Automated Plantation Mapping in Southeast Asia Using MODIS Data and Imperfect Visual Annotations

Abstract

Expansion of large-scale tree plantations for commodity crop and timber production is a leading cause of tropical deforestation. While automated detection of plantations across large spatial scales and with high temporal resolution is critical to inform policies to reduce deforestation, such mapping is technically challenging. Thus, most available plantation maps rely on visual inspection of imagery, and many of them are limited to small areas for specific years. Here, we present an automated approach, which we call Plantation Analysis by Learning from Multiple Classes (PALM), for mapping plantations on an annual basis using satellite remote sensing data. Due to the heterogeneity of land cover classes, PALM utilizes ensemble learning to simultaneously incorporate training samples from multiple land cover classes over different years. After the ensemble learning, we further improve the performance by post-processing using a Hidden Markov Model. We implement the proposed automated approach using MODIS data in Sumatra and Indonesian Borneo (Kalimantan). To validate the classification, we compare plantations detected using our approach with existing datasets developed through visual interpretation. Based on random sampling and comparison with high-resolution images, the user’s accuracy and producer’s accuracy of our generated map are around 85% and 80% in our study region.

Plantation Mapping in Southeast Asia

“…While the visual validation of generated maps is beyond the scope of this paper, we used a sampling-based approach for a more detailed examination of locations discussed in the above three scenarios and measured the accuracy of the proposed method and existing plantation products. The results were discussed in our previous report (Jia et al, 2016).…”

Toward Large-Scale Mapping of Tree Crops with High-Resolution Satellite Imagery and Deep Learning Algorithms: A Case Study of Olive Orchards in Morocco

“…Timely and accurate monitoring of the extent and productivities of tree crops are necessary for enhancing their management [16]. With the increasing accessibility of remote sensing data and the progress in algorithms, there is a surge of studies focusing on mapping different tree crops, e.g., oil palm [17][18][19][20][21][22], rubber [23][24][25], and coffee [26][27][28][29], with multisource satellite imagery. Despite the varying level of success, these studies mainly fall into two categories.…”

“…Comparatively, tree species with obvious spatial or spectral features on medium to coarse resolution imagery are much better studied than the small crown trees. For instance, in southeast Asia, mature oil palm trees can have a crown diameter around 10 meters and are often cultivated regularly in plantations that span a few kilometers, thus even coarseresolution imagery, e.g., MODIS (250 m), can be used to map oil palm plantations [17,22]. In contrast, olive trees in the Mediterranean usually have crown sizes less than 5 m and are more sparsely planted in orchards, therefore are hardly detectable from MODIS imagery or even moderate resolution products like Landsat and Sentinel-2.…”