For my Master thesis I am producing a trend analysis of NDVI in a rather large study area over the last decades using Landsat sensors in Google Earth Engine. My idea is to produce a median NDVI for each month and afterwards median / max NDVI for the summer (green) months. However, many images contain areas with full cloud coverage for the entire month and applying a cloud mask results in quite extensive no data areas. Is it acceptable for the analysis to include these patchy NDVI images while, for example, adding a histogram explaining the cloud coverage? Or is there a better way for this type of analysis? For demonstration I uploaded a PNG of my study area showing a NDVI map of one month containing no data areas.
Aggregate monthly NDVI by pixel maximum and for summer (green) months use the mean or median.
Hello,
You may use the following codes in the given links to remove the cloud from NDVI datasets:
https://gis.stackexchange.com/questions/333883/removing-clouds-from-sentinel-2-surface-reflectance-in-google-earth-engine
https://gis.stackexchange.com/questions/332455/removing-clouds-from-landsat-8-surface-reflectance-in-google-earth-engine
https://www.earthdatascience.org/courses/use-data-open-source-python/multispectral-remote-sensing/landsat-in-Python/remove-clouds-from-landsat-data/
Also, the following article maybe useful to you:
Article The Grass Is Not Always Greener on the Other Side: Seasonal ...
Cheers
Thanks for the answers! In fact, I was able to remove the clouds, my concern is more about reducing all of the images of the summer months (may - september) to one, while many pixels are missing values because of cloud coverage. Is it for example suitable for calculating a median of a specific pixel which has no data for may, july and august?
- Badges
- Science topic
- Similar topics
- Statistics
- Time Series