GEE: Construction of long time series based on Landsat images (1985-2020 NDVI annual synthetic time series)

Author: CSDN @ _Yakult_

The code documented in this article is a script for building an annual collection of synthetic imagery . It obtains a collection of Landsat images within a given time range by calling a series of functions, and performs preprocessing and synthesis. It includes steps such as spectral index calculation, band adjustment, and median synthesis of remote sensing images.

The result is shown in the figure below,

The main steps of the script are as follows:

1. Define the dataset parameters, including the time series start year, end year, and start and end dates for each year.

2. Define the spectral index calculation function timeSeriesIndex, which is used to calculate the spectral index (NDVI, etc.).

3. Create a dummy image collection dummyCollectionthat fills in missing years.

4. Defines a function that builds a collection of annual synthetic images buildMosaicCollection, calls buildMosaicthe function to build a median composite image for each year, and adds it to the collection.

5. Define the L8 to L7 band adjustment function harmonizationRoyto band adjust Landsat 8 imagery according to a set of slope and intercept parameters to achieve spectral consistency with Landsat 7.

6. Defines a function that applies median compositing to a collection of images buildMosaicby calling other functions to obtain the