AI-generated Key Takeaways
-
Vegetation Change Tracker (VCT) is an automated method for reconstructing recent forest disturbance history using dense Landsat time series.
-
VCT outputs a 2D array per pixel for multiple years, providing information on input years, landcover mask, and the magnitude of disturbance in different indices (UD composite, B4, NDVI, dNBR).
-
The
ee.Algorithms.TemporalSegmentation.VCTalgorithm is used to apply VCT, requiring a time series collection with specific bands and a land cover collection. -
The VCT algorithm includes parameters to control the detection of forest and the number of years analyzed.
The output is a 2D array per pixel containing 6 rows x N years. The output rows contain: input years, VCT landcover mask, magnitude in term of the UD composite, magnitude of distubance in B4, magnitude of distubance in NDVI, magnitude of distubance in dNBR.
See: Huang, C., Goward, S.N., Masek, J.G., Thomas, N., Zhu, Z. and Vogelmann, J.E., 2010. An automated approach for reconstructing recent forest disturbance history using dense Landsat time series stacks. Remote Sensing of Environment, 114(1), pp.183-198.
| Usage | Returns |
|---|---|
ee.Algorithms.TemporalSegmentation.VCT(timeSeries, landCover, maxUd, minNdvi, forThrMax, nYears) | Image |
| Argument | Type | Details |
|---|---|---|
timeSeries | ImageCollection | Collection from which to extract VCT disturbances, containing the bands: B3, B4, B5, B7, thermal, NDVI, DNBR, and COMP. This collection is expected to contain 1 image for each year, sorted by time. |
landCover | ImageCollection | Collection from which to extract VCT masks. This collection is expected to contain 1 image for each image in the timeSeries, sorted by time. |
maxUd | Float, default: 4 | Maximum Z-score composite value for detecting forest. |
minNdvi | Float, default: 0.45 | Minimum NDVI value for forest. |
forThrMax | Float, default: 3 | Maximum threshold for forest. |
nYears | Integer, default: 30 | Maximum number of years. |