The package provides tools to fill-in missing values in satellite data.
It can be used to gap-fill, e.g., MODIS NDVI data and
is helpful when developing new gap-fill algorithms.
The methods are tailored to data (images) observed at equally-spaced points in time.
This is typically the case for MODIS land surface products and AVHRR NDVI data, among others.
The predictions of the missing values are based on a subset-predict procedure, i.e.,
each missing value is predicted separately by
(1) selecting subsets of the data that are in a neighborhood around the missing point and
(2) predicting the missing value based on the subset.
The main function of the package is Gapfill.
Features
Gap-filling can be executed in parallel.
Users may define new Subset and Predict functions
and run alternative prediction algorithms with little effort.
See Extend for more information and examples.
Visualization of space-time data are simplified through the ggplot2-based
function Image.
F. Gerber, R. Furrer, G. Schaepman-Strub, R. de Jong, M. E. Schaepman, 2016,
Predicting missing values in spatio-temporal satellite data.
http://arxiv.org/abs/1605.01038.