Synergy of Active and Passive Remote Sensing Data for Effective Mapping of Oil Palm Plantation in Malaysia

Abstract

Oil palm is recognized as a golden crop, as it produces the highest oil yield among oil seed crops. Malaysia is the world’s second largest producer of palm oil; 16% of its land is planted with oil palm. To cope with the ever-increasing global demand on edible oil, additional areas of oil palm are forecast to increase globally by 12 to 19 Mha by 2050. Multisensor remote sensing plays an important role in providing relevant, timely, and accurate information that can be developed into a plantation monitoring system to optimize production and sustainability. The aim of this study was to simultaneously exploit the synthetic aperture radar ALOS PALSAR 2, a form of microwave remote sensing, in combination with visible (red) data from Landsat Thematic Mapper to obtain a holistic view of a plantation. A manipulation of the horizontal–horizontal (HH) and horizontal–vertical (HV) polarizations of ALOS PALSAR data detected oil palm trees and water bodies, while the red spectra L-band from Landsat data (optical) could effectively identify built up areas and vertical–horizontal (VH) polarization from Sentinel C-band data detected bare land. These techniques produced an oil palm area classification with overall accuracies of 98.36% and 0.78 kappa coefficient for Peninsular Malaysia. The total oil palm area in Peninsular Malaysia was estimated to be about 3.48% higher than the value reported by the Malaysian Palm Oil Board. The over estimation may be due the MPOB’s statistics that do not include unregistered small holder oil palm plantations. In this study, we were able to discriminate most of the rubber areas.

Sensitivity Analysis of Sentinel-1 Backscatter to Oil Palm Plantations at Pluriannual Scale: A Case Study in Gabon, Africa

“…However, the mean VH power reported in [10] for the oil palm area is −15.75 dB, which is too low and very similar to the backscattering expected from a bare surface or a field with dominant direct scattering from soil, where low and sparse vegetation is planted. This last explanation is compatible with our observations and analysis performed in the present paper (as well as in agreement with the findings in [12]). It must be pointed out that potential differences in density plantation, morphology or growth stage explain the disagreements observed among different studies.…”

“…However, the first and last images (i.e., 2015 and 2021, respectively) show a slightly noticeable variability, contributing to enlarge the dynamic range to 1-1.5 dB, being the variations of environmental conditions a possible explanation. It is noted, however, that these backscattering levels exhibit very similar values to those reported in a number of previous works at C-band [8,12,29,30]. These values are also higher than from the low and sparse vegetation characterising savannah areas, especially for VV channel.…”

“…Although we fully agree on the general idea of the utility of Sentinel-1 data, a successful use of Sentinel-1 data would require the combination of both VV and VH polarisation channels, as shown in [8] for a one-year period time series. The combination of optical and radar sensors has been also addressed more recently [12] where ALOS-2 PALSAR-2, Sentinel-1 and Landsat TM images are employed. Several images covering 2016 were used and high classification accuracy was obtained.…”

Perspective: Improving the accuracy of plant phenology observations and land-cover and land-use detection by optical satellite remote-sensing in the Asian tropics

“…9 Such future developments in optical sensors on board satellites with high observation frequency and high spatial resolution will remarkably improve the accuracy of phenology observations in tropical rain forests, where the phenological timing and patterns differ among the numerous and highly diverse tree species (Osada, 2018;Reich et al, 2004). (Ichikawa, 2007;Fitzherbert et al, 2008;GEAS, 2011;Koh et al, 2011;Hansen et al, 2013;Carlson et al, 2014;Nagai et al, 2014a;Estoque et al, 2019;Najib et al, 2020). Here, we defined deforestation as having occurred at points where the ratio of number of days observed GRVI < 0 under clear sky conditions to total observed GRVI under clear sky conditions was above 80% (Nagai et al, 2014a).…”

“…However, SAR cannot observe plant phenology, which is mainly shown as a characteristic of color change of canopy surface on satellite remote-sensing in optical signals. Thus, advancement in SAR technology and/or the integration of SAR and optical sensors will be needed for the accurate detection of land cover and land use (Najib et al, 2020).…”