1

I am using Landsat 5, 7, 8 images and RStudio. My goal is to compute the disturbance in a given area using the Vegetation Change Tracker (VCT) algorithm. I would like to ask you if you could provide me a general guidance on how to construct it using R Studio. I computed the IFZ of each individual image, I created the LTSS, I have the NDVI and the NBR of each year. My problem is on how to create the rules required form the VCT (see the image below).Decision rules used by eh vct algorithm (Huang, 2010)

The acronyms CLIV, MNCLIV and MNCHIV means:

  • the pixel was classified as a forest at least in 2 sequentially images,
  • the minimum years of sequentially images needed for a pixel to be classified as a forest (2 years) and
  • minimum number of consecutive high IFZ values for a pixel to be classified as a non-forest pixel.

The paper named An automated approach for reconstructing recent forest disturbance history using dense Landsat time series stack.

  • 1
    What is the form of your data? How are the various bands and time-point rasters stored? – Spacedman Feb 17 '19 at 20:37
  • I had yearly Landsat images of every band in raster format (.tif) which I downloaded using Google Earth Engine. Also, I had the NDVI and NBR indices in the same format for every year. Based on those indices I created random points in areas which I knew that it was dense forest (NDVI > 0.6) or a disturbance was occurred (e.g. a fire). Then, I used the Point Sampling Tool in QGIS in order to check the spectral value in the bands B3, B5 & B7 and finally I computed the FZ and IFZ indices for every image. Lastly, I created a time series stack of the yearly IFZ indices in R using the rts package. – Vasilis Ntoufas Feb 18 '19 at 14:08
  • Edit your question and describe the form of the R objects you have - their names, their layers, their time structure, a description of what the values are and so on. Give the output of summary or print on the objects. At the moment we can't get past the first step since we have no idea where your water classification is. – Spacedman Feb 18 '19 at 14:38