Get feature values from ImageCollection to list

Question

I am calculating the mean precipitation value over a given area from CHIRPS precipitation data. My goal is to get a list containing a date and precipitation value for every Image of the ImageCollection.

Using .getInfo() on the reduced ImageCollection I can get a dict representing the ImageCollection and from there select the values on the client side.

import ee
from datetime import datetime

ee.Initialize()

dataset = ee.ImageCollection('UCSB-CHG/CHIRPS/DAILY').filter(ee.Filter.date('2016-01-01','2016-12-31'))
area = ee.Geometry.Polygon([[[29.045341997822412, -2.1339228457039368],
                                               [29.07984593458999, -2.1339228457039368],
                                               [29.07984593458999, -2.0966977113182073],
                                               [29.045341997822412, -2.0966977113182073],
                                               [29.045341997822412, -2.1339228457039368]]])

 # Image reduction applied to each image.
def reduce_dataset_region(image):
    # Calculate mean of precipitation on defined area.
    local_precipitation_image = image.reduceRegion(
        reducer=ee.Reducer.mean(),
        geometry=area,
        scale=20
    )

    return image.set('mean', local_precipitation_image)

# Apply region reduction to ImageCollection
reduced_dataset = dataset.map(reduce_dataset_region, True)

# Request server-side ImageCollection as dict
reduced_dataset_dict = reduced_dataset.getInfo()

# Create list to hold daily precipitation values.
precipitation_list = []

# Loop through features in image collection dict. One feature represents one image.
for feature in reduced_dataset_dict['features']:
    # Get date and turn into datetime
    # Get precipitation value and add with datetime to list.
    precipitation_list.append((timestamp, feature['properties']['mean']['precipitation']))

print(precipitation_list)

This seems slow and not the right way to go, especially since I want to use data from multiple years.

How can I create a list containing the precipitation value for each Image and send only that, and not the whole ImageCollection, to the client? What is the better alternative to .getInfo here?

Answer 1

I found a solution using the iterate() function (https://developers.google.com/earth-engine/apidocs/ee-imagecollection-iterate). It calls the reduce_dataset_region function and returns a list containing the precipitation values.

This example helped understanding how to use it.

import ee

ee.Initialize()

dataset = ee.ImageCollection('UCSB-CHG/CHIRPS/DAILY').filter(ee.Filter.date('2016-01-01', '2016-12-31'))
area = ee.Geometry.Polygon([[[29.045341997822412, -2.1339228457039368],
                                               [29.07984593458999, -2.1339228457039368],
                                               [29.07984593458999, -2.0966977113182073],
                                               [29.045341997822412, -2.0966977113182073],
                                               [29.045341997822412, -2.1339228457039368]]])

# Create list to hold precipitation values
precipitation_list = ee.List([])

 # Image reduction applied to each image.
def reduce_dataset_region(image, list):
    # Calculate mean of precipitation on defined area.
    local_precipitation_image = image.reduceRegion(
        reducer=ee.Reducer.mean(),
        geometry=area,
        scale=20
    )

    return ee.List(list).add(local_precipitation_image)

# Apply region reduction to ImageCollection and return result in List
reduced_dataset = dataset.iterate(reduce_dataset_region, precipitation_list)

# Request server-side List  as dict
reduced_dataset_dict = reduced_dataset.getInfo()