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.

Measuring and monitoring forest canopy height are essential for effective forest management and ecosystem assessment. We evaluated the accuracy of forest canopy height based on Digital Surface Models (DSMs) derived from three different optical earth observation satellite stereo images: WorldView-3 along-track stereo pair images, BlackSky Gen-2 along-track 5-shot stereo images, and Newsat cross-track multiple stereo images. BlackSky Gen-2 5-shot stereo images achieved the highest accuracy on the forest canopy height measurement. Over coniferous forest areas with minimal seasonal height variation, the accuracy metrics were as follows: for WorldView-3, the standard deviation was 1.87 meters, Root Mean Square Error (RMSE) was 2.12 meters, and Normalized Median Absolute Difference (NMAD) was 0.97 meters; for BlackSky Gen-2, the standard deviation was 1.65 meters, RMSE was 1.79 meters, and NMAD was 1.24 meters; and for Newsat, the standard deviation was 1.59 meters, RMSE was 1.84 meters, and NMAD was 1.34 meters. The topographic azimuth affected the decreasing accuracy of the DSM derived from the stereo pair images of WorldView-3 on slopes facing away from the satellite's azimuth. The results highlighted the importance of acquiring stereo images from multiple suitable viewing angles to ensure accurate forest measurements in mountainous area. These findings demonstrate the applicability of DSMs derived from satellite stereo images, particularly multi-view stereo images from small satellites, for measuring forest canopy height.

The Greenhouse Gases Observing Satellite-2 (GOSAT-2), launched in October 2018, is equipped with the Cloud and Aerosol Imager-2 (CAI-2), which is used to estimate aerosol properties. In this study, the error characteristics of the estimated products were investigated through comparison with ground-based observations. Worldwide AERONET data from all available sites were primarily used to evaluate the optical properties of aerosol for the period March 2019 to December 2024. The comparison results for each year were similar, and most error statistics showed little or no trend. Consequently, the results from 2019 were considered to be representative and are presented here.

Comparing the satellite data with the AERONET data and calculating the correlation coefficients between them, the correlation for aerosol optical depth (AOD) was generally moderate, but the correlations for other parameters (Ångström exponent (AE_ext), absorption AOD (AOD_abs), absorption AE (AE_abs), single scattering albedo (SSA), PM2.5, and BC amount) were either absent or weak. For example, in a comparison between AOD at 550 nm with AERONET_AOD (all) (i.e., over land and ocean areas), the bias (BIAS), standard deviation (SD), and root mean squared error (RMSE), were 0.090, 0.149, and 0.174, respectively, and the correlation coefficient (R) was 0.593, indicating a moderate correlation. The satellite estimates tended to be overestimated when AOD was low and underestimated when AOD was high. In the comparison with AERONET_AOD (ocean) (i.e., over ocean areas), the BIAS, SD, and RMSE were 0.138, 0.071, and 0.155, respectively, and R was 0.880, indicating a strong correlation. Both the BIAS and R were higher than the values observed over land.

The results of the comparisons varied by dataset, region, target area, and time period. For ground-based AOD values below approximately 0.2, the satellite-derived AOD estimates were insensitive to changes in AOD, suggesting that low AOD values may not be reliably estimated. The estimated AE_ext and AE_abs values were limited to a narrow range, and the SSA values tended to cluster around specific values.

In recent years, organic farming has been increasingly promoted in Japan due to a shift toward sustainable agriculture as well as the rising cost of chemical fertilizers. However, the nitrogen efficacy of organic materials shows considerable seasonal variation, complicating the adjustment of fertilization rates. This has resulted in farm management issues such as delayed crop growth and insufficient harvest volumes to meet contractual delivery requirements for leafy vegetables. To address these challenges, it is essential to record crop growth trajectories and biomass accumulation, and to develop fertilization management techniques based on historical data. In this study, we developed and validated a non-destructive monitoring method to capture temporal changes in the plant height of komatsuna (Brassica rapa var. perviridis) cultivated under greenhouse conditions. Four types of close-range remote sensing devices were used: a ropeway-type mobile imaging system, depth cameras, 2D light detection and ranging (LiDAR), and time-of-flight (ToF) sensors. Structure-from-Motion/Multi-View Stereo (SfM/MVS) analysis using the ropeway-type system achieved the highest accuracy in estimating plant height (root mean square error [RMSE]: 1.4 cm), enabling the visualization of growth heterogeneity across the field. The depth cameras also showed high accuracy (RMSE: 2.45～2.75 cm). The ToF sensors were effective for nighttime monitoring (RMSE: 3.83 cm), while the 2D LiDAR provided cross-sectional profiles of canopy height but had the lowest accuracy (RMSE: 4.96 cm). Continuous day-night monitoring confirmed the presence of a physiological response, circadian leaf angle changes influenced by light conditions, and showed that plant height continued to increase during the nighttime. These growth characteristics of komatsuna may affect the accuracy of plant height estimation.

Synthetic Aperture Radar (SAR) satellites play a critical role in disaster response by providing all-weather, day-and-night observation capabilities. SAR is particularly valuable for flood disasters, which often occur under cloud cover and at night when optical satellites are ineffective. This paper presents the application of single-image flood inundation analysis using Synspective’s small SAR satellite “StriX” to three flood events that occurred in 2025: Kamiamakusa City in Kumamoto Prefecture, Japan (August 2025); Calbayog City in the Philippines (December 2025); and Hoi An in Vietnam (October 2025). The single-image inundation analysis method extracts inundated areas from a single SAR image without requiring pre-event archive data, making it particularly suitable for rapid disaster response and applications in regions with limited archive coverage. In Kamiamakusa, an emergency observation was conducted by StriX-3, and the entire workflow from tasking registration to delivery of the analysis report was completed within approximately three hours, with the extracted inundation areas showing distributions consistent with on-site live camera footage. In Calbayog, four consecutive observations were performed twice daily over two days, capturing the temporal evolution of inundation areas and demonstrating the value of high-frequency monitoring for evacuation and flood response decision-making. In Hoi An, the inundation map produced by StriX was compared with that from ESA’s Sentinel-1, confirming a generally consistent distribution while delivering the result the day after the event, independent of the 12-day revisit cycle. These cases demonstrate that the combination of single-image inundation analysis and the expanding StriX constellation enables rapid and reliable flood disaster response both domestically and internationally.

This study examined the relationships between burn severity levels classified from dNBR derived from Sentinel-2 MSI imagery and post-fire canopy height model (CHM) and canopy cover (CC) obtained from airborne LiDAR (LP) data, using the March 2025 wildfire in Minamiku, Okayama City, Japan (total burned area: 486 ha) as a case study. Post-fire airborne LiDAR data were acquired on March 29, 2025. CHM was calculated as the difference between the digital surface model (DSM) and digital elevation model (DEM) at 1 m resolution, and CC was derived by binarizing CHM at a 2 m threshold. Both CHM and CC were resampled to a 10 m grid to match the dNBR pixel size. dNBR was classified into five burn severity levels based on thresholds from Key and Benson (2006). Buildings and roads were excluded using mask polygons derived from base map data. Zonal statistics of CHM and CC were computed for each burn severity level. The results showed a clear decreasing trend in both CHM and CC with increasing burn severity: the Unburned level had a median CHM of 8.3 m and median CC of 1.00, while the Moderate-High level showed values of 3.6 m and 0.77, respectively. Spatial distributions of dNBR and CHM also showed visual correspondence across the study area. These findings indicate that dNBR broadly reflects post-fire changes in canopy height and cover derived from airborne LiDAR data, even in Japanese wildfire conditions.

Three Japan Forest Technology Association, case studies that, demonstrate the utility of remote sensing technology for monitoring forest ecosystems, mapping the site index distributions of forest plantations, and monitoring harvested areas are presented. The first case pertains to the application of Unmanned Aerial Vehicle (UAV) imagery in the Ogasawara Islands. This approach was used to assess ecosystem conservation initiatives and to conduct vegetation surveys in regions of the Islands that are otherwise inaccessible. This case’s quantitative assessment of the expansion of the native tree canopy and the effects of the control of invasive alien plants in Takinoura, Anijima was based in a visual interpretation and analysis of time-series orthomosaic images. High-resolution UAV images were used to (i) identify the distribution of rare and exotic plant species on the island’s inaccessible cliffs and landslide areas, and (ii) analyze the relationships between the vegetation distribution and topographic conditions. The second case study presents a methodology for generating site index distribution maps based on the integration of canopy-height values derived from airborne LiDAR with forest planning maps. This approach is shown to facilitates the visualization of disparities in forest productivity across different forest subcompartments. The case provides valuable information for identifying the optimal areas for harvesting and reforestation, as well as for enhancing forest planning. The third case study focuses on the estimation of harvest area in private forests across Japan, leveraging optical satellite imagery. In this method, harvested area maps were stratified, and the presence or absence of a harvested area at randomly allocated sample points in each stratum was visually interpreted. The visual interpretation results were then used to estimate harvest areas based on statistical analyses. To mitigate the effects of interpretation bias and input errors, each sample point was independently interpreted by at least three interpreters.

This case study examines the differences between ground subsidence measurements taken from space (interferometric SAR analysis results) and measurements taken on land (leveling survey results). The target area is the Nobi Plain, spanning Aichi, Gifu, and Mie prefectures, with an area of approximately 40 km north-south and 30 km east-west. This case study uses interferometric SAR analysis results from Sentinel-1 satellite data, interferometric SAR analysis results from ALOS-2 data loaned by the Geospatial Information Authority of Japan, and leveling survey observation results compiled by the Tokai Three Prefectures Ground Subsidence Survey Association. The verification period is five years, from 2016 to 2021. The verification results showed that the interferometric SAR analysis results of the SBAS method at 30 m from the Sentinel-1 satellite were in good agreement with the leveling survey results. In the future, it is desirable to conduct comparative verification of the interferometric SAR analysis results of the SBAS method at 30 m from the ALOS-2 satellite and the leveling survey results.

Ground deformation and road subsidence in urban areas are critical social concerns that require effective monitoring and infrastructure maintenance. Time-series Interferometric Synthetic Aperture Radar (InSAR) is a powerful technology capable of detecting continuous minute surface displacements over wide areas. However, it remains challenging to utilize the resulting multi-dimensional point cloud data in practical operations because conventional static drawings and two-dimensional geographic information system (GIS) environments cannot effectively display or share the integrated spatial and temporal information.

This paper introduces a practical approach to ground deformation monitoring and visualization. First, a permanent scatterer InSAR (PSInSAR) time-series analysis was conducted on the urban road subsidence incident that occurred in Chofu City, Tokyo, in October 2020. Utilizing Sentinel-1 C-band synthetic aperture radar (SAR) data (2019–2021) combined with a 10-meter digital elevation model (DEM), 2.5-dimensional (2.5-D) analysis was performed by integrating ascending and descending orbits to separate displacements into quasi-upward and quasi-eastward components. The analysis successfully captured a localized quasi-upward subsidence of up to approximately 15 mm and distinctive opposing quasi-eastward horizontal movements concentrated around the subsidence area from September 2020, demonstrating the high accuracy and spatial-temporal resolution of the method.

Second, to overcome information-sharing barriers among diverse stakeholders, this paper presents two intuitive visualization solutions: an online WebGIS environment utilizing ArcGIS Dashboards and a secure, offline client-side Web application built with Leaflet.js. The WebGIS dashboard integrates multi-dimensional datasets into a code-free interactive interface, allowing non-experts to instantly view displacement time-series graphs linked to specific geographic points. Meanwhile, the client-side Web application enables standalone high-security data analysis—including attribute-based color mapping, dynamic threshold screening, and on-the-fly linear regression—directly within standard web browsers without data leakage risks.

In conclusion, integrating advanced time-series InSAR analysis with interactive visualization platforms and high-quality geospatial data significantly enhances risk assessment, disaster mitigation, and infrastructure lifecycle management.

Following the 2021 landslide disaster in Atami, Shizuoka, national demand for embankment management grew, leading to the enforcement of the Embankment Regulation Act in 2023. However, surveying vast areas to identify embankment sites is costly in terms of personnel, time, and expense. This study developed an AI model combining JAXA’s ALOS-2 SAR data with deep learning to extract embankment locations. Using pre- and post-construction SAR intensity images, the model classifies each pixel as “embankment” or “non-embankment,” enabling stable observation even in cloudy or nighttime conditions. Validation in Tottori and Okayama achieved 60-70 % recall at a threshold of 0.1 for sites exceeding 3,000 m². Future work will develop a hybrid model integrating Sentinel-2 optical data to improve accuracy and generalizability.

Construction sites change rapidly and therefore require timely observation for effective management. Although unmanned aerial vehicles (UAVs) provide high-resolution data at flexible timings, their practical use is hindered by fragmented workflows and labor-intensive manual processing. This study developed an integrated Digital Twin Platform (DTPF) that involves an automated UAV system to streamline the entire workflow from field measurement, photogrammetry, point cloud filtering, and data integration. The system incorporates schedule-linked data management, automatic ground control point matching, and purpose-specific filtering. When our proposed system was applied to an earthwork site over a period of three months, measurement tasks for progress management and site confirmation were reduced by up to 80% compared to conventional methods, allowing managers to reallocate time to more critical activities such as construction planning and safety management. These results demonstrate that the DTPF successfully addresses operational efficiency and labor reduction demands in construction. It may also have applications in agriculture and forestry.

Indonesia has the sixth largest Exclusive Economic Zone (EEZ) in the world. Although fisheries offshore of the Indonesian EEZ are active and produce high levels of products, several illegal, unreported and unregulated (IUU) fishing activities are also recognized in this region. IUU fishing activities are monitored and reported by the Ministry of Marine Affairs and Fisheries of Indonesia (Kementerian Kelautan dan Perikanan in Indonesian, referred to as the ‘KKP’) and its institute, i.e., the Institute of Information Management of Marine and Fisheries Resources (Balai Pengelolaan Informasi Sumber Daya Kelautan dan Perikanan, the ‘BPISDKP’) with the use of satellite data. PALSAR-2, the Japanese synthetic aperture radar (SAR) sensor on the Japanese satellite ALOS-2, has a ScanSAR Normal Mode (PALSAR-2/ScanSAR) which has its observation area as 350.5 km×355 km, and PALSAR-2/ScanSAR has been found to be effective for monitoring wide offshore areas of Indonesia. However, the average size of fishing vessels is <20 m, which is smaller than the 50-m resolution of PALSAR-2/ScanSAR. The present study analyzed and confirmed the abilities of PALSAR-2/ScanSAR for IUU fishing detection. Authors brought GNSS receivers and took a 17.6-m-long vessel to the northern sea of West Bali, Indonesia on the date of PALSAR-2/ScanSAR observation. Using the observation from the GNSS signal collection on the vessel as the ground truth and the vessel observation by PALSAR-2 sensor, authors also analyzed the data product and the GNSS signals to detect the vessel. The results revealed that the fishing vessel was detected as a single pixel. Our findings confirmed that (i) PALSAR-2/ScanSAR has useful and effective abilities for monitoring small-scale fishing activities offshore, and (ii) PALSAR-2/ScanSAR can be used for practical monitoring operations in Indonesian marine waters. This study thus provides one example ensuring the abilities of PALSAR-2/ScanSAR, and authors note that improvements of the accuracy of PALSAR-2/ScanSAR could be achieved by conducting observations under various environmental conditions in order to enhance its capacities for monitoring IUU fishing activities in Indonesian oceans.

This paper examines the integration of satellite remote sensing in designing triggers for parametric insurance to bolster global disaster resilience. Unlike traditional indemnity insurance, parametric insurance utilizes predefined indices for rapid payouts, bypassing lengthy damage assessments. While conventional triggers rely on limited ground-based sensors, satellite data provides objective, high-frequency, and wide-area observations. The study reviews international applications in agriculture and proposes three novel scenarios: compensating snowfall deficits for ski resorts, monitoring maritime activity in the Strait of Hormuz via AIS data to mitigate energy price spikes, and leveraging real-time thermal data for immediate wildfire response. By utilizing AI to reduce development costs and precise trigger design to minimize basis risk, satellite-based insurance addresses labor shortages and offers a transparent, scalable solution for climate-related financial risks.

Companies dealing with forest-derived products-such as palm oil, paper/pulp, cocoa, and coffee-face an urgent management issue: ensuring sustainable raw material procurement. This is driven by increasingly stringent regulations like the EU Deforestation Regulation (EUDR) and disclosure frameworks such as the Task Force on Nature-related Financial Disclosures (TNFD). However, tracing vast supply chains back to the farm level poses significant practical and technical challenges. This paper introduces “MIMAMORI GLOBAL,” a service that leverages satellite imagery and geospatial analysis to visualize and monitor environmental risks, including deforestation and biodiversity loss. The discussion focuses on the technical overview of the service, particularly the spatial analysis logic developed to estimate potential procurement areas (Sourcing Areas) even in cases where perfect traceability is lacking.

This paper introduces Deep Forestry's below-canopy autonomous drone and cloud-based AI software. It focuses on the product concept, workflow, expected outputs, and potential applications in Japan. The drone is intended to acquire three-dimensional information from within forests, while the AI software transforms captured data into usable information for survey, management, and future digital forest platforms.