Mapping over image collection

Mapping over an ImageCollection

Reference: https://developers.google.com/earth-engine/guides/ic_mapping

To apply a function to every Image in an ImageCollection use imageCollection.map(). The only argument to map() is a function which takes one parameter: an ee.Image.

In [1]:

import ee
import geemap

import ee import geemap

In [2]:

Map = geemap.Map()

Map = geemap.Map()

Define a function¶

In [3]:

def minus1000(num):
    result = ee.Number(num).subtract(1000)
    return result

def minus1000(num): result = ee.Number(num).subtract(1000) return result

In [4]:

result = minus1000(2000)
print(result.getInfo())

result = minus1000(2000) print(result.getInfo())

1000

Apply the map function¶

In [5]:

years = ee.List.sequence(2000, 2010)
print(years.getInfo())

years = ee.List.sequence(2000, 2010) print(years.getInfo())

[2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010]

In [6]:

results = years.map(minus1000)
print(results.getInfo())

results = years.map(minus1000) print(results.getInfo())

[1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009, 1010]

Use anonymous/lambda function¶

In [7]:

years = ee.List.sequence(2000, 2010)

results = years.map(lambda x: ee.Number(x).subtract(1000))
print(results.getInfo())

years = ee.List.sequence(2000, 2010) results = years.map(lambda x: ee.Number(x).subtract(1000)) print(results.getInfo())

[1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009, 1010]

Filter an ImageCollection¶

In [8]:

roi = ee.Geometry.Point(-83.92, 35.96)  # Knoxville, Tennessee

collection = (
    ee.ImageCollection('LANDSAT/LC08/C01/T1_SR')
    .filterBounds(roi)
    .filterDate("2019-01-01", "2020-01-01")
    .sort("CLOUD_COVER")
)

print(collection.size().getInfo())

roi = ee.Geometry.Point(-83.92, 35.96) # Knoxville, Tennessee collection = ( ee.ImageCollection('LANDSAT/LC08/C01/T1_SR') .filterBounds(roi) .filterDate("2019-01-01", "2020-01-01") .sort("CLOUD_COVER") ) print(collection.size().getInfo())

17

In [9]:

image = collection.first()

image = collection.first()

In [10]:

# image.getInfo()

# image.getInfo()

In [11]:

geemap.image_props(image).getInfo()

geemap.image_props(image).getInfo()

Out[11]:

{'CLOUD_COVER': 0.06,
 'CLOUD_COVER_LAND': 0.06,
 'EARTH_SUN_DISTANCE': 0.994981,
 'ESPA_VERSION': '2_23_0_1b',
 'GEOMETRIC_RMSE_MODEL': 7.624,
 'GEOMETRIC_RMSE_MODEL_X': 5.066,
 'GEOMETRIC_RMSE_MODEL_Y': 5.698,
 'IMAGE_DATE': '2019-03-17',
 'IMAGE_QUALITY_OLI': 9,
 'IMAGE_QUALITY_TIRS': 9,
 'LANDSAT_ID': 'LC08_L1TP_019035_20190317_20190325_01_T1',
 'LEVEL1_PRODUCTION_DATE': 1553541347000,
 'NOMINAL_SCALE': 30,
 'PIXEL_QA_VERSION': 'generate_pixel_qa_1.6.0',
 'SATELLITE': 'LANDSAT_8',
 'SENSING_TIME': '2019-03-17T16:11:33.5001919Z',
 'SOLAR_AZIMUTH_ANGLE': 144.944595,
 'SOLAR_ZENITH_ANGLE': 43.08416,
 'SR_APP_VERSION': 'LaSRC_1.3.0',
 'WRS_PATH': 19,
 'WRS_ROW': 35,
 'system:asset_size': '641.99612 MB',
 'system:band_names': ['B1',
  'B2',
  'B3',
  'B4',
  'B5',
  'B6',
  'B7',
  'B10',
  'B11',
  'sr_aerosol',
  'pixel_qa',
  'radsat_qa'],
 'system:id': 'LANDSAT/LC08/C01/T1_SR/LC08_019035_20190317',
 'system:index': 'LC08_019035_20190317',
 'system:time_end': '2019-03-17 16:11:33',
 'system:time_start': '2019-03-17 16:11:33',
 'system:version': 1563396666124498}

In [12]:

collection.aggregate_array("CLOUD_COVER").getInfo()

collection.aggregate_array("CLOUD_COVER").getInfo()

Out[12]:

[0.06,
 0.42,
 0.57,
 0.76,
 1.26,
 6.67,
 9.69,
 13.14,
 22.55,
 26.67,
 29.89,
 33.11,
 51.76,
 74.72,
 78.86,
 85.73,
 95.02]

In [13]:

vis_params = {'bands': ['B5', 'B4', 'B3'], 'min': 0, 'max': 3000, 'gamma': 1.0}

Map.addLayer(image, vis_params, "First image")
Map.centerObject(image)
Map

vis_params = {'bands': ['B5', 'B4', 'B3'], 'min': 0, 'max': 3000, 'gamma': 1.0} Map.addLayer(image, vis_params, "First image") Map.centerObject(image) Map

Out[13]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Select the best image each year¶

In [14]:

roi = ee.Geometry.Point(-83.92, 35.96)  # Knoxville, Tennessee


def best_image(year):
    start_date = ee.Date.fromYMD(year, 1, 1)
    end_date = start_date.advance(1, 'year')

    image = (
        ee.ImageCollection('LANDSAT/LC08/C01/T1_SR')
        .filterBounds(roi)
        .filterDate(start_date, end_date)
        .sort("CLOUD_COVER")
        .first()
    )

    return image

roi = ee.Geometry.Point(-83.92, 35.96) # Knoxville, Tennessee def best_image(year): start_date = ee.Date.fromYMD(year, 1, 1) end_date = start_date.advance(1, 'year') image = ( ee.ImageCollection('LANDSAT/LC08/C01/T1_SR') .filterBounds(roi) .filterDate(start_date, end_date) .sort("CLOUD_COVER") .first() ) return image

In [15]:

start_year = 2013
end_year = 2020
years = ee.List.sequence(start_year, end_year)
year_list = years.getInfo()
print(year_list)

start_year = 2013 end_year = 2020 years = ee.List.sequence(start_year, end_year) year_list = years.getInfo() print(year_list)

[2013, 2014, 2015, 2016, 2017, 2018, 2019, 2020]

In [16]:

images = years.map(best_image)

images = years.map(best_image)

In [17]:

count = images.size().getInfo()
print(count)

count = images.size().getInfo() print(count)

8

In [18]:

ee.ImageCollection(images).aggregate_array("CLOUD_COVER").getInfo()

ee.ImageCollection(images).aggregate_array("CLOUD_COVER").getInfo()

Out[18]:

[0.09, 0.13, 0.03, 0.01, 0.03, 0.01, 0.06, 4.61]

In [19]:

for index in range(0, count):
    image = ee.Image(images.get(index))
    layer_name = "Image " + str(year_list[index])
    Map.addLayer(image, vis_params, layer_name, False)

for index in range(0, count): image = ee.Image(images.get(index)) layer_name = "Image " + str(year_list[index]) Map.addLayer(image, vis_params, layer_name, False)

In [20]:

Map

Map

Out[20]:

Make this Notebook Trusted to load map: File -> Trust Notebook

Set image properties¶

In [21]:

collection = ee.ImageCollection(images)

collection = ee.ImageCollection(images)

In [22]:

collection.aggregate_array("system:time_start").getInfo()

collection.aggregate_array("system:time_start").getInfo()

Out[22]:

[1382804022390,
 1399392695740,
 1442247113710,
 1454688719630,
 1494778283540,
 1525191067880,
 1552839093500,
 1591546292340]

In [23]:

collection = collection.map(
    lambda img: img.set(
        {"DATE": ee.Date(img.get("system:time_start")).format("YYYY-MM-dd")}
    )
)

collection = collection.map( lambda img: img.set( {"DATE": ee.Date(img.get("system:time_start")).format("YYYY-MM-dd")} ) )

In [24]:

collection.aggregate_array("DATE").getInfo()

collection.aggregate_array("DATE").getInfo()

Out[24]:

['2013-10-26',
 '2014-05-06',
 '2015-09-14',
 '2016-02-05',
 '2017-05-14',
 '2018-05-01',
 '2019-03-17',
 '2020-06-07']

---

Last update: 2022-03-25