HANTS

Researchers at the Cdentre for Geoinformation in the Netherlands needed to process NDVI time-series from the SPOT-VGT sensor, using ENVI, over a period of five years to investigate phenology differences between landcover types, and how weather conditions affected the differences.

Researchers used IDL and ENVI together to implement the HANTS algorithm in conducting their research. The Hants algorithm filters the time-series based on fourier transforms. The team used IDL, as it an easy solution for implementing the iterative, FFT (Fast Fourier Transform)-based filter. Using the filter allowed them to filter the clouded points in the NDVI time-series. After filters were applied, ENVI was used to implement intuitive visualization options, such as image tiling. Researchers at the Centre chose ENVI for its flexibility and capability of handling diverse data types, "particularly including complex values and plot Z-profiles were very practical," said Allard de Wit of the Centre, "The result of the HANTS algorithm is a efficient way of cleaning up the NDVI-timeseries which could be used successfully to derive the start-of-growing season for different land cover types and years in the Netherlands."