SPLITS (Spline Analysis of Time Series) is a software package to analyze time series of remotely sensed parameters such as vegetation indices. It has been designed for handling large datasets on Linux and UNIX platforms. Initially developed during my PhD thesis , SPLITS has been constantly developed and refined. The various spline models available in SPLITS can handle irregular time series by using time stamped data available now in more and more archives. MODIS vegetation index products , e.g. are time sta
The user interface of SPLITS follows a classical UNIX toolbox concept: There is one program for every purpose, and several programs may be combined using shell scripts to manage more complex tasks. With only a few exceptions, all programs are command line based. Of course, there are programs with graphical interfaces to explore your time series imagery and try different spline models on your data before you settle for a model that you want to apply to your entire dataset using the command line tools. The co
The typical workflow with SPLITS is to first prepare a time series dataset, using either the preprocessing tool ( prepro ) that comes with SPLITS, or your own tools. The time series dataset can then be viewed and explored using splview to select an appropriate model. To apply the model to your entire dataset, use splfit . The result is an intermediate dataset describing the fitted spline model. Finally, there exist a number of analysis programs to make use of the model, e.g. phencal to calculate phenologica