Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales
Meng Yang,Yanting Zhou,Yong Xie,Wen Shao,Fengyu Luo
DOI: https://doi.org/10.1109/jstars.2024.3385998
IF: 4.715
2024-04-24
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:Numerous spatiotemporal fusion (STF) methods have been developed to generate surface reflectance data with high spatial and temporal resolutions for dynamic monitoring. Although comparative studies have been conducted to assess various fusion methods, selecting the most suitable fusion method for acquiring long-term time series data remains a challenge. This article compared four representative STF methods based on the effect of 8 × n-day (n = 1, 2, ..., 7) time scales between base and predicted data. These methods included the spatial and temporal adaptive reflectance fusion model (STARFM), flexible spatiotemporal data fusion (FSDAF), enhanced STARFM (ESTARFM), and sensor-bias driven spatio-temporal fusion model (BiaSTF). Accuracy was assessed using metrics such as the root-mean-square error, correlation coefficients, erreur relative globale adimensionnelle de synthèse, and spectral angle mapper. The results indicate that as the time scale increases, fusion accuracy decreases, with a significant drop observed at the 40-day mark. Compared with the 8-day scale, at the 40-day scale, the ERGAS of STARFM decreased by 20.66%, that of FSDAF by 17.00%, that of ESTARFM by 14.37%, and that of BiaSTF by 11.48%. Furthermore, STF methods based on two pairs of images demonstrate a notable advantage in capturing data from distant temporal phases. In regions with pronounced phenological changes and longer time scales, BiaSTF consistently exhibits the best fusion performance (ERGAS = 1.67), followed by ESTARFM (ERGAS = 1.85). These findings can aid in determining the most suitable STF methods and provide guidelines for the development of new methods.
imaging science & photographic technology,remote sensing,engineering, electrical & electronic,geography, physical