Journal of The Remote Sensing Society of Japan
Online ISSN : 1883-1184
Print ISSN : 0289-7911
ISSN-L : 0289-7911
Regular Paper
Improvement of JAXA’S High-Resolution Land Use and Land Cover Map of Japan with Species-Level Orchard Categories
Shun Ito, Yuki Mizuno, Shoki Shimada, Sota Hirayama, Takeo Tadono, Koichi Imamura, Naoyoshi Hirade, Fumi Ohgushi, Ryo Satake, Kenlo Nishida Nasahara
Author information
JOURNAL FREE ACCESS

2026 Volume 46 Issue 3 Pages 159-175

Details
Abstract

Land use and land cover (LULC) maps are crucial for sustainable agricultural management, environmental conservation, etc. While several nationwide LULC maps for Japan have been developed, JAXA HRLULC-Japan is notable for its high resolution and a categorization system that reflects Japan’s unique landscape. However, despite these strengths, its latest version (2024JPN_v25.04) lacks detailed information on orchards. This limitation has become increasingly significant as the demand for such spatial data grows across various societal contexts. Specifically, mapping species such as persimmon and chestnut—which are highly attractive to bears—is critical for mitigating human-wildlife conflicts. Furthermore, spatial data for perennial fruit trees is essential for climate change adaptation, given the long-term nature of their cultivation and their inherent immobility. To address these needs, this study developed a nationwide, species-level orchard map for Japan by enhancing JAXA’s framework through a hierarchical classification approach, expanding the system to 27 categories. The 1st level classification established a 20-category base map including “Deciduous orchard” and “Evergreen orchard”. The 2nd level classification further subdivided the “Deciduous orchard” category into 8 major species: “Apple”, “Pear”, “Persimmon”, “Peach”, “Cherry”, “Japanese apricot”, “Grape”, and “Chestnut”. Using multi-temporal Sentinel-2/MSI data and ALOS-2/PALSAR-2 data from 2024 with the classifier “SACLASS2.5” developed by JAXA, the 1st level classification achieved an overall accuracy (OA) of 93.2±0.5 %, and the 2nd level classification attained an OA of 85.2±3.5 %. This research provides the first high-resolution nationwide LULC map with species-specific orchard categories, offering a vital resource for wildlife management, climate-resilient agriculture, etc.

Content from these authors
© 2026 The Remote Sensing Society of Japan
Previous article Next article
feedback
Top