search

loicdtx / bfastSpatial

Set of utilities and wrappers to perform change detection on satellite image time-series (Landsat and MODIS). Includes pre-processing steps and functions for spatial implementation of bfastmonitor change detection and post processing of the results.
93 stars 39 forks source link

how to create time series raster layers with self-produced rasters #80

Open myidream opened 6 years ago

myidream commented 6 years ago

Hi guys! I am a new user of “bfastSpatial” R package to detect the breakpoint of vegetation gross primary productivity (GPP) of China during the past 15 years based the monthly spatial GPP data that we produced using VPM model with MODIS images. I meet the problem in the step to create time series GPP data using the timeStack() command. I am not sure whether the “bfastSpatial” package can be applied in raster layers that not from Landsat or MODIS. Any reponse would be helpful.

loicdtx commented 6 years ago

That's right, timeStack() is specific to Landsat and MODIS land data so that it won't work on data that follow other naming conventions. The good thing is that timeStack() is really just a thin wrapper around raster::stack() + raster::setZ(), so that it should be easy to adapt or bypass. See the code here. The key in your case will be to get a vector of dates corresponding to each layer that will compose your stack.