New JAXA High-Resolution Land Use/Land Cover Map for Vietnam Aiming for Natural Forest and Plantation Forest Monitoring

Abstract

Highly detailed and accurate forest maps are important for various applications including forest monitoring, forestry policy, climate change, and biodiversity loss. This study demonstrates a comprehensive and geographically transferable approach to produce a 12 category high-resolution land use/land cover (LULC) map over mainland Vietnam in 2016 by remote sensing data. The map included several natural forest categories (evergreen broadleaf, deciduous (mostly deciduous broadleaf), and coniferous (mostly evergreen coniferous)) and one category representing all popular plantation forests in Vietnam such as acacia (Acacia mangium, Acacia auriculiformis, Acacia hybrid), eucalyptus (Eucalyptus globulus), rubber (Hevea brasiliensis), and others. The approach combined the advantages of various sensor data by integrating their posterior probabilities resulting from applying a probabilistic classifier (comprised of kernel density estimation and Bayesian inference) to each datum individually. By using different synthetic aperture radar (SAR) images (PALSAR-2/ScanSAR, PALSAR-2 mosaic, Sentinel-1), optical images (Sentinel-2, Landsat-8) and topography data (AW3D30), the resultant map achieved 85.6% for the overall accuracy. The major forest classes including evergreen broadleaf forests and plantation forests had a user’s accuracy and producer’s accuracy ranging from 86.0% to 95.3%. Our map identified 9.55 × 106 ha (±0.16 × 106 ha) of natural forests and 3.89 × 106 ha (±0.11 × 106 ha) of plantation forests over mainland Vietnam, which were close to the Vietnamese government’s statistics (with differences of less than 8%). This study’s result provides a reliable input/reference to support forestry policy and land sciences in Vietnam.

Synergistic Use of Sentinel-1 and Sentinel-2 to Map Natural Forest and Acacia Plantation and Stand Ages in North-Central Vietnam

“…Similarly a study in Indonesia [18] found accuracies of 86% using Cband SAR in distinguishing oil palm and acacia plantation from natural forest. Similarly, for S-2 our overall accuracy of 92.5% is similar to previous studies, with 90% accuracy achieved in mapping acacia plantations in Indonesia [66] using optical SPOT satellite (10-20 m resolution) imagery, and 86-88% in Vietnam [23] distinguishing natural forest and plantation (acacia, rubber and eucalyptus combined) using optical and radar data.…”

“…These indices have been proved to be effective in earlier work classifying plantation and natural forest [23,43,44]. Further, we used a form of the radar vegetation index (RVI) [45] We used the results from the time-series (see Section 2.3.1 above) to select 2 vegetation indices: first, the Normalised Burn Ratio (NBR) vegetation index:…”

“…Therefore, these two sensors can usefully complement each other. A number of studies have used satellite imagery both optical [14,15] and radar [16][17][18] and combined optical and radar [19][20][21][22][23] to identify plantation areas, chiefly in Malaysia, Indonesia, and China and largely focused on palm oil, acacia and rubber plantations. Classifying acacia plantations has proved problematic in another study due to their short life-cycle [18].…”

Ensemble learning updating classifier for accurate land cover assessment in tropical cloudy areas

“…Several attempts have been made to monitor land cover in the entire VMD, where is covered by clouds up to 93% in peak seasons (Paul Lawson and Zuidema 2009;Li et al 2013;Jackson et al 2018). Truong et al and Hoang et al have successfully quantified comprehensive land cover in the delta, using the fusion of multiple remotely sensed sources (Truong et al 2019;Hoang et al 2020). However, these results contain relatively a few misclassifications among non-forest classes such as barren land and open water.…”

Rural land cover mapping using short-term Sentinel-2 temporal composites and machine learning

“…Implementing multispectral data, studies have long been conducted, especially in oil palm plantations 6 or rubber estates. 7 While multispectral is invaluable for rural mapping, problems of atmospheric disturbances such as cloud cover remain, especially when the changing climate is considered. The option of using synthetic aperture radar (SAR) has been investigated in an attempt to solve these problems.…”