Image collection images to a 1D vector where there is one value corresponding to each image

Question

I am currently working with MODI's land surface temperature image collection dataset of 20 years of data. I want to convert each image to a single value representation using mean/max function. I tried the map function, but it returns only images, whereas I am looking for a vector of float values where each value is the max/mean of the temperature of that region.

How can I solve this issue?

Answer 1

You are on the right track with map(). Hopefully this code explains how you can do this:

// Your area of interest
var region = ee.Geometry.Point([12.492362, 41.890232]).buffer(500).bounds()

// Some collection of images. You're using MODIS here
var imageCollection = ee.ImageCollection('COPERNICUS/S2')
  .filterDate('2019-01-01', '2019-02-01')
  .filterBounds(region)

// A feature collection where each feature contains
// mean and max of each image in the image collection
var featureCollection = imageCollection.map(toMeanMax)
print(featureCollection)

// Calculates mean and max for an image
function toMeanMax(image) {
  var meanMax = image.reduceRegion({
     // Reduce the image to mean and max. Can be done separately,
     // but combining two reducers is more efficient.
    reducer: ee.Reducer.mean()
      .combine(ee.Reducer.max(), null, true),
    geometry: region, // The area of the image you want to reduce
    scale: 10 // The resolution in meters to use
  })
  // When mapping over a collection you must return an 
  // image (map function returns an image collection)
  // or a feature (map function returns a feature collection)
  return ee.Feature(null, meanMax) 
}

https://code.earthengine.google.com/68a509283944836705afc9dd989f92e6