Multi-Type Forest Change Detection Using BFAST and Monthly Landsat Time Series for Monitoring Spatiotemporal Dynamics of Forests in Subtropical Wetland

Abstract

Land cover changes, especially excessive economic forest plantations, have significantly threatened the ecological security of West Dongting Lake wetland in China. This work aimed to investigate the spatiotemporal dynamics of forests in the West Dongting Lake region from 2000 to 2018 using a reconstructed monthly Landsat NDVI time series. The multi-type forest changes, including conversion from forest to another land cover category, conversion from another land cover category to forest, and conversion from forest to forest (such as flooding and replantation post-deforestation), and land cover categories before and after change were effectively detected by integrating Breaks For Additive Seasonal and Trend (BFAST) and random forest algorithms with the monthly NDVI time series, with an overall accuracy of 87.8%. On the basis of focusing on all the forest regions extracted through creating a forest mask for each image in time series and merging these to produce an ‘anytime’ forest mask, the spatiotemporal dynamics of forest were analyzed on the basis of the acquired information of multi-type forest changes and classification. The forests are principally distributed in the core zone of West Donting Lake surrounding the water body and the southwestern mountains. The forest changes in the core zone and low elevation region are prevalent and frequent. The variation of forest areas in West Dongting Lake experienced three steps: rapid expansion of forest plantation from 2000 to 2005, relatively steady from 2006 to 2011, and continuous decline since 2011, mainly caused by anthropogenic factors, such as government policies and economic profits. This study demonstrated the applicability of the integrated BFAST method to detect multi-type forest changes by using dense Landsat time series in the subtropical wetland ecosystem with low data availability.

Exploring BFAST to detect forest changes in Portugal

“…The map of Figure 7(a) was reclassified to represent two classes, "change" and "no-change" (Figure 8). The accuracy assessment estimates an overall accuracy of 0.85±0.02 at the 95% confidence level, which is similar to some previous research 3 . Omission and commission errors of class "change" are 0.30±0.02 and 0.19±0.02.…”

“…This paper describes results obtained with Landsat 8 time series and Breaks For Additive Season and Trend (BFAST) 1 in a region of central Portugal. BFAST has been used to analyze Normalized Difference Vegetation Index (NDVI) time series and other vegetation indices to detect deforestation in a range of biomes 2,3 , but rarely in Mediterranean and Atlantic biomes (but see 4 ). BFAST decomposes a time series into trend and seasonal components with methods for detecting abrupt changes within the two components.…”

Deep Learning for Regular Change Detection in Ukrainian Forest Ecosystem With Sentinel-2

“…With remote sensing, deforestation monitoring systems use a bunch of automatic methods, based on the change detection techniques [12,21,26] e.g. time-series analysis of vegetation indices [38], image preprocessing to detect changes between a pair of images [49], machine learning classification [24,57]. But recently, the deep learning has demonstrated great potential in remote sensing, due to its ability to extract features from the spectral-spatial-temporal image data [40], providing state-of-the-art results in remote sensing change detection.…”

Spatial Agreement among Vegetation Disturbance Maps in Tropical Domains Using Landsat Time Series

“…Higher NDVI values suggest higher amounts of photosynthetic active biomass. As the most popular index, NDVI time series have been used in many forest disturbance detection and monitoring efforts [38,39].…”

“…However, the high accuracies presented by NDVI is somewhat surprising given the fact that studies showed low accuracies related to this index in tropical forests [13,47]. This is a particularly useful result indicating why NDVI is still the most frequently used index in remote sensing, and has been presented in forest disturbances detection and monitoring [38,39].…”