Announcement: All noncommercial projects registered to use Earth Engine before April 15, 2025 must verify noncommercial eligibility to maintain Earth Engine access.
[[["Easy to understand","easyToUnderstand","thumb-up"],["Solved my problem","solvedMyProblem","thumb-up"],["Other","otherUp","thumb-up"]],[["Missing the information I need","missingTheInformationINeed","thumb-down"],["Too complicated / too many steps","tooComplicatedTooManySteps","thumb-down"],["Out of date","outOfDate","thumb-down"],["Samples / code issue","samplesCodeIssue","thumb-down"],["Other","otherDown","thumb-down"]],["Last updated 2023-10-06 UTC."],[[["\u003cp\u003e\u003ccode\u003eImageCollection.median()\u003c/code\u003e reduces an image collection to a single image by calculating the median pixel value across all images in the collection for each band.\u003c/p\u003e\n"],["\u003cp\u003eBands are matched by name when computing the median, ensuring that the same band from each image is used in the calculation.\u003c/p\u003e\n"],["\u003cp\u003eThe result is a single image representing the median values of the input image collection, useful for summarizing central tendencies in the data.\u003c/p\u003e\n"],["\u003cp\u003eThis method is applicable to any type of numeric band in the input image collection.\u003c/p\u003e\n"]]],["The `median()` method reduces an `ImageCollection` to a single `Image` by calculating the median value for each pixel across all images in the collection. This is done for all bands that are matched by name. The method takes one argument, which is the `ImageCollection` itself. Each image pixel value that is available for a band will be used to calculate the median. This median is the value that will be output in the band of the new `Image`.\n"],null,["# ee.ImageCollection.median\n\nReduces an image collection by calculating the median of all values at each pixel across the stack of all matching bands. Bands are matched by name.\n\n\u003cbr /\u003e\n\n| Usage | Returns |\n|----------------------------|---------|\n| ImageCollection.median`()` | Image |\n\n| Argument | Type | Details |\n|--------------------|-----------------|---------------------------------|\n| this: `collection` | ImageCollection | The image collection to reduce. |\n\nExamples\n--------\n\n### Code Editor (JavaScript)\n\n```javascript\n// Sentinel-2 image collection for July 2021 intersecting a point of interest.\n// Reflectance, cloud probability, and scene classification bands are selected.\nvar col = ee.ImageCollection('COPERNICUS/S2_SR')\n .filterDate('2021-07-01', '2021-08-01')\n .filterBounds(ee.Geometry.Point(-122.373, 37.448))\n .select('B.*|MSK_CLDPRB|SCL');\n\n// Visualization parameters for reflectance RGB.\nvar visRefl = {\n bands: ['B11', 'B8', 'B3'],\n min: 0,\n max: 4000\n};\nMap.setCenter(-122.373, 37.448, 9);\nMap.addLayer(col, visRefl, 'Collection reference', false);\n\n// Reduce the collection to a single image using a variety of methods.\nvar mean = col.mean();\nMap.addLayer(mean, visRefl, 'Mean (B11, B8, B3)');\n\nvar median = col.median();\nMap.addLayer(median, visRefl, 'Median (B11, B8, B3)');\n\nvar min = col.min();\nMap.addLayer(min, visRefl, 'Min (B11, B8, B3)');\n\nvar max = col.max();\nMap.addLayer(max, visRefl, 'Max (B11, B8, B3)');\n\nvar sum = col.sum();\nMap.addLayer(sum,\n {bands: ['MSK_CLDPRB'], min: 0, max: 500}, 'Sum (MSK_CLDPRB)');\n\nvar product = col.product();\nMap.addLayer(product,\n {bands: ['MSK_CLDPRB'], min: 0, max: 1e10}, 'Product (MSK_CLDPRB)');\n\n// ee.ImageCollection.mode returns the most common value. If multiple mode\n// values occur, the minimum mode value is returned.\nvar mode = col.mode();\nMap.addLayer(mode, {bands: ['SCL'], min: 1, max: 11}, 'Mode (pixel class)');\n\n// ee.ImageCollection.count returns the frequency of valid observations. Here,\n// image pixels are masked based on cloud probability to add valid observation\n// variability to the collection. Note that pixels with no valid observations\n// are masked out of the returned image.\nvar notCloudCol = col.map(function(img) {\n return img.updateMask(img.select('MSK_CLDPRB').lte(10));\n});\nvar count = notCloudCol.count();\nMap.addLayer(count, {min: 1, max: 5}, 'Count (not cloud observations)');\n\n// ee.ImageCollection.mosaic composites images according to their position in\n// the collection (priority is last to first) and pixel mask status, where\n// invalid (mask value 0) pixels are filled by preceding valid (mask value \u003e0)\n// pixels.\nvar mosaic = notCloudCol.mosaic();\nMap.addLayer(mosaic, visRefl, 'Mosaic (B11, B8, B3)');\n```\nPython setup\n\nSee the [Python Environment](/earth-engine/guides/python_install) page for information on the Python API and using\n`geemap` for interactive development. \n\n```python\nimport ee\nimport geemap.core as geemap\n```\n\n### Colab (Python)\n\n```python\n# Sentinel-2 image collection for July 2021 intersecting a point of interest.\n# Reflectance, cloud probability, and scene classification bands are selected.\ncol = (\n ee.ImageCollection('COPERNICUS/S2_SR')\n .filterDate('2021-07-01', '2021-08-01')\n .filterBounds(ee.Geometry.Point(-122.373, 37.448))\n .select('B.*|MSK_CLDPRB|SCL')\n)\n\n# Visualization parameters for reflectance RGB.\nvis_refl = {'bands': ['B11', 'B8', 'B3'], 'min': 0, 'max': 4000}\nm = geemap.Map()\nm.set_center(-122.373, 37.448, 9)\nm.add_layer(col, vis_refl, 'Collection reference', False)\n\n# Reduce the collection to a single image using a variety of methods.\nmean = col.mean()\nm.add_layer(mean, vis_refl, 'Mean (B11, B8, B3)')\n\nmedian = col.median()\nm.add_layer(median, vis_refl, 'Median (B11, B8, B3)')\n\nmin = col.min()\nm.add_layer(min, vis_refl, 'Min (B11, B8, B3)')\n\nmax = col.max()\nm.add_layer(max, vis_refl, 'Max (B11, B8, B3)')\n\nsum = col.sum()\nm.add_layer(\n sum, {'bands': ['MSK_CLDPRB'], 'min': 0, 'max': 500}, 'Sum (MSK_CLDPRB)'\n)\n\nproduct = col.product()\nm.add_layer(\n product,\n {'bands': ['MSK_CLDPRB'], 'min': 0, 'max': 1e10},\n 'Product (MSK_CLDPRB)',\n)\n\n# ee.ImageCollection.mode returns the most common value. If multiple mode\n# values occur, the minimum mode value is returned.\nmode = col.mode()\nm.add_layer(\n mode, {'bands': ['SCL'], 'min': 1, 'max': 11}, 'Mode (pixel class)'\n)\n\n# ee.ImageCollection.count returns the frequency of valid observations. Here,\n# image pixels are masked based on cloud probability to add valid observation\n# variability to the collection. Note that pixels with no valid observations\n# are masked out of the returned image.\nnot_cloud_col = col.map(\n lambda img: img.updateMask(img.select('MSK_CLDPRB').lte(10))\n)\ncount = not_cloud_col.count()\nm.add_layer(count, {'min': 1, 'max': 5}, 'Count (not cloud observations)')\n\n# ee.ImageCollection.mosaic composites images according to their position in\n# the collection (priority is last to first) and pixel mask status, where\n# invalid (mask value 0) pixels are filled by preceding valid (mask value \u003e0)\n# pixels.\nmosaic = not_cloud_col.mosaic()\nm.add_layer(mosaic, vis_refl, 'Mosaic (B11, B8, B3)')\nm\n```"]]