If you have not calculated the image on GEE, or you have uploaded the image to GEE's personal asset Assets, how to perform a linear regression on the GEE platform?
This article records the method and code for unary linear regression on the GEE platform using local data image .
The specific steps are as follows:
①. Upload images of different years to Assets;
②. Call multiple images, and add a time band to each scene (used to express the time sequence and prepare for regression);
③. Construct multiple images into one ImageCollection;
④, use ee.Reducer.linearFit() to run ImageCollection;
⑤, output linear regression results, visualize and download the three images of slope, intercept and delta_y;
⑥, calculate Pearson correlation through ee.Reducer.pearsonsCorrelation() Coefficient Pearson correlation and p-value.
Example 1:
Rate of change in ozone (o3) from 2015 to 2019:
Change in ozone (o3) from 2015 to 2019: