Back to search

Boston University

Monitoring tropical forests using the fusion of multi-sensor time series

Abstract

dc:description.abstract

Tropical moist forests provide crucial ecosystem services. However, monitoring tropical forests is challenging in many regions of ecological importance due to persistent cloud cover and the lack of receiving station coverage for the Landsat satellites. This dissertation explores the use of multi-sensor fusion methods that combine optical and radar time series for monitoring tropical forests in West Africa. Southern Ghana is one of the cloudiest regions of the world. A multi-sensor monitoring algorithm was developed to provide maps and area estimates of forest disturbance from 2000 to 2022. In particular, plantation management was separated from native forest disturbances. The results reveal that 11,121 km2 (± 1,827 km2) of native forest were disturbed during this period, equivalent to 52.4% ± 8.6% of the forest cover in 2000. The results indicate that previous reports overestimated the loss of native forest due to the lack of information on plantations. Near-Real-Time (NRT) alert systems are powerful tools to reduce illegal forest activities. Using the Fusion Near Real-Time (FNRT) algorithm in the Protected Areas (PAs) of Ghana reveals alarmingly that 704 (±177) km2 of native forest were lost in 2023 and 2024, representing a 3% annual disturbance rate. Comparison of four tropical NRT alert systems, including GLAD-L, RADD, DIST-ALERT, and FNRT, shows remarkably little agreement; the highest level of agreement between any two datasets was less than 30%. Accuracy assessment based on reference samples indicates that these four products have very different patterns of accuracy. Interestingly, alerts in some products are highly reliable, yet many forest disturbances are often missed. The methods developed here can be used to monitor tropical forests over large scales and long time periods, particularly in cloudy and complicated landscapes. The results contribute to a better understanding of tropical forest dynamics in West Africa and can support improved forest management and decision-making. Overall, this research demonstrates the great potential of multi-sensor fusion for monitoring tropical forests.

Degree

thesis:*
Name thesis:degree_name
Doctor of Philosophy
Level thesis:degree_level
doctoral
Discipline thesis:degree_discipline
Earth & Environment
Grantor
Boston University
Year dc:date.issued
2026

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Dong, Luofan
Advisors dc:contributor.advisor
  • Woodcock, Curtis E.
  • Tang, Xiaojing
ORCID dc:identifier.orcid

Subjects

dc:subject × 5

Rights

dc:rights
Statement dc:rights
  • Attribution 4.0 International
Language dc:language.iso
en_US

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/2144/53770

Chain of custody

source
Harvested from
Boston University
Base URL
open.bu.edu/oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Dong, Luofan. Monitoring tropical forests using the fusion of multi-sensor time series. doctoral thesis, Boston University, 2026. https://hdl.handle.net/2144/53770