An optimization of parameter settings in HANTS for global NDVI time series reconstruction

The applications of Multi-temporal Normalized Difference Vegetation Index (NDVI), a critical proxy for analysing vegetation dynamics, have been long hindered by prevalent noise. The Harmonic ANalysis of Time Series (HANTS) has been widely applied to reconstruct noise- and cloud-free NDVI time series data set from regional to global scales. The reconstruction performance of HANTS is severely dependent on the inherent parameter settings in HANTS. However, most applications set the parameters of HANTS based on users' experience. This study analysed the sensitivity of the reconstruction performance to several critical parameters and weighting scheme of HANTS. The results document the optimized parameter settings for global NDVI time series reconstruction.