Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales
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 ac...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2024-01-01
|
Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/10494345/ |
_version_ | 1797197331039256576 |
---|---|
author | Meng Yang Yanting Zhou Yong Xie Wen Shao Fengyu Luo |
author_facet | Meng Yang Yanting Zhou Yong Xie Wen Shao Fengyu Luo |
author_sort | Meng Yang |
collection | DOAJ |
description | 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 × <italic>n</italic>-day (<italic>n</italic> = 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. |
first_indexed | 2024-04-24T06:42:16Z |
format | Article |
id | doaj.art-74831192e114446498bb7b8008b95e1e |
institution | Directory Open Access Journal |
issn | 1939-1404 2151-1535 |
language | English |
last_indexed | 2024-04-24T06:42:16Z |
publishDate | 2024-01-01 |
publisher | IEEE |
record_format | Article |
series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
spelling | doaj.art-74831192e114446498bb7b8008b95e1e2024-04-22T23:00:11ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing1939-14042151-15352024-01-01178291830110.1109/JSTARS.2024.338599810494345Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time ScalesMeng Yang0https://orcid.org/0009-0002-0065-1607Yanting Zhou1https://orcid.org/0009-0001-6653-4369Yong Xie2https://orcid.org/0000-0002-7863-7170Wen Shao3https://orcid.org/0000-0003-1203-0883Fengyu Luo4https://orcid.org/0009-0004-8427-2536School of Geographical Sciences, Nanjing University of Information Science and Technoloy, Nanjing, ChinaSchool of Geographical Sciences, Nanjing University of Information Science and Technoloy, Nanjing, ChinaSchool of Geographical Sciences, Nanjing University of Information Science and Technoloy, Nanjing, ChinaSchool of Geographical Sciences, Nanjing University of Information Science and Technoloy, Nanjing, ChinaSchool of Geographical Sciences, Nanjing University of Information Science and Technoloy, Nanjing, ChinaNumerous 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 × <italic>n</italic>-day (<italic>n</italic> = 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.https://ieeexplore.ieee.org/document/10494345/Accuracy evaluationhigh spatiotemporal surface reflectance (SR)spatiotemporal fusion (STF)time series data |
spellingShingle | Meng Yang Yanting Zhou Yong Xie Wen Shao Fengyu Luo Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Accuracy evaluation high spatiotemporal surface reflectance (SR) spatiotemporal fusion (STF) time series data |
title | Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales |
title_full | Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales |
title_fullStr | Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales |
title_full_unstemmed | Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales |
title_short | Accuracy Evaluation of Four Spatiotemporal Fusion Methods for Different Time Scales |
title_sort | accuracy evaluation of four spatiotemporal fusion methods for different time scales |
topic | Accuracy evaluation high spatiotemporal surface reflectance (SR) spatiotemporal fusion (STF) time series data |
url | https://ieeexplore.ieee.org/document/10494345/ |
work_keys_str_mv | AT mengyang accuracyevaluationoffourspatiotemporalfusionmethodsfordifferenttimescales AT yantingzhou accuracyevaluationoffourspatiotemporalfusionmethodsfordifferenttimescales AT yongxie accuracyevaluationoffourspatiotemporalfusionmethodsfordifferenttimescales AT wenshao accuracyevaluationoffourspatiotemporalfusionmethodsfordifferenttimescales AT fengyuluo accuracyevaluationoffourspatiotemporalfusionmethodsfordifferenttimescales |