High-resolution global map of closed-canopy coconut palm

Abstract

Demand for coconut is expected to rise, but the global distribution of coconut palm has been studied little, which hinders the discussion of its impacts. Here, we produced the first 20 m global coconut palm layer using a U-Net model that was trained on annual Sentinel-1 and Sentinel-2 composites for the year 2020. The overall accuracy was 99.04 ± 0.21 %, which was significantly higher than the no-information rate. The producer's accuracy for coconut palm was 71.51 ± 23.11 % when only closed-canopy coconut palm was considered in the validation, but this decreased to 11.30 ± 2.33 % when sparse and dense open-canopy coconut palm was also taken into account. This indicates that sparse and dense open-canopy coconut palm remains difficult to map with accuracy. We report a global coconut palm area of 12.66 ± 3.96 × 106 ha for dense open- and closed-canopy coconut palm, but the estimate is 3 times larger (38.93 ± 7.89 × 106 ha) when sparse coconut palm is included in the area estimation. The large area of sparse coconut palm is important as it indicates that production increases can likely be achieved on the existing lands allocated to coconut. The Philippines, Indonesia, and India account for most of the global coconut palm area, representing approximately 82 % of the total mapped area. Our study provides the high-resolution, quantitative, and precise data necessary for assessing the relationships between coconut production and the synergies and trade-offs between various sustainable development goal indicators. The global coconut palm layer is available at https://doi.org/10.5281/zenodo.8128183 (Descals, 2023).