/MCD43A4')
.filterDate(start_date, end_date)
.filterBounds(roi)
.sort('system:time_start')
)

modis_collection2 = (ee.ImageCollection('MODIS/006/MOD13Q1')
.filterDate(start_date, end_date)
.filterBounds(roi)
.sort('system:time_start')
)

evi_image = modis_collection1.map(calculate_MODIS_EVI1) #or
evi_image = modis_collection2.map(calculate_MODIS_EVI2)

evi_selected = evi_image.select('EVI')

evi_array1 = geemap.ee_to_xarray(evi_image, crs="EPSG:4326", geometry=modis_collection1.first().geometry())
print(evi_array.values)
evi_array2 = geemap.ee_to_xarray(evi_image, crs="EPSG:4326", geometry=modis_collection2.first().geometry())
print(evi_array.values)`

In both the result is as follows:

<bound method Mapping.values of <xarray.Dataset> Size: 949MB
Dimensions: (time: 244, lon: 360, lat: 180)
Coordinates:

• time (time) datetime64[ns] 2kB 2022-...
• lon (lon) float64 3kB -179.5 ... 179.5
• lat (lat) float64 1kB -89.5 ... 89.5
  Data variables: (12/15)
  Nadir_Reflectance_Band1 (time, lon, lat) float32 63MB ...
  Nadir_Reflectance_Band2 (time, lon, lat) float32 63MB ...
  Nadir_Reflectance_Band3 (time, lon, lat) float32 63MB ...
  Nadir_Reflectance_Band4 (time, lon, lat) float32 63MB ...
  Nadir_Reflectance_Band5 (time, lon, lat) float32 63MB ...
  Nadir_Reflectance_Band6 (time, lon, lat) float32 63MB ...
  ... ...
  BRDF_Albedo_Band_Mandatory_Quality_Band3 (time, lon, lat) float32 63MB ...
  BRDF_Albedo_Band_Mandatory_Quality_Band4 (time, lon, lat) float32 63MB ...
  BRDF_Albedo_Band_Mandatory_Quality_Band5 (time, lon, lat) float32 63MB ...
  BRDF_Albedo_Band_Mandatory_Quality_Band6 (time, lon, lat) float32 63MB ...
  BRDF_Albedo_Band_Mandatory_Quality_Band7 (time, lon, lat) float32 63MB ...
  EVI (time, lon, lat) float32 63MB ...
  Attributes: (12/20)
  date_range: [950832000000, 1653868800000]
  description:

  Provider:
  

  Cade...
  keywords: ['albedo', 'brdf', 'daily', 'global', 'mcd43a4', ...
  period: 1
  period_mapping: [950832000000, 1653868800000]
  product_tags: ['albedo', 'brdf', 'reflectance']
  ... ...
  visualization_0_bands: Nadir_Reflectance_Band1,Nadir_Reflectance_Band4,N...
  visualization_0_gamma: 1.5
  visualization_0_max: 4000.0
  visualization_0_min: 0.0
  visualization_0_name: True Color
  crs: EPSG:4326>

[giswqs]

Do not use modis_collection2.first().geometry(), because each MODIS covers the whole world. Specify a polygon or bbox geometry

[jevegabriones]

Thanks for the fast reply. I'm using the same code provided by @yakubhariana70: Earth Engine to XArray, but replacing the part I pasted.
Unfortunately, the clip (.filterBounds(roi)) is not working and providing the whole world as a xr.DataSet output.
I made a copy with the code