Prof.  Long  Li
China University of Mining and Technology,  China

Title: Applying Remote Sensing Multitemporal Analysis with Caution

Abstract:

Thanks to its capability of repetitive observations, remote sensing has been extensively utilized for detecting and mapping the evolution of the Earth’s surface over time. Multitemporal analysis (MTA), where several remote sensing images over the same geographic area but acquired on different dates are compared, has constantly been performed in various applications, for example, examining land use/cover change, and monitoring urban heat islands. It is often that such analysis was built on only one image (or scene) for selected months or years. A question arises whether such practice is appropriate in all the remote sensing applications that compare multitemporal images.

 

To address this question, I revisited previous studies, and assessed the reliability and applicability of the above-mentioned practice in remote sensing studies. Results show that it is suitable for land use/cover, which changes annually, but inappropriate for biophysical variables, e.g., land surface temperature, that change on an hourly or a daily basis. It would be imprudent to use images acquired on a day to represent an entire month, season, or year to reveal the monthly, seasonally or yearly variability of the constantly changing variables. This indicates different findings and conclusions would be obtained if more remote sensing images were included and compared in many previous studies on urban heat islands.

 

It is concluded that performing multitemporal analysis using remote sensing image data often requires the consideration of how frequently remotely sensed variables change over time. If insufficient or unrepresentative images are compared for change detection in remote sensing MTA, caution should be exerted when making conclusions from the detected changes.

 

Keywords: remote sensing, multitemporal analysis, land use/cover, urban heat islands, surface temperature

Biography:

Prof. Dr. Long Li is an associate professor at the Department of Land Resources Management, China University of Mining and Technology. His main research interests include remote sensing of land resources, land-use planning, and natural risk assessment. Dr. Li obtained his PhD degree in geography at the Vrije Universiteit Brussel, Belgium in 2016. He has previously worked at the University of Portsmouth, the UK as a visiting researcher and the Vrije Universiteit Brussel as a geographic researcher. He has published more than 40 papers in peer-reviewed journals, addressing various environmental challenges in the urban and natural contexts using geographic information systems and remote sensing. He has delivered oral and poster presentations at numerous international and national conferences in Europe, Asia, Australia, and Africa. He is a Reviewer Board member of two MDPI journals, /Remote Sensing/ and /Land/, and has served as an anonymous reviewer for over 40 reputed journals. In addition, Dr. Li is currently guest-editing two special issues, Environmental Development: Monitoring, Assessment, and Adaptation for the journal /Sustainability/, and Remote Sensing for Sustainable Land Use and Management for the journal /Frontiers in Remote Sensing/.