Conference Program

Prepare your trip to the 2026 Pacific Island GIS and Remote Sensing User Conference!

Save the dates:

Last update : 05 March 2026 – 09:30 am

  • Program Overview
  • Day 1 – 30.11.2026
  • Day 2 – 01.12.2026
  • Day 3 – 02.12.2026
  • Day 4- 03.12.2026
  • Day 5 – 04.12.2026
2026-11-30
10:25 – 11:45
session 1.b
Institutionnal
12:45 – 14:05
session 4.b
Institutionnal
14:30 – 14:50
session 1.d
New Technologies (AI, Deep Learning, Cloud)
2026-12-01
08:00 – 09:20
session 2.a
Maritime Environment
09:45 – 10:45
session 2.b
Disaster
10:45 – 11:05
session 4.a
Disaster
12:05 – 13:25
session 2.c
GIS software and technics
13:50 – 14:00
lgt 2.d
General
2026-12-02
08:00 – 09:20
session 3.a
Remote Sensing data and software
09:45 – 11:05
session 3.b
General
12:05 – 13:25
session 3.c
General
13:50 – 15:10
session 3.d
General
16:00 – 18:00
PGRSC AGM
General
2026-12-03
08:00 – 08:20
session 2.b
Disaster
08:20 – 09:20
session 4.a
Disaster
09:45 – 11:05
session 1.c
Mapping
12:05 – 13:25
session 4.c
New Technologies (AI, Deep Learning, Cloud)
13:50 – 15:10
session 4.d
General
2026-12-04
08:00 – 10:00
session 5.a
Workshop
10:15 – 12:15
session 5.b
Land Environment
13:15 – 15:15
session 5.c
Workshop
2026-11-30
07:00 – 09:00
Registration
09:00 – 10:00
Conference opening
10:00 – 10:25
Coffee break
10:25 – 10:45
Bradley Eichelberger
The Pacific GIS and Remote Sensing Council (PGRSC)
Institutionnal
The Pacific GIS & Remote Sensing Council (PGRSC) is a non-profit organisation duly established under the processes within the Charitable Trust Act [Cap 67] of the Republic of Fiji. PGRSC was established in 2018 to oversee the activities of the Pacific GIS and Remote Sensing User Conference and promote the use of Geographic Information Systems (GIS) and Remote Sensing (RS) technologies in the Pacific Islands Region. To date, PGRSC consists of over 560 individual Members from 48 countries (18 PICT) and over 60 Institutional Members, including government agencies, international and regional organisations, research institutions, universities, and private companies. The presentation will outline the Council\'s governance structure and its strategy for promoting capacity building and knowledge sharing throughout the region. Successes will be shared along with progress towards strategic goals and planned actions for 2027 that will support PGRSC Member benefits and geospatial needs in the region.
10:45 – 11:05
Andiswa Mlisa | Bradley Eichelberger
From partnership to action: Strengthening Pacific geospatial networks through SPC–PGRSC collaboration
Institutionnal
Strong geospatial networks connect regional capabilities with national priorities, expertise and practical applications. The partnership between the Pacific Community (SPC) and the Pacific GIS and Remote Sensing Council (PGRSC), formalised through a Memorandum of Understanding signed in August 2025, provides a framework for coordinated outreach, capacity development, knowledge sharing and strengthened regional and international connections. Through Digital Earth Pacific, SPC contributes Earth observation data, tools and technical support, complemented by PGRSC’s community-convening role and national focal points network. This presentation examines how the partnership is being translated into a shared programme of action. At the sixth Digital Earth Pacific Steering Committee meeting, held jointly with PGRSC national focal points in Nadi, Fiji, on 11–13 February 2026, participants endorsed a joint workplan aligned with country needs. Priorities include webinars showcasing Digital Earth Pacific products and applications, a focal point mapathon to support product validation, development and communication of country use cases, exploration of university-based knowledge sharing, youth engagement, and support for the annual Pacific Islands GIS and Remote Sensing User Conference. The presentation will report progress against this action plan and discuss lessons emerging from implementation, including the role of national focal points in connecting users, communicating country priorities and supporting feedback on the relevance and usability of regional products. It will consider opportunities to strengthen coordination, peer learning and sustained engagement between regional technical teams and national GIS and remote sensing communities. By sharing the partnership’s approach and implementation experience, the presentation contributes to the conference theme of Strengthening and Expanding Geospatial Networks in the Pacific Islands, highlighting how institutional collaboration can support national ownership and practical use of Earth observation information.
11:05 – 11:25
Wolf Forstreuter
Memories of GIS and Remote Sensing Development in Pacific Island Countries: The 1990s in Fiji
Institutionnal

Memories of GIS and Remote Sensing Development in Pacific Island Countries: The 1990s in Fiji

Wolf Forstreuter, Pacific Wolf Consult

Abstract

The development of geographical information systems (GIS) and remote sensing (RS) in the Pacific Islands during the 1990s was shaped by technological innovation, international development assistance, institutional cooperation and the commitment of a relatively small community of specialists. This paper documents personal experiences and observations of these developments, focusing on Fiji as an early centre for introducing digital satellite image processing and establishing regional GIS and remote sensing networks. Although it cannot cover every important initiative, it records milestones that contributed to the development of GIS and remote sensing capabilities across the Pacific Islands.

Figure 01: From left to right: Head of SOPAC IT, Les Allinson; Head of the Forest Management Services Division, Osea Tuinivanua; GTZ inventory project leader, Wolf Forstreuter.

A major milestone was Fiji’s national forest inventory, initiated in 1991 with financial support from German development aid. The project aimed to map the forest cover of Fiji’s seven main islands using digital satellite imagery. Most Pacific Island Countries are outside the coverage of satellite data reception stations. Through the involvement of EOSAT and the Tracking and Data Relay Satellite (TDRS) system, Landsat IV imagery could be transmitted to a ground station in the United States. After two years of acquiring sufficiently cloud-free imagery, Fiji became the first Pacific Island Country to be mapped using satellite image data.

Figure 02: GIS&RS User Meeting at SOPAC in the 90ties.

The inventory introduced digital image processing capabilities within Fiji’s Forestry Department. More than 540 field plots provided training data for supervised maximum likelihood classification, supported by a digital terrain model derived from existing topographic maps. However, the project also revealed the limitations of early remote sensing technology, particularly the effects of atmospheric conditions and the importance of local geographical knowledge. Furthermore, inconsistencies in coordinate systems and map projections demonstrated that forest mapping was not solely a forestry activity but a national geospatial issue requiring cooperation between government agencies.

The decision to transform the inventory into a forest monitoring system strengthened GIS and remote sensing capacity within Fiji. During the first half of the 1990s, GIS units were established in several government departments, the University of the South Pacific (USP) and regional organisations. Monthly GIS and remote sensing user meetings began in Fiji, providing an informal platform for exchanging data, sharing technical knowledge, coordinating activities and presenting new applications. These meetings also encouraged the documentation of developments through a regional newsletter and provided the foundation for the Pacific Islands GIS and Remote Sensing User Conference.

Regional cooperation accelerated through initiatives supported by the South Pacific Applied Geoscience Commission (SOPAC). The GIS and Remote Sensing Newsletter, first published in 1993, expanded from documenting Fiji’s user meetings to distributing technical knowledge across the Pacific. In 1996, GIS-PacNet emerged as an email distribution network connecting GIS and remote sensing practitioners across the region. Training activities, software support and cooperation among regional organisations further strengthened these connections. MapInfo became a widely used GIS platform, while ERDAS supported digital remote sensing applications.

The establishment of GIS and remote sensing units in 16 Pacific Island Countries through a European Union-funded project in 2002 marked another important step in regional capacity development. Meanwhile, the user conference evolved from Fiji’s Year-end User Conference in 1998 into the Pacific Islands GIS and Remote Sensing User Conference in 2004. It became an international forum for exchanging experiences and presenting practical applications from across the region.

Looking back, these developments demonstrate that technological progress alone was insufficient to establish sustainable GIS and remote sensing capabilities. Informal professional networks, practical training, access to data and software, institutional cooperation and the exchange of local experience were equally important. The experience also illustrates how differences in institutional mandates and professional interests could create challenges, highlighting the need for effective coordination and clearly defined roles.

The paper concludes that documenting these early developments is important for understanding the foundations of the Pacific GIS and remote sensing community. The lessons learned remain relevant to current efforts to strengthen regional cooperation, develop technical capacity and sustain geospatial services in Pacific Island Countries.

Keywords: GIS; remote sensing; Pacific Islands; Fiji; Landsat; capacity development; regional cooperation; GIS-PacNet; PGRSC.

11:25 – 11:45
Fiu Etike Penjueli
International cooperation to increase the value and impact of open-source geospatial data and software in Fiji
Institutionnal
Fiu Penjueli, Marcus Blake, Sam Quinsey, Luis De La Rua, Diego Díaz Milanés, Niel Coffee, Jesse Whitehead Fiji Bureau of Statistics;  Centre for Australian Research into Access (CARA), Deakin Rural Health, Australia;  The Pacific Community (SPC); Background Open-source geospatial data and infrastructures are critical for small island states such as Fiji. Geographic dispersion, natural hazard risk, and constrained data resources place high demands on spatial evidence for health decision-making. Global open-source data now offer levels of completeness, accuracy, and timeliness that can rival some official datasets, enabling new opportunities to address spatial data gaps. However, open data benefits are not evenly distributed. Enhancing the value of open geospatial data requires clear application pathways, institutional collaboration, and capacity building to ensure data supports essential human activities, including service delivery and disaster resilience. Methods We demonstrate how open-source geospatial data can be integrated with national administrative and census-derived datasets to develop population reference frames, transport network models, and service location registers tailored to the Fijian context. Working in collaboration the Fiji Bureau of Statistics (FBoS), the Pacific Community (SPC), Centre for Australian Research into Access (CARA), Te Ngira Institute for Population Research and Universidad Loyola combined census-based population grids with open data sources to produce experimental service accessibility metrics. Results Fiji is an archipelago of more than 330 islands, with population distribution and service access strongly shaped by geography. Integrating population grids derived from the 2017 Population and Housing Census with open-source transport and service data has enabled new insights into spatial inequalities in service accessibility. These experimental metrics demonstrate the feasibility of using open data to support evidence-based planning. Conclusion This collaborative case study shows that open-source geospatial data, when combined with national statistical systems and supported by international cooperation, can substantially enhance service accessibility analysis and resilience planning in small island contexts. Strengthening local capability and institutional partnerships is essential to realising the full value of open data and ensuring equitable, policy-relevant outcomes under current and future climate risks.
11:45 – 12:45
Lunch
12:45 – 13:05
Sachiyo Nakamura
Deloitte SORA-NEX Project
Institutionnal
Across the Pacific Islands, the need for broad-area and continuous access to information is becoming increasingly important in areas such as climate change, disaster management, maritime monitoring, food security, communications, energy, and infrastructure. Given the region’s characteristics—including vast maritime areas, widely dispersed islands, and the remoteness of many communities—satellite data has strong potential as a practical means of complementing existing local knowledge, datasets, and operational systems.   SORA-NEX, selected under JAXA’s Space Strategic Fund, is a Japan–Australia satellite utilization program led by Deloitte. At the same time, the broader vision associated with this effort includes strengthening collaboration among Japan, Australia, and Pacific Island Countries through an open innovation approach spanning business partnerships, co-creation, policy dialogue, networking, events, and education.   In this context, Deloitte is exploring, together with Pacific stakeholders, how a similar collaborative approach may be shaped in ways that are grounded in Pacific priorities, institutions, and ongoing initiatives. This is not intended as a top-down transfer of an external model, but as a dialogue-based effort to co-identify realistic and sustainable use cases with governments, universities, regional organizations, and private-sector actors across the region.   Priority areas currently under consideration include disaster management, climate change, fisheries and maritime domain awareness, and data infrastructure development. Potential applications include monitoring floods, droughts, and volcanic activity; tracking coastal erosion and sea-level rise; supporting efforts to address illegal, unreported and unregulated (IUU) fishing; and improving fishing ground prediction and broader ocean monitoring. These application areas reflect both the Pacific’s urgent resilience needs and the practical strengths of satellite-enabled observation and communications.   Some elements of this collaborative approach have already begun on an exploratory basis through stakeholder dialogue, use-case discussions, and consideration of practical conditions for adoption in Pacific Island contexts. These early activities are intended to understand local realities more clearly, including data availability, human capacity, connectivity, operational arrangements, and institutional coordination needed to support practical collaboration.   In doing so, Deloitte seeks to work in a way that respects and complements the important roles already played by Pacific regional organizations and universities, including institutions, which are already active in areas such as environmental analysis, fisheries monitoring, capacity building, and geospatial education. The aim is to help widen opportunities for Pacific-led and mutually beneficial collaboration, rather than duplicate or displace existing efforts.   At PGRSC, we hope to share this perspective and discuss how satellite data can support Pacific priorities through partnerships built on dialogue, co-creation, and mutual respect. We also hope to explore how Deloitte can contribute as a trusted bridge among Pacific institutions, Japanese stakeholders, and international partners in shaping practical next steps for satellite-enabled cooperation in the region.
13:05 – 13:25
jonah sullivan
Rise of the Machines, Rise of the Workers
Institutionnal
ABSTRACT: The fourth wave of the industrial revolution has been seen by many as a very bad idea. Fortunately we’ll soon all be replaced by sentient machines ruled by tech oligarchs. Don\'t fear, there\'s some compelling parallels and lessons from good old fashioned organised labour! DESCRIPTION: The new ways of working in our workplaces have many benefits - and many new challenges: burnout, tech debt, remote colleagues, rapid hiring/firing, and changing career trajectories. These challenges may be accelerating, but they\'re hardly new, and neither are the solutions. There\'s wisdom tucked away in the pages of history, where workers once organized, bargained, and fought for better conditions. We\'d do well to dust off those old playbook pages and see what lessons still shine today. As Cory Doctorow noted at PyCon 2025, \"Tech workers\' power never came from solidarity; it came from scarcity\". With AI changing the industry, collective organizing is becoming more attractive to developers seeking job security and workplace protections. This presentation will provide an overview of common workplace issues, such as burnout, tech debt, remote colleagues, rapid hiring/firing, and changing career trajectories and how they have been addressed in the past through techniques developed by organised labour movements, that will help you have a better work life.
13:25 – 13:45
Vaipo Mataora, Meizyanne Hicks, Andrick Lal
Pacific Geospatial & Surveying Council (PGSC)
Institutionnal
Established in 2014, the Pacific Geospatial & Surveying Council (PGSC) brings together national leaders, practitioners, and technical experts in surveying, hydrography, and geospatial information management. Pacific Island countries and territories are highly vulnerable to coastal hazards including storm surge, king tides, and sea level rise. It provides a platform for Pacific Island Countries and Territories (PICTs) to collectively address challenges, share resources, and ensure that geospatial and surveying services are aligned to regional and global frameworks such as:
  • United Nations Convention on the Law of the Sea (UNCLOS)
  • United Nations Integrated Geospatial Information Framework (UN-IGIF)
  • Global Geodetic Reference Frame (GGRF)
  • International Hydrographic Organization (IHO) standards
The PGSC Strategy (2017–2027) set the foundation for strengthening Pacific capacity, collaboration, and governance. The long-term goal of PGSC is that the Pacific Island survey and geospatial services are sufficiently resourced to respond to member country needs and geospatial requirements for sustainable development. This long-term goal is supported by four priority areas for action identified in the Strategy and in particular address the goals as highlighted in regional meetings. These goals are:
  • Fostering regional leadership, direction and support for member states to engage stakeholders and the community on geospatial and surveying activities.
  • Adopting a modern Geodetic Reference Frame (GRF) and improving the technology underpinning geospatial systems and applications.
  • Increasing support and diversity of resources for geospatial and surveying activities at the national and regional level.
  • Building a self-reliant geospatial and surveying community supportive of learning, innovation and gender equity.
13:45 – 14:05
Thomas Tilak
Beyond the Geoportal: Building Sustainable Geospatial Data Services for the Pacific
Institutionnal
Across the Pacific, geospatial platforms are frequently developed as part of projects to meet an immediate need: publishing datasets, delivering a map portal, supporting a specific programme, or building a decision-support tool. The technical solution may be successful, but a more difficult question emerges when the project finishes: who keeps the service running, the data discoverable and the platform useful? Drawing on practical experience supporting geospatial and data platforms across Pacific Island countries and territories, this presentation explores how we can move from project-based geoportals towards sustainable geospatial data services. Rather than treating a geoportal as a standalone application, the presentation proposes thinking about geospatial infrastructure as a set of reusable capabilities: data storage, cataloguing and metadata, access control, APIs, visualisation, processing and long-term stewardship. It will discuss lessons from GeoNode-based deployments and emerging approaches using cloud-optimised formats and standards. The presentation will also examine an often overlooked part of geospatial architecture: the operating model around the technology. Sustainable services require clear ownership, governance, support responsibilities and realistic handover arrangements, not simply a server and an application. The goal is to start a practical conversation about how Pacific organisations can build interoperable geospatial services that survive individual projects, reduce duplication and progressively form a stronger regional geospatial network.
14:05 – 14:30
Coffee break
14:30 – 14:50
Sarah Pitcher-Campbell
AI and GIS for Pacific Organisations
New Technologies (AI, Deep Learning, Cloud)
Artificial intelligence is changing how organisations create, explore and use geographic information, but understanding where it can genuinely add value can be difficult amongst the rapid pace of new technology and AI hype. For Pacific organisations already using GIS to manage land, infrastructure, natural resources, hazards, field operations and other essential services, there are practical opportunities to make existing information and workflows easier to use.
This presentation explores several ways AI and GIS are beginning to work together, from improving field data collection and extracting information from imagery, through to allowing people to ask questions of maps and geographic data using natural language. It will also look at emerging agentic AI approaches, where AI can connect trusted organisational data with established GIS tools and analysis to support more complex workflows. Through practical demonstrations and Pacific-focused examples, the session will consider not only what is becoming possible, but also what is needed to use these capabilities responsibly. The focus will remain on authoritative data, appropriate governance and human judgement, with AI supporting rather than replacing the people who understand the organisation, its data and the decisions that need to be made.
18:00 – 22:00
Opening Cocktail
2026-12-01
07:00 – 08:00
Registration
08:00 – 08:20
Sébastien Lagarde
From Space to Sea: Multi-Source Maritime Awareness for Pacific Ocean States. Lessons from the Coral Sea Natural Park System and the MARIS Approach
Maritime Environment
Pacific Island countries and territories are large ocean states: their exclusive economic zones are vast, contiguous and largely unwatched, while patrol vessels and aircraft are few. Marine protected areas are multiplying under the 30x30 target and the BBNJ Agreement, yet many lack the means to be monitored, with the risk of becoming “paper parks”. No single technology can watch millions of square kilometres all the time: satellites provide a wide-area baseline, but revisit times, cloud cover and small wooden or fibreglass boats leave blind spots. Since 2021, MAGIS has operated S²P (Sentinelle du Parc), the surveillance system of the Coral Sea Natural Park, which covers 1.3 million km² of New Caledonia’s EEZ. Five years on, the system delivers:
  • A single operational picture — a web map fusing ten data sources (satellite AIS, VMS, Sentinel-1 SAR and optical imagery, weather, field observations) that shows every vessel, its track and its activity.
  • Civilian sentinels — 13 longline fishing vessels carry onboard sensors (radar, RF, etc.) whose contacts are relayed to the system via satellite, revealing small boats that satellites miss.
  • Automated alerts — rule-based detection of course changes, abnormal speeds, AIS gaps and unauthorised zone entries, sent directly to managers.
  • A long-term memory — more than 70 million vessel positions archived, used to map pressure zones, count entries and exits across the park boundaries and support the assessment of management measures.
Five years of operations have taught us four lessons: no single source is enough; quality control of heterogeneous data streams is a job in itself; nobody can watch a map 24 hours a day; and sea users are allies, not only subjects of control. Building on this operational experience, we are scaling S²P up into MARIS (Maritime Awareness and Response Information System), a modular and replicable solution for other territories, structured around six actions: See (fuse satellite and in-situ data), Involve (cooperative vessels as sensors), Cross-check (automated alerts today; tomorrow, AI agents trained on the park archive to learn normal activity patterns and escalate the right alert to the right person), Act (send assets to confirmed targets rather than blind patrols, decisions remaining human), Prove (long-term spatio-temporal analysis and before/after metrics on management rules) and Cooperate (connected EEZs sharing alerts, while each territory keeps sovereignty over its data). The presentation will walk through the system as it runs today, with real examples from the Coral Sea Natural Park, before presenting the MARIS roadmap. We will discuss how such a network could connect Pacific territories, in line with the conference theme of strengthening and expanding geospatial networks across the region. Keywords: maritime domain awareness; multi-source data fusion; SAR; AIS; in-situ sensors; AI agents; marine protected areas; Pacific
08:20 – 08:40
Nicholas Metherall
Coral Bleaching Extent Detection (C-BED) Copernicus Marine Data and Sentinel-2 spectral analyses to monitor the impacts of the 2026 marine heatwave in Tonga and Fiji.
Maritime Environment
Coral reef biogenic georesources have a foundational role in sustaining the livelihoods of coastal communities in Pacific Island Countries and Territories (PICTs) by providing food security, income, coastal protection, minerals and cultural benefits through fisheries and other ecosystem services. Yet, these ecosystems are becoming increasingly exposed to a warming ocean, which subsequently leads to marine heatwave events that trigger coral bleaching. Therefore, developing an Early-Warning System (EWS) to closely monitor coral bleaching in these communities is more important than ever before. Because the Pacific Ocean covers around one-third of the Earth, blending in-situ and remote sensing data is critical for early detection, monitoring, and timely action. This study presents the Coral Bleaching Extent Detection (C-BED) framework, which integrates Copernicus Marine Data Services, Pacific Ocean Portal, and NOAA Coral Reef Watch datasets with satellite data, including Sentinel-2 imagery, as well as observations of local field monitoring networks. The framework was implemented at six reef sites, three in Tonga and three in Fiji, to examine the correlation between SST anomalies, marine heatwave conditions, and spectral changes in coral reef ecosystems both before and after the 2026 marine heatwave event. The results provide a preliminary estimate of coral bleaching for these sites.
08:40 – 09:00
Rémi Andreoli
VeeSat: Coastal Early Warning from Space – Applications to the TAHATAI Neo Project
Maritime Environment

Pacific Island countries face growing coastal challenges driven by climate change, coastal erosion, extreme weather events, and increasing human pressures on shorelines and lagoon ecosystems. While Earth observation technologies offer unprecedented capabilities to monitor these dynamic coastal environments, the challenge is shifting from periodic observation to near-real-time geospatial intelligence, enabling early detection of coastal changes, anticipation of emerging threats, and timely decision-making.

Developed by Tauhã, a New Caledonian company specializing in geospatial intelligence, VeeSat introduces a new approach to Coastal Early Warning from Space through Virtual Sentinel Stations. These stations continuously monitor strategic coastal locations by integrating multi-source geospatial data, automated analysis, and intelligent change detection. The ambition is to transform complex observations into operational geospatial intelligence, actionable alerts, and ultimately real-time coastal early warning capabilities, supporting faster and more informed responses to emerging coastal risks. This presentation will introduce the VeeSat concept, its technological architecture, and its integration within the QVX GeoAI platform, highlighting how automated geospatial intelligence can support the transition from conventional coastal monitoring to proactive early warning systems. The TAHATAI Neo project, supported by the Space for Climate Observatory (SCO France), will illustrate the practical application of this approach in French Polynesia. Building upon the initial TAHATAI project, TAHATAI Neo explores how Virtual Sentinel Stations can combine multi-source geospatial intelligence, automated coastal change detection, and operational monitoring services to better identify and anticipate threats affecting coastal and lagoon environments. Beyond this regional application, the presentation will discuss the potential of VeeSat to establish a scalable network of Virtual Sentinel Stations across Pacific Island countries, supporting a transition toward continuous, predictive, and ultimately real-time coastal early warning services. This approach aims to bridge the gap between advanced geospatial technologies and the operational needs of coastal managers, local authorities, and decision-makers. Keywords: Geospatial Intelligence, VeeSat, Virtual Sentinel Stations, Coastal Early Warning, TAHATAI Neo, French Polynesia, Pacific Islands.
09:00 – 09:20
Chan-Su Yang | Dae-Woon Shin | Sree Juwel Kumar Chowdhury | Gulam Kibria Robbani
MULTISENSOR REMOTE SENSING FOR IUU FISHING SURVEILLANCE IN PACIFIC ISLAND COUNTRIES: 2026 PROGRESS AND PHASE II
Maritime Environment
Pacific Island countries (PICs) manage extensive exclusive economic zones (EEZs), where limited surveillance capacity and gaps in vessel reporting complicate efforts to address illegal, unreported, and unregulated (IUU) fishing. This presentation summarizes the KIOST–FFA project’s 2020–2025 results and its transition to Phase II, launched in 2026. In Phase I, KIOST developed vessel detection framework using Sentinel-1 synthetic aperture radar (SAR), Sentinel-2, and Landsat imagery. The dual-polarization Sentinel-1 method targeted vessels under 30 m and achieved a probability of detection of 0.84, including a 10 m vessel in a field experiment. YOLO-based optical detection was piloted in Fiji, Vanuatu, Samoa, and Tonga, and standardized CSV and PNG vessel-position products were test-shared with FFA. In a complementary study over Korean coastal waters, a random forest stage for Sentinel-2 residual cloud screening achieved 95.05% F1 and raised recovery of reference cloud pixels from 43.1% to 97.3%. In five scenes with archived YOLOv5 results, it removed 8.2% of detections, 96.5% of which were not ships. For nighttime monitoring, the KIOST VIIRS Boat Detection (VBD) algorithm processed Suomi-NPP, NOAA-20, and NOAA-21 Day/Night Band imagery in an average of 83.9 seconds per image. A 2026 parameter-adjustment study using 14 scenes reduced potentially noise-associated detections in low-radiance areas, although independent vessel-location data are needed to confirm false detections or establish improved accuracy. Analysis of Earth Observation Group VBD data for 14 PICs over 2017–2025 found a 37% increase in detected nighttime fishing activity, with Papua New Guinea showing the highest activity and distinct seasonal patterns across the region. AIS- and V-Pass-based fishing-activity and vessel-type methods were tested mainly in Korean waters, with a sample satellite-AIS application in the Pacific. A related study of satellite radio-frequency (RF) observations in congested Busan waters found low RF–AIS association rates across L-, S-, and X-bands (6.5%, 1.7–24.6%, and 7.7–17.2%, respectively), with signals concentrated in less congested areas. This highlights the need to interpret RF detections with vessel density and other data sources. Phase II (2026–2030) will integrate SAR, optical, VIIRS, RF, AIS, and VMS data for vessel matching, fishing-activity mapping and prediction, and an upgraded Regional Surveillance Picture (RSP), alongside capacity building with FFA and its Members. Pacific validation and evidence of operational use remain priorities; satellite detections alone do not establish IUU activity or enforcement impact. Keywords: IUU fishing, Pacific Island Countries, vessel detection, residual cloud screening, VIIRS Boat Detection, satellite radio frequency, maritime surveillance
09:20 – 09:45
Coffee break
09:45 – 10:05
Nicolae Cristian Popescu, Romulus Costache, Nicu Ciobotaru
Flash Flood Warning System Service in Romania
Disaster
The Flash-Flood Warning System Service (Disaster Risk Management theme) project which will be presented at this conference, was developed within the EU-funded MULTICARE project and presents the design, implementation, and testing of an operational flash-flood warning system for the municipality of Tecuci, Romania. The system aims to improve urban resilience to increasingly frequent and severe flood events driven by climate change. The Flash-Flood Warning System Service (FFWSS) is built around three core components: hydrological and hydraulic modelling, hazard and risk assessment, and warning dissemination. The hydrological component uses the EF5 (Ensemble Framework for Flash-Flood Forecasting) model together with meteorological forecasts from sources such as ICON, GFS, and MSWX to predict river discharges in the Tecucel River basin. When predefined flow thresholds are exceeded, the hydraulic model (HEC-RAS 2D) is automatically activated to simulate flood extent, water depth, and flow velocity across the city. A major part of the work focuses on creating flood hazard and risk maps. Hazard maps were generated for seven flood-return periods, while risk maps combine flood intensity with exposure indicators such as building characteristics, construction materials, building age, number of floors, and road-network density. The analysis identifies the most vulnerable urban areas and helps estimate potential impacts on infrastructure and residents. The warning system employs a four-level color-coded alert scheme (yellow, orange, red, and violet) corresponding to increasing flood severity. Once forecasts indicate a significant threat, automated warning messages are generated and sent to a hydrological expert for validation before being distributed to local authorities and emergency-response organizations. Messages include expected flood severity, affected areas, and recommended actions. An innovative aspect of the system is its integration with GIS-based transport and accessibility analysis. By combining flood forecasts with road-network data, the service can estimate travel times, identify isolated areas, and support the planning of evacuation and emergency-intervention routes during flood events. Overall, the report concludes that the FFWSS provides a practical and scalable solution for urban flood preparedness. Tested in Tecuci, the system demonstrates how advanced forecasting, risk assessment, and automated communication can be integrated to reduce human and economic losses from flash floods, while offering a model that can be replicated in other cities facing similar risks.
10:05 – 10:25
Cassie Stelow
From Geospatial Data to Disaster Decisions: Strengthening Risk Intelligence in the Pacific
Disaster

Effective disaster management depends on more than access to hazard data. Decision-makers need to understand what is happening, who and what may be affected, why some communities are more vulnerable than others, and what information is most important for preparedness, response, and recovery. Across the Pacific, answering these questions requires bringing together hazard information, population and infrastructure data, risk and vulnerability information, local knowledge, and operational context from a wide range of organizations and systems.

This presentation will explore how Pacific Disaster Center (PDC) works to transform geospatial data into actionable information for disaster management. Drawing on PDC’s experience supporting disaster management organizations throughout the Pacific, the presentation will examine the connection between baseline risk information, real-time early warning, impact analysis, and operational decision support.

A key focus will be the importance of understanding the underlying conditions that shape disaster impacts. Hazard intensity alone does not determine disaster outcomes. Population characteristics, critical infrastructure, access to essential services, coping capacity, and disaster management capabilities all influence how communities prepare for, respond to, and recover from an event. PDC’s National Disaster Preparedness Baseline Assessment (NDPBA) methodology provides one example of how these factors can be brought together to develop a more comprehensive understanding of disaster risk. Current work in Samoa will be used to illustrate this approach and provide a Pacific context for the discussion.

The presentation will also examine how baseline risk information can be connected with real-time hazard monitoring and impact analytics through PDC’s DisasterAWARE platform. Integrating these different types of information allows disaster managers to move from simply identifying where a hazard is occurring toward understanding potential impacts to populations, infrastructure, and communities and using that information to support operational decisions.

New technologies are increasing the ability to make these connections quickly. Automated data processing, impact modeling, machine learning, and emerging AI capabilities can help identify hazards, process large volumes of information, and communicate complex analysis in more understandable ways. These technologies, however, depend on trusted data, local expertise, transparent sources, and sustained relationships with the organizations responsible for managing and using the information.

Ultimately, strengthening geospatial networks in the Pacific is not only a technical challenge. It is a disaster management challenge. Connecting people, data, science, and technology creates the foundation for better preparedness, more effective early warning, faster impact assessment, and more informed decision-making before, during, and after disasters.
10:25 – 10:45
Jonelle Sayama | Keanno Fausto
Developing a UAV-Based Framework for Post-Typhoon Damage Assessment in Guam
Disaster
The U.S. territory of Guam is regularly threatened by tropical cyclones due to its location in the western Pacific Ocean. High-intensity storms can cause widespread impacts to buildings, transportation networks, utilities, vegetation, and other critical infrastructure, with cascading effects on public health, economic activity, natural resources, and community recovery. Rapid and comprehensive post-storm damage assessment is therefore critical for identifying impacted communities, prioritizing response activities, and supporting local and federal disaster assistance.  Traditionally, typhoon impacts on Guam have been documented using satellite imagery, radar, official weather observations, and field-based damage assessments. While ground surveys provide essential information, they can be labor-intensive and time-consuming, particularly following a major storm. Downed vegetation, construction debris, flooding, damaged roads, and other hazards can also restrict access to affected areas, creating potential gaps in the spatial coverage of damage assessments. To address these gaps, uncrewed aerial vehicles (UAVs) can serve as a rapidly deployable tool that can collect high-resolution imagery in areas that may be difficult or unsafe to access. Following the passage of Typhoon Mawar over Guam on 24-25 May 2023,  the University of Guam (UOG) Drone Corps deployed UAVs to provide the National Oceanic and Atmospheric Administration (NOAA) and the National Weather Service (NWS) Guam Forecast Office with imagery of immediate post-typhoon impacts that would support their efforts to validate storm-impact models and observations.  The surveys produced high-resolution, site-specific imagery of areas heavily impacted by the storm, providing insights such as wind direction and coastal inundation. While initially intended to support storm model validation for the NWS Guam Forecast Office, the resulting imagery also proved useful to other partner agencies for documenting localized impacts that could inform their own adjacent damage assessments. The Guam Power Authority, for example, used the imagery to determine the amount of solar panel installations that were blown off during the storm. Together with other partner-driven assessments following the storm, these applications revealed the broader operational role that UAVs can play in supporting post-disaster response in Guam.  Since Typhoon Mawar, this approach has evolved from an individual post-disaster deployment into a repeatable framework for UAV-supported storm assessment in Guam. Following Typhoons Sinlaku and Bavi in 2026, UOG Drone Corps applied the same general methodology in coordination with the Guam Bureau of Statistics and Plans (BSP). Following Typhoon Sinlaku, BSP identified a need for pre- and post-storm imagery to support public assistance disaster assessments, with high-resolution UAV imagery identified as the preferred method for documenting site conditions. Priority locations were identified following each storm, UAV imagery was collected over affected communities and infrastructure, and geospatial products were developed to document conditions and facilitate information sharing among partner agencies. Repeated implementation has also provided opportunities to refine interagency coordination, site prioritization, field operations, data processing, and dissemination procedures.  This presentation examines the progression of UOG Drone Corps’ post-typhoon response from Typhoon Mawar through subsequent storm events, highlighting how high-resolution UAV imagery can complement conventional ground-based and remotely sensed damage assessments. Particular attention is given to interagency coordination, rapid data acquisition, identification of vulnerable or difficult-to-access areas, and the dissemination of geospatial information for disaster response and recovery. Lessons learned across multiple deployments are used to identify opportunities for developing a more standardized UAV damage-assessment workflow for Guam.  The development of this capacity is timely as Guam and the broader western Pacific experience changing seasonal climate conditions, including the development of El Niño conditions in 2026. Recent tropical cyclone activity reinforces the importance of maintaining locally available and rapidly deployable technologies for documenting storm impacts and supporting recovery. By establishing a repeatable framework that combines local knowledge, interagency coordination, UAV technology, and high-resolution geospatial data, the UOG Drone Corps model provides a foundation for integrating emerging technologies into broader disaster resilience strategies for Guam and other vulnerable Pacific Island communities. 
10:45 – 11:05
Jumpei Takami
From Satellite Elevation to Digital Twins: Building a Reliable Height Foundation for Disaster Resilience in the Pacific Islands
Disaster

Flood risk in Pacific Island countries is decided by small differences in height. On low and flat islands such as Tongatapu, a few tens of centimetres can decide which houses are flooded, which roads can be used and which buildings are safe for vertical evacuation. Good elevation data is therefore the starting point for any disaster planning, but many islands still lack it.

This talk presents the \"Pacific Digital Twin for Disaster Resilience\" project, implemented by UNOOSA/UN-SPIDER with funding from the Government of Japan in the Cook Islands, Tonga and Palau, together with national partners EMCI, NDRMO and PALARIS. The project builds a foundation data layer from Pleiades Neo satellite imagery (15 cm) and a 1 m satellite stereo digital surface model, and uses it to create 3D digital twins of buildings (LOD2), flood simulations and flood dependent evacuation routes, linked with low cost IoT weather stations and a WebGIS platform.

We share three lessons from the first year.

First, satellite stereo DSM and LiDAR are complementary. LiDAR gives centimetre level accuracy and sharp building edges, and is the best choice for engineering scale work. Satellite stereo DSM gives consistent coverage across many islands at lower cost, needs no aircraft or ground access, and can be updated after a disaster. For archipelagos, a practical approach is LiDAR where it already exists, and satellite DSM to fill gaps and cover outer islands. Persistent cloud cover is the main challenge, and we explain how we have managed it.

Second, all elevation data, whether from LiDAR, drones or satellites, depends on the same geodetic foundation. GNSS and satellites measure height above the ellipsoid, while floods follow height above mean sea level. Without a geoid model, a unified vertical datum and working CORS stations, datasets from different sources cannot be combined reliably. Japan only completed its own move to a satellite based height system in April 2025. Pacific countries are not behind, and can move directly to the latest approach. We outline a simple pathway: conversion models, a national geoid, a unified vertical datum, and then planning decisions such as relocation.

Third, digital twins only last when national agencies own them. We describe how the project shares data with national and regional partners, trains national staff, and connects with SPC, JICA, JAXA and other initiatives so that the products continue to be used after the project.

The talk ends with an invitation to the UN-SPIDER regional workshop on 4 and 5 December at SPC, where participants can try the tools hands on.

11:05 – 12:05
Lunch
12:05 – 12:25
Yuming Wen
Salinity Issues in Hatgana Basin, Guam, USA
GIS software and technics
There are six basins located in northern Guam. They almost occupies the whole northern Guam lens aquifer (NGLA), which accounts for about 50% area of Guam. NGLA provides about 90% drinking water for Guam. Considering the importance of the groundwater resources in NGLA, it’s crucial to evaluate water quality in NGLA. This research focuses on the spatio-temporal analysis of patterns and trends of salinity measured by chloride concentrations in the Hagåtña Basin, Guam. Hatgana Basin is the southernmost basin in NGLA, and occupies 22.37% total area of the basins. Chloride concentration data from the wells with earliest data collection back to 1973 are used for processing and analysis. The wells with data of chloride concentration tested are located in the basal zone or the para-basal zone. Based on spatial and temporal analysis of the trends of chloride concentration in the study area, decadal trends indicate that salinity levels from the wells demonstrate general increasing patterns though the most recent data shows that the salinity levels display a general decreasing trend from 2010 to 2020. Twelve of twenty seven wells studied have chloride concentrations that surpass the local benchmark maximum contaminant level (MCL) of 150 mg/L and USEPA’s Safe Drinking Water Guideline MCL of 250 mg/L, and nine of those twelve wells have significant occurrences of chloride concentrations over the local benchmark value, while the chloride concentration from seven of those nine wells exceed US EPA’s MCL. Most of the wells located in the southeastern side of the basin have chloride concentrations that are out of standard. In order to analyze the geospatial and temporal trends and patterns of salinity levels in the study area, driving forces such as precipitation, well pumping rates, climate factors and human induced activities will be considered in the future, and therefore an overall understanding of characteristics of salinity in Hatgana Basin can be achieved.
12:25 – 12:45
Neha Naidu
Where Does the Water Go? Using High-Resolution LiDAR, GIS and Local Knowledge to Understand Urban Flooding in Fiji
GIS software and technics
Urban flooding is a recurring challenge in Pacific Island towns, where intense rainfall, low-lying terrain, urban development and drainage constraints combine to cause localised flood impacts. Understanding how water moves across the landscape is essential for drainage management, infrastructure planning and community resilience. This presentation shows how high-resolution LiDAR data from Kahuto Pacific’s Moana Data Service can be combined with GIS hydrological analysis and local knowledge to identify surface-water flow pathways and flood-susceptible areas, using the Nasinu urban area as a case study. Using a 1 m resolution Digital Elevation Model (DEM) 276.6 km², captured in 2025 the analysis maps terrain depressions, flow direction and flow accumulation to identify natural drainage pathways and areas where runoff is likely to concentrate. These results are then compared with existing drainage infrastructure and urban development. A key part of the work is validation with local stakeholders, including [Nasinu Town Council / Water Authority of Fiji / community representatives]. Initial consultations confirmed flooding at Nasinu Town Council locations identified by the analysis, and further local observations and drainage information are now being incorporated to refine the results. Because terrain analysis alone does not account for pipe capacity or rainfall intensity, this local input is essential. The case study demonstrates how collaboration between data providers, local authorities and communities can turn high-resolution geospatial data into practical tools for drainage assessment, disaster preparedness and resilient urban development. It offers a scalable approach for other Pacific Island towns seeking to understand and manage urban flood risk.
12:45 – 13:05
jonah sullivan
Git Blame: AI Edition
GIS software and technics
ABSTRACT: AI is really helpful to assist in writing code…but who taught AI to write code? Turns out it was trained on millions of open source software repositories that have applied a multitude of software licenses. And how do you copyright your AI-assisted code? Is it even your intellectual property? DESCRIPTION: This presentation explores the legal complexities using large language models that have been trained on open-source Python code, with a focus on how LLMs can inadvertently reproduce copyrighted or licensed Python snippets or functions. It examines the implications of common Python‑centric software licenses such as MIT, BSD, Apache 2.0, and GPL, and how these licenses apply to AI‑generated code. There will also be an examination of the legal concept of \"human creative contribution\" for copyright eligibility. Attendees will learn from recent court outcomes and Python‑relevant case studies, along with practical strategies for protecting Python repositories, and reviewing AI‑generated code. The presentation concludes with guidance on selecting and applying appropriate open‑source licenses to Python projects to mitigate infringement risks while maintaining developer productivity.
13:05 – 13:25
Neha Naidu
Connecting Moana Data, FME and ArcGIS for Geospatial Decision-Making in the Pacific
GIS software and technics
 

High-resolution geospatial data offers valuable opportunities for planning and resilience across the Pacific. However, turning diverse datasets into accessible information requires effective integration and processing. This presentation explores a workflow connecting Moana Data Service, FME and ArcGIS. Moana provides LiDAR-derived terrain data and aerial imagery, FME supports data validation, transformation and automation, and ArcGIS enables spatial analysis, web mapping and dashboards. Using examples from Fiji, the presentation will explore how these tools can support repeatable workflows for flood assessment and infrastructure planning. It highlights the potential of collaboration and connected geospatial platforms to improve data accessibility and support informed decisions across Pacific communities.

13:25 – 13:50
Coffee break
13:50 – 13:55
Desy Nuryunarsih
[Lightning batch] Where Can I Walk to Be Healthier? A Demo of Personalised 15-Minute Neighbourhood Health Maps Linking QRISK3 and WCRF/AICR Scores to Local Resources
Human landscape
Cardiovascular disease and cancer share modifiable risk factors and a shared geography, yet clinical tools such as QRISK3 and the WCRF/AICR cancer prevention score identify who is at risk without showing whether the resources needed to reduce that risk are within walking distance of home. We present an open-source tool, developed for the SULSA-funded GRAP pilot study, that links both scores to individual-level neighbourhood analysis. The workflow combines a REDCap survey, a Python script that calculates the 5-point WCRF/AICR score, and pedestrian network analysis of OpenStreetMap data to generate 5-, 10- and 15-minute walking zones around each home. Health-relevant resources are shown on an interactive map, and those matching a person\'s low-scoring components are highlighted as personal recommendations; for example, low physical activity highlights nearby parks and gyms, and a QRISK3 score of 10% or more highlights GP surgeries and pharmacies. The tool also computes isochrone compactness and a composite walkability score from 0 to 10. The scoring script passed all 19 validation checks across low, medium and high adherence test cases, and the full pipeline produced personalised maps for nine synthetic participants in about 90 seconds each on a standard laptop. Walking-zone compactness ranged from 0.41 to 0.78 and walkability scores from 2.8 to 8.6, with less walkable, less resource-rich neighbourhoods among higher-deprivation profiles. Using only free software and open data, and keeping all data within the institution, the tool turns risk scores into a personal, walkable action plan and is ready for testing with real participants in the GRAP feasibility study.
13:55 – 14:00
Mohammad Arar
Integrating Drone-Based Remote Sensing and GIS to Investigate Nitrous Oxide Dynamics in New Zealand’s Braided Rivers
Remote Sensing data and software
Geographic Information Systems (GIS) and drone-based remote sensing provide powerful tools for mapping channel morphology and quantifying hydrological characteristics in braided rivers. High-resolution imagery enables the measurement of channel dimensions, mapping of wetted areas, and characterisation of spatial variability across complex channel networks. Although braided rivers occur in specific geographical settings, their dynamic channel structures may create distinct environments for nitrous oxide (N2O), a potent greenhouse gas that contributes to global warming, accumulation and emissions. Understanding how channel morphology and hydrological conditions influence N2O accumulation and release is important for improving estimates of greenhouse gas emissions from these river systems. This research combines drone imagery, GIS-based spatial analysis, field measurements, and biogeochemical data to investigate channel morphology and N2O dynamics across eight gravel-bed braided rivers in New Zealand. The rivers represented contrasting hydrological regimes: large, high-discharge alpine-fed rivers and smaller, low-discharge hill-fed rivers. Sampling was conducted during winter and summer across three channel types: fast-flowing main channels, low-flow side channels, and backwaters with little or no flow. This design enabled investigation of how river type, season, and channel-scale variability influence N2O dynamics. Drone imagery was acquired using a DJI Matrice 210 v2 equipped with a MicaSense Altum multispectral sensor at an altitude of 120 m, achieving a ground sampling distance of 5.05 cm per pixel. Imagery was processed using Esri Drone2Map to generate georeferenced orthomosaics. These orthomosaics were classified in ArcGIS Pro using machine learning classification tools to distinguish four land-cover classes: water, gravel, barren ground, and forest. Wetted channel area was calculated from the mapped water class, while channel slope was derived from elevation data using the Summarize Elevation tool in ArcGIS Pro. Channel widths derived from drone imagery were combined with field measurements of water depth and average flow velocity to calculate channel discharge. Dissolved N2O was supersaturated relative to atmospheric equilibrium across all rivers and seasons. Mean saturation was higher in hill-fed rivers (231 ± 364%) than in alpine-fed rivers (116 ± 37%). Backwaters exhibited the highest N2O saturation and estimated fluxes, whereas main channels generally exhibited the lowest values. Mean estimated N2O fluxes were 72.2 µmol N2O m-2 d-1 in hill-fed rivers and 8.9 µmol N2O m-2 d-1 in alpine-fed rivers. Multi-model analyses identified nitrate availability as the primary predictor of N2O saturation, while temperature also influenced seasonal patterns. Integrating high-resolution drone imagery, GIS-based classification, and field measurements provides a spatial framework for characterising braided river environments and investigating their environmental variability. By quantifying channel dimensions, wetted areas, and hydrological characteristics, these methods support the interpretation of spatial differences in N2O accumulation and emissions. Combining geospatial analysis with biogeochemical measurements can improve understanding of greenhouse gas dynamics in braided rivers and contribute to more representative assessments of freshwater N2O emissions.
18:00 – 22:00
Talanoa Session, Kava and Music
2026-12-02
07:00 – 08:00
Registration
08:00 – 08:20
Pascal Antoine Dieujuste
Satellite IoT and Geospatial Integration for Environmental Monitoring in Fiji
Remote Sensing data and software
Environmental observation across Fiji\'s dispersed islands requires communications, maintainable field equipment, and data that national institutions can use. During 2026, ArkEdge Space\'s Fiji DX project advanced from an initial satellite Internet of Things demonstration to remote island deployment and technical planning with the Fiji Meteorological Services (FMS). This presentation examines the progress, practical constraints, and institutional arrangements involved in connecting ground observations through Low Earth Orbit (LEO) satellites to a geospatial platform, and outlines a pathway toward broader environmental monitoring across the Pacific. An initial demonstration in January 2026 connected weather observations at FMS facilities with ArkEdge\'s satellite communications system and AE Insights platform. More than 2,000 data packets were transmitted during the demonstration period, and compatibility between the AWS Weather Station and ArkEdge Space’s IoT transmitter was validated. Follow-up discussions with FMS and the National Disaster Management Office (NDMO) in February helped identify priorities for remote island observations, geospatial visualization, data integration, and partnership development. These activities established a basis for evaluating how satellite connectivity could complement existing observation infrastructure. In May 2026, the project extended field testing to Rotuma, deploying a weather sensor, ArkEdge IoT transmitter, and solar and battery power system. The deployed configuration connected the sensor directly to the transmitter, with data relayed through satellite and ground infrastructure to AE Insights. This provided a remote island setting in which to assess communications, power management, installation logistics, and data availability. The work also highlighted the need to plan maintenance and support around the realities of island access and local technical capacity. AE Insights provides the geospatial interface for reviewing satellite-delivered observations. The project has explored combining these observations with contextual information such as elevation, rainfall, and flood susceptibility to support environmental assessment. Engagement with FMS\' Climate Team in October translated this work into more specific integration requirements: access to historical records, incremental API retrieval of newly arrived data, station metadata, maintenance history, and clarification of pressure calculations. The agreed direction is for FMS to retrieve data through an ArkEdge API and manage its own reformatting and ingestion, initially for verification. The experience demonstrates that successful transmission is only one part of a usable monitoring service. Current satellite delivery is intermittent, and the project does not yet provide guaranteed next-day availability or the one-hour latency desired for priority observations. Rotuma records are currently available at hourly resolution, while finer data resolution requires further technical assessment. Existing test sensor compliance with World Meteorological Organization requirements has not been confirmed. FMS nevertheless recognizes a role for experimental observations in monitoring gaps, while distinguishing these from observations suitable for official products and analysis. Future work will focus on improving delivery timeliness and data completeness, testing compatibility with FMS-preferred sensors and data loggers, and evaluating connectivity for hydrological stations experiencing communications difficulties. Further platform development will support clearer station provenance and more useful data access. Training, documented maintenance responsibilities, and local ownership will be essential to sustain future deployments. A comprehensive MoU and NDA are being developed to formalize collaboration and data sharing, alongside exploration of funding for subsequent expansion. The longer-term ambition is a scalable service that complements national networks, reaches remote observation black spots, and connects satellite IoT with geospatial information used by Pacific institutions. Fiji\'s experience offers practical lessons for moving from experimental connectivity toward sustainable environmental monitoring: define user requirements early, validate equipment and data together, and develop technical capability alongside institutional agreements. The presentation will share these lessons and identify opportunities for regional collaboration on climate resilience and disaster preparedness.
08:20 – 08:40
Ahmad Soleymani
Transforming Satellite Earth Observation into Actionable Geospatial Services for Disaster Resilience
Remote Sensing data and software
Satellite Earth Observation has become an essential source of timely and spatially continuous information for disaster risk reduction and resilience building. However, the availability of satellite data alone does not necessarily translate into effective disaster management. The critical challenge is to transform Earth Observation data into reliable, accessible, and actionable geospatial services that can support informed decision-making before, during, and after disasters. This presentation introduces an integrated approach for transforming satellite-based Earth Observation into operational geospatial services for disaster resilience, drawing on experiences and applications from Iran. The approach considers the full value chain from satellite data acquisition and preprocessing to remote sensing analysis, geospatial information generation, service delivery, and decision support. It highlights the complementary use of multi-source Earth Observation data, GIS, cloud-based processing, artificial intelligence, and emerging GeoAI approaches to enhance the monitoring and assessment of environmental and climate-related hazards. The Iranian experience provides practical examples across diverse hazard contexts, including drought monitoring, flood assessment, dust-storm monitoring, land-subsidence detection, and environmental change analysis. These applications demonstrate how satellite-derived information can be transformed into user-oriented geospatial services that support situational awareness, risk assessment, early warning, and post-disaster assessment. Particular attention is given to the transition from individual satellite-data applications toward scalable, interoperable, and user-driven geospatial services that can be integrated into institutional decision-support frameworks. The presentation further explores how lessons from Iran can contribute to international knowledge exchange and be adapted to the specific geographic, environmental, and institutional conditions of Pacific Island countries. By connecting satellite Earth Observation capabilities with operational geospatial services, emerging technologies, and user needs, the proposed approach demonstrates how space-based information can move beyond data provision toward practical decision support and more resilient disaster risk management.
08:40 – 09:00
Iurii Cherniak
Tracing of ionospheric responses to Space Weather events with IGS ROTI maps
Remote Sensing data and software
Occurrence of the ionospheric plasma density irregularities as a response to geomagnetic storm manifestation represents a specific aspect of the Sun – Earth interactions. Strong geomagnetic storms lead to an occurrence of intense ionospheric irregularities at high latitudes as a result of a significant increase in the auroral particle precipitation and ionospheric electric fields. During space weather events, the zone of the intense ionospheric irregularities expands equatorward from high latitudes and can reach midlatitudes. At the same time, storm-induced plasma irregularities of equatorial origin that appear at midlatitudes remain largely unexplored. In this study, we demonstrate performance of GNSS-based ROTI maps to represent key features of ionospheric irregularity occurrence on a global scale. We are investigating intensity as well as latitudinal and temporal extensions of ionospheric irregularities detected by ROTI maps related with the auroral processes and storm-time EPB developments. The plasma irregularities developing at high/mid and low latitudes have entirely different physical mechanisms of their generation and further evolution. Approaches based on trans-ionospheric propagation of GNSS signals throughout these structures can effectively detect all of them. It allows to analyze the main features of such structures’ development under geomagnetic storm conditions and to estimate physical drivers leading to them generation. But ionospheric irregularities development is still not completely predictable in terms of onset time, intensity, and location, and can be dangerous for navigation and communication satellite systems. As the ground-based GNSS networks provide continuous measurements of the navigation signals parameters, it allows us to consider these facilities as a global observatory for monitoring of ionospheric plasma density spatial distribution and its dynamics toward better understanding of ionospheric plasma irregularities phenomena. Key words:  Ionosphere, plasma irregularities, GNSS, ROTI maps
09:00 – 09:20
Vasemaca Naraukawa | Fayth Tikotani | Elenoa Ciboiwasawasa
Sentinel-1 SAR Analysis for Disaster Monitoring and Environmental Monitoring
Remote Sensing data and software
Radar remote sensing provides all-weather, day-and-night monitoring capabilities for environmental and urban applications. By penetrating cloud cover, Synthetic Aperture Radar (SAR) enables reliable soil moisture estimation and building detection through backscatter analysis, with potential applications in disaster risk monitoring and response. This paper evaluates the capability of Sentinel-1 SAR data for mapping soil moisture as a proxy of flood inundation extents and detecting damaged buildings. Using the Digital Earth Pacific platform, Sentinel-1 Ground Range Detected (GRD) imagery underwent radiometric calibration, speckle filtering, and DEM analysis. Subsequent analysis of Vertical-Vertical (VV) and Vertical-Horizontal (VH) polarization backscatter responses enabled the characterization of soil moisture patterns and the detection of built-up areas. The findings demonstrate Sentinel-1 SAR\'s potential for soil moisture mapping and damaged building detection, supporting environmental monitoring and disaster risk management while informing future applications in Fiji and the Northern Mariana Islands.
09:20 – 09:45
Coffee break
09:45 – 10:05
Joseph Atalig, Bradley Eichelberger
Drone mounted thermal imaging to support natural resource management in the Northern Mariana Islands
UAV
Enter descriThe Commonwealth of the Northern Mariana Islands Division of Fish and Wildlife has been integrating a variety of remote sensing techniques for managing natural resources within the island chain, including monitoring species and associated threats. The use of drones/unmanned aerial systems (UAS) allow capture of information with minimal disturbance to the species and at large scales that can be used to supplement surveys conducted on-the-ground. In particular, DFW has been increasingly exploring the use of drone mounted thermal imaging systems to monitor the status of both endangered and invasive species. We present three use cases demonstrating how a Skydio X-10 UAS with paired FLIR and high-resolution optical cameras can support natural resources management at various spatial scales. Based on these use cases, the CNMI DFW will deploy large scale UAS mounted thermal imaging campaigns in the future for both the populated southern islands and more remote northern islands within the archipelago. ption here.
10:05 – 10:25
Ryan Perroy
Cultural and Natural Resource Mapping in Hawaiʻi
Mapping
As in many locations across the Pacific, cultural and natural resources within the Hawaiian Islands are under threat from pressures related to development, climate change, natural disasters, and invasive species. Low-altitude aerial lidar data, collected over densely vegetated areas using drone and helicopter platforms, provide an unprecedented ability to assess current conditions, quantify changes over time, and predict future impacts under various scenarios. Here, we share examples from Hawai’i related to native forest conservation, cultural resource mapping, and coastal monitoring. Lidar-derived metrics of forest structure, including Plant Area Index (PAI) and Foliage Height Diversity (FHD), can reveal sub-canopy differences relating to the presence and density of understory invasive species like strawberry guava (Psidium cattleyanum) and feral ungulates, differences challenging to detect with passive optical and multispectral remote sensing datasets alone. Similarly, rock wall structures and other cultural resources can be readily mapped underneath impenetrable stands of vegetation and geomorphic changes following hurricanes and other disturbances, including cliff erosion, can be quantified and shared with local agencies.
10:25 – 10:45
Jayson Inau
Remote Sensing and GIS approach towards Groundwater Investigation in the Kairuku District of Central Province, Papua New Guinea
Remote Sensing data and software
EThe geology of Papua New Guinea is among the most complex in the world, shaped by the collision of the Pacific and Australian plates. This collision led to the formation of the hard-rock terrain that dominates the Central Highlands and the New Islands of Papua New Guinea. The hard-rock terrain is one of the obstacles to groundwater investigation in Papua New Guinea. To overcome this obstacle, the Remote Sensing and Geographical Science approach was used for groundwater investigation rather than the traditional method of ground survey using the geophysical, geological, and hydrogeological methods, which are time-consuming and very expensive. The Remote sensing and GIS approach used multicriteria analysis to delineate the groundwater potential zone. Ten thematic layers, lithology, rainfall, soil, slope, drainage density, vegetation, land use and land cover, elevation, landform, and inundation, were used for the delineation of the groundwater zone in the Kairiku district of the Central province of Papua New Guinea. A pairwise comparison matrix was constructed to calculate the weight of the individual thematic layers. The thematic layers, with their individual weights, were integrated into ArcGIS using the weighted-sum overlay method to generate the groundwater potential map. The map shows zones of very poor, poor, moderate, good, and very good groundwater occurrence. Very poor zone occupies 1.58%, poor 42.19%, moderate 10.57%, Good 28.53%, and very good 17.13% in the study area. The groundwater potential map was validated using resistivity data collected during ground investigations by the Mineral Resources Authority. The resistivity data confirm the presence of groundwater in the areas indicated on the map. Keywords Groundwater, Geographic Information System, Kairiku, Remote Sensing, Resistivity nter description here.
10:45 – 11:05
Mark Edwin Tupas | Vioghn Jendrik Aquino | Keith Cabello | Noel Jerome Borlongan
Mapping Cascading Hazards Using Multi-Source Remote Sensing: Lessons from Recent Philippine Earthquakes and Floods
Disaster
Disasters rarely occur as isolated events. Earthquakes, extreme rainfall, flooding, landslides, coastal change, and infrastructure disruptions can interact across space and time, creating cascading impacts that a single Earth observation sensor or dataset can\'t fully capture. Recent disasters in the Philippines demonstrate how multi-source remote sensing, supported by international data-sharing mechanisms and ground observations, can provide a more comprehensive and timely picture of these complex events.   This paper presents the Philippine Space Agency\'s (PhilSA) experience applying multi-source Earth observation to recent disasters, focusing on the 8 June 2026 Mw 7.8 Sarangani and the 30 September 2025 Mw 6.9 Cebu earthquakes, and the sequence of hydro-meteorological events that affected both areas. The earthquakes generated destructive ground shaking, while earthquake-related mass movements and ground deformation presented additional hazards requiring rapid assessment. Both areas experienced Typhoons, Typhoon Tino (Kalmaegi) and Typhoon Inday (Bavi), which produced severe flooding and further illustrated how communities and infrastructure can be exposed to multiple hazards in a relatively short period.   PhilSA activated international space-based disaster response mechanisms, including Sentinel Asia and the International Charter on Space and Major Disasters, to access diverse satellite datasets for rapid assessment. Multi-sensor analysis combined nighttime-light observations from the Visible Infrared Imaging Radiometer Suite (VIIRS) with Synthetic Aperture Radar (SAR) and multispectral and very high-resolution (VHR) optical imagery. Nighttime lights provided an initial means of examining spatial patterns of electricity disruption and changes in human activity, while SAR and optical imagery supported assessment of potential structural damage, surface disturbance, and flooding. Optical change detection also identified coastal changes associated with vertical deformation. Inland, satellite imagery enabled the identification of earthquake-triggered landslides, including slope failures that obstructed streams and rivers, and supported subsequent delineation of potential upstream inundation.   Experience from the 2025 Cebu earthquake provided an opportunity to complement satellite-derived products with field observations. PhilSA used Sentinel-1, ALOS-2, and PlanetScope imagery for damage-proxy mapping in northern Cebu and subsequently conducted ground validation using geotagged observations of surface ruptures, collapsed buildings, and other earthquake impacts. These observations demonstrate the importance of linking satellite-derived indicators with field information to assess the reliability and limitations of rapid damage products. When considered alongside satellite-based flood mapping undertaken during major hydro-meteorological events in 2025, the cases illustrate how a common Earth observation capability can be adapted to different hazard types and phases of disaster response. PhilSA\'s 2025 disaster-monitoring activities included both flood and earthquake damage-proxy products, reflecting this multi-hazard approach.   Across the cases, no single sensor fully represented disaster impacts. Instead, the complementary characteristics of SAR, high-resolution optical imagery, nighttime lights, and field observations enabled assessment of different dimensions of cascading hazards. The experience also demonstrates that the effectiveness of Earth observation depends not only on sensor capability but on the networks that enable timely access to data, analysis, validation, and dissemination.   These Philippine experiences offer lessons relevant to disaster-prone Pacific Island countries, where earthquakes, tsunamis, tropical cyclones, floods, and landslides can produce similarly interconnected impacts. Strengthening regional and international geospatial networks, including mechanisms for rapid satellite tasking, data sharing, collaborative analysis, and validation, can therefore substantially improve situational awareness. The cases demonstrate how multi-source remote sensing can evolve from producing individual hazard maps towards an integrated framework for observing cascading disasters, providing decision-makers with a more complete understanding of rapidly evolving events.
11:05 – 12:05
Lunch
12:05 – 12:25
Teja Kattenborn
3Dtrees.earth: Free cloud-based processing, visualisation and analysis of forest LiDAR data
Forestry

Laser scanning captures trees and forests as three-dimensional collections of points, revealing their size, shape and spatial arrangement. These point clouds can support forest inventories and studies of forest structure, biomass and habitat. However, large files, complex software and demanding computations often make the data difficult to use. 3Dtrees.earth provides a freely accessible, web-based platform for bringing forest point clouds into a shared environment for processing, visualisation and analysis.

The platform supports close-range forest observations from terrestrial laser scanning (TLS), mobile laser scanning (MLS) and drone-based laser scanning (ULS). Users can upload datasets, explore them interactively in three dimensions and access cloud-based processing workflows through their browser. Depending on the data and workflow, analyses can identify individual trees and derive measurements describing tree dimensions and forest structure. Remote computing resources support these tasks, while shared datasets and documented workflows facilitate collaboration and reuse.

This presentation walks through the main steps, from uploading a point cloud to viewing trees and interpreting derived measurements. We explain how different scanning approaches capture complementary aspects of a forest and why data quality, coverage and method choice affect the results. We also discuss opportunities for researchers and practitioners to contribute datasets and analysis tools. By connecting accessible visualisation with cloud-based processing, 3Dtrees.earth aims to help a wider community turn detailed laser scans into useful information about trees and forests.

12:25 – 12:45
Chan-Su Yang | Sree Juwel Kumar Chowdhury | Won-Jun Choi
Development of Decision Ready Tools to Support Coastal and Marine Spatial Planning (II)
Maritime Environment
Pacific Island countries require current, consistent information on coastlines, land use, waves, shallow-water bathymetry, and marine habitats to support coastal and marine spatial planning (CMSP). This study presents progress during the first reporting period of the second phase of the Development of Decision Ready Tools to Support Coastal and Marine Spatial Planning project (August–December 2025), implemented by the Pacific Community (SPC) and the Korea Institute of Ocean Science and Technology (KIOST). The project connects satellite-derived products and national capacity development with Digital Earth Pacific (DEP), with activities spanning Fiji, the Republic of the Marshall Islands (RMI), Tonga, and Tuvalu. KIOST tested its Sentinel-2 coastline-extraction method on Tongatapu, Tonga, and its satellite wave-parameter retrieval method near the Ha’apai Islands. The coastline application demonstrated the method’s potential for island monitoring while exposing errors along rocky shores and artificial infrastructure, where land–water boundaries are difficult to distinguish; refinement is therefore required before consistent country-scale use. The wave method was transferred to a new island setting, extending testing beyond the initial Funafuti case. For Funafuti, Tuvalu, KIOST manually digitized a 2025 land-use map from high-resolution CAS500 imagery, complementing a 2023 map and providing a basis for subsequent land-use change analysis. In parallel, SPC conducted preliminary satellite-based land-cover classification in Fiji, Tonga, Tuvalu, and RMI, testing random-forest and foundation-model approaches. Alpha marine benthic-habitat classification products were generated for the four countries, alongside initial data collection and model-prediction work. KIOST also continued development of Sentinel-2 satellite-derived bathymetry, currently using the Korean Peninsula as a methodological study area, and an LSTM-based bleaching-susceptibility approach for coral reefs in Tonga, Tuvalu, and RMI. These activities broaden the project from coastline and wave products toward an integrated coastal and marine information portfolio. SPC facilitated land-cover and marine-habitat workshops in Fiji during the Pacific GIS and Remote Sensing Conference to help national stakeholders interpret derived products and consider their use in decision-making and DEP. The reporting period identified practical constraints: rocky and built-up shorelines complicate coastline extraction, similar spectral signatures challenge land-cover classification, and limited calibration and validation data restrict confidence in habitat maps. Accordingly, the products reported here are preliminary or under development; independent accuracy assessment, additional field reference data, algorithm refinement, and country-level validation remain necessary. The Phase II work establishes a staged pathway for improving satellite products, validating them with national partners, and integrating decision-relevant coastal information into DEP.
12:45 – 13:05
Charlington Leo
Vanuatu Data Science Driving Innovation in Climate Change and Natural Disasters
Disaster
The Vanuatu Data Science project is a strategic government initiative, supported by SPC and funded by the World Bank, aimed at strengthening the country\'s capacity to collect, manage, analyse and utilize reliable data for climate change adaptation and natural disaster management. Through partnerships between key government institutions, the project introduces innovative digital tools, machine learning, statistical analysis and an integrated Management Information System (MIS) to improve disaster impact assessments, resource mapping, preparedness and emergency response. Ultimately, the initiative promotes evidence-based decision-making, strengthens institutional coordination and improves resilience and public service delivery for vulnerable communities across Vanuatu.
13:05 – 13:25
Angela Manchester
Reading the Shifts: What Job Listing Trends Suggest About the Future of GIS and Remote Sensing
New Technologies (AI, Deep Learning, Cloud)

This talk looks at directional patterns observed in an ongoing dataset of sourced GIS, remote sensing, and Earth observation job listings, tracked across global regions including Oceania, and what they might suggest about where the field is heading.

A few relationships appear worth watching. As AI and machine learning tools become more embedded in geospatial workflows, there appears to be a parallel rise in demand for people who can bridge traditional GIS skills with applied data science, rather than a straightforward replacement of one by the other. At the same time, a decline in purely manual mapping or digitising roles seems to coincide with growth in roles centred on data management, quality assurance, and systems administration, suggesting the work is shifting rather than shrinking.

Open-source tools also seem to be gaining ground as a parallel track alongside established commercial platforms, which may lower the barrier to entry for individuals and institutions with limited budgets, a dynamic particularly relevant in the Pacific.

Remote work appears to remain the exception rather than the norm in this sector, which has implications for how location and mobility factor into career planning.

These are directional observations from one evolving dataset, not firm conclusions, and the talk is intended to prompt discussion rather than present definitive findings. It will walk through the trends and invite the audience to weigh in on what they are seeing in their own markets.

13:25 – 13:50
Coffee break
13:50 – 14:10
Andrick Lal, Veenil Rattan, Marika Kalouniviti, Rusiate Turuva
LiDAR and Geodetic Surveys
Mapping
Pacific Island countries and territories are highly vulnerable to coastal hazards including storm surge, king tides, and sea level rise. Reliable coastal inundation forecasting requires high-resolution elevation data, particularly through Light Detection and Ranging (LiDAR). While several Pacific countries have undertaken LiDAR surveys, coverage remains uneven across the region. There is a need for a coordinated, evidence-based Pacific Site Prioritization Framework to guide investments in LiDAR acquisition to support coastal inundation forecasting, early warning systems, and climate resilience planning. This regional consultation will bring together technical experts and decision-makers to co-develop a prioritization framework that reflects Pacific needs, risks, data gaps, and operational forecasting requirements. In February 2015, the UN General Assembly adopted the resolution (A/69/L.53) “A Global Geodetic Reference Frame (GGRF) for Sustainable Development” - the first resolution recognizing the importance of a globally coordinated approach to geodesy. In the Pacific, Fiji, Papua New Guinea, Samoa, Solomon Islands, Tuvalu and Vanuatu have adopted the UN resolution for GGRF. Geodetic surveys are critical for LiDAR because they provide the precise positioning framework that ensures LiDAR data is accurate, reliable, and usable for engineering, planning, environmental management, and disaster resilience applications. The geodetic surveys establish accurate positioning and heighting reference system to correctly position the LiDAR data on the earth and relate it to the ocean, through sea level observations.
14:10 – 14:30
Andrick Lal, Veenil Rattan, Marika Kalouniviti
Pacific Sea Level & Geodetic Monitoring
Mapping
The Pacific Sea Level and Geodetic Monitoring project began in 1991 as an Australian response to concerns raised by the member countries of the South Pacific Forum over the potential impacts of global warming on climate and sea levels in the Pacific. Now under the under the Climate and Oceans Support Program in the Pacific (COSPPac), Australia and New Zealand, has been supporting the 15 Pacific Island countries (Cook Islands, Federated States of Micronesia, Fiji, Kiribati, Marshall Islands, Nauru, Niue, Palau, Papua New Guinea, Samoa, Solomon Islands, Tokelau, Tonga, Tuvalu and Vanuatu) to measure, record and analyse long-term sea level and land motion for over 25 years. The land motion data is collected continuously at one or two Global Navigation Satellite System (GNSS) stations in each of the 15 PICs. The primary goal “to generate an accurate record of variance in long-term sea level for the Pacific and to establish methods to make the data readily available and usable by Pacific Island Countries
14:30 – 14:50
Rémi Andreoli, Bradley Eichelberger, Maria Kottermair, Arti Pratap, Kasaqa Tora, Salote Viti, Wolf Forstreuter
PGRSC Digital Platform: Connecting Members, Expertise and Opportunities Across the Pacific
Institutionnal

The Pacific GIS and Remote Sensing Council (PGRSC) is introducing its redesigned website and digital platform, developed to strengthen collaboration, professional networking, and knowledge sharing across the Pacific GIS and remote sensing community.

More than a communication website, the new platform aims to become a central hub connecting PGRSC members, institutions, professionals, and regional partners, while facilitating access to expertise, resources, and professional opportunities. This presentation will introduce the platform\'s key features, including the new Member Dashboard, providing members with a personalized space to manage their profiles, access membership services, and engage with the PGRSC community. Particular attention will be given to the Consultant Roster, designed to increase the visibility of regional expertise and facilitate connections between qualified professionals and organizations seeking specialized skills. The Job Opportunities section will also be presented as a tool for sharing employment, consultancy, and career development opportunities across the region. Additional functionalities supporting member engagement, institutional visibility, news sharing, and conference activities will be demonstrated through practical examples. By bringing together people, expertise, and opportunities within a single digital environment, the PGRSC aims to strengthen its role as a regional professional network and support the development of a more connected and collaborative Pacific geospatial community.
14:50 – 15:10
Teja Kattenborn
deadtrees.earth: Free cloud-based mapping of forest cover and standing dead trees from aerial imagery
Forestry
Forests are changing rapidly, and aerial photographs can reveal where trees have died and how forest cover is changing. Drones make it possible to collect detailed images locally, but turning these images into useful maps often requires specialist software, computing resources and technical expertise. deadtrees.earth provides a freely accessible, web-based platform that brings data processing, visualisation and analysis together in one place. Users can contribute aerial and drone imagery, explore datasets through an interactive map, and use cloud-based tools to analyse forest cover and standing deadwood. Artificial intelligence helps identify tree cover and visible dead tree crowns in high-resolution images. Processing runs on remote computing infrastructure, allowing users to work through a web browser. The platform also supports sharing imagery and analysis results with the wider community, helping individual surveys contribute to a growing resource for forest research and monitoring. This presentation introduces the platform through a practical workflow, from contributing imagery to inspecting maps and interpreting the results. It explains what aerial images can reveal about tree mortality, why image quality and local conditions matter, and how users can assess the outputs. We discuss opportunities for researchers, forest managers and local organisations to apply the platform, contribute data and help improve its tools. The aim is to make detailed forest monitoring more accessible and to connect local observations with collaborative efforts to understand forest change.
16:00 – 18:00
PGRSC AGM
18:00 – 22:00
Through the Fiji Night clubs
2026-12-03
07:00 – 08:00
Registration
08:00 – 08:20
Lorant Czaran | Jumpei Takami
Space for Disaster Resilience in the Pacific: How UNOOSA and UN-SPIDER Support Island Countries
Disaster

The United Nations Office for Outer Space Affairs (UNOOSA) works to make the benefits of space available to all countries. Its UN Platform for Space-based Information for Disaster Management and Emergency Response (UN-SPIDER) was established by the UN General Assembly in 2006 (resolution 61/110). Its mission is to help all countries, especially developing countries, use space-based information across the full disaster management cycle.

This talk explains what UNOOSA and UN-SPIDER do, and how Pacific Island countries can work with us. UN-SPIDER operates from offices in Vienna, Bonn and Beijing, supported by a global network of Regional Support Offices and National Focal Points. Our main services are technical advisory missions, capacity building, knowledge sharing through the UN-SPIDER Knowledge Portal, and support during emergencies.

In emergencies, UNOOSA is a Cooperating Body of the International Charter \"Space and Major Disasters\". Since 2018, UNOOSA can request Charter activations directly on behalf of national disaster management agencies, and it helps these agencies become Authorized Users so they can request satellite data themselves. We also work closely with regional mechanisms such as Sentinel Asia, led by JAXA, which provides satellite based emergency observation for the Asia Pacific region.

In the Pacific, UN-SPIDER carried out a technical advisory mission to Tonga in 2023, followed by the Tonga Disaster Preparedness Pilot Project with CEOS in 2024. We are now implementing the \"Pacific Digital Twin for Disaster Resilience\" project in the Cook Islands, Tonga and Palau, funded by the Government of Japan, and a digital twin project in Solomon Islands supported by the Philippine Space Agency. Each of these activities is linked to national ownership and to regional partners such as SPC and PGRSC.

The talk will close with practical pathways for Pacific agencies: how to request a technical advisory mission, how to access the Charter and Sentinel Asia, how to join training, and how to connect national work with global frameworks such as the Sendai Framework and Early Warnings for All. The following talk will present the Pacific Digital Twin project in more detail.

08:20 – 08:40
Tomoyuki KITAMOTO | Moeka YAMAJI | Natsumi UEDA | TBD
JAXA's Satellite-Based Tools and International Cooperation for Disaster Management: GSMaP, Today's Earth, GLInT, and Sentinel Asia – Part I
Disaster
JAXA will introduce several systems and initiatives through which JAXA supports disaster management activities using Earth observation satellite data. These include details and use cases of utilizing the following systems and initiative. GSMaP: Real-time global rainfall monitoring using multiple satellite observations. GSMaP has multi-satellite global precipitation map under the Global Precipitation Measurement (GPM) Mission, by using Dual-frequency Precipitation Radar (DPR) onboard GPM core satellites, other GPM constellation satellites, and Geostationary satellites. This website offers hourly global precipitation information Today’s Earth: Satellite-driven hydrological modeling for flood and water-cycle monitoring. Today\'s Earth (TE)\" is JAXA\'s terrestrial hydrological simulation system developed under the joint research with University of Tokyo. Various hydrological products relating to the conditions of land surface and river are obtained as the results of numerical simulation. GLInT: The GSMaP Local Integration Tool with gauges (GLInT) is a tool that improves GSMaP satellite-based precipitation estimates by incorporating ground-based rain gauge observations. By combining satellite and rain gauge data, GLInT provides more accurate and reliable quantitative rainfall information, supporting applications such as local weather warnings and flood management. Sentinel Asia: International cooperation framework that provides satellite-based information for disaster management across the Asia-Pacific region. As of October 2026, Sentinel Asia are composed of 130 member organizations from 30 countries/regions and 21 international organizations including the Pacific Community (SPC), NDRMO Fiji, NDRMO Tonga, and UNOCHA\'s Office of the Pacific (OoP) from the Pacific. When disasters occur, member organizations can activate Sentinel Asia and receive not only satellite data but also action-ready information for their disaster activities. Members are variety from space organizations, universities, disaster management organizations and international organizations in the Asia-Pacific. Note: If possible, we would appreciate having two separate presentation slots (20 minutes each) rather than a single combined session for 20 mins. In addition, due to our travel schedule, we will arrive in Suva on the morning of 3 December, therefore we would be grateful if our presentations could be scheduled for the afternoon on 3 December, if possible. Thanks.
08:40 – 09:00
Tomoyuki KITAMOTO | Moeka YAMAJI | Natsumi UEDA | TBD
JAXA's Satellite-Based Tools and International Cooperation for Disaster Management: GSMaP, Today's Earth, GLInT, and Sentinel Asia – Part II
Disaster
JAXA will introduce several systems and initiatives through which JAXA supports disaster management activities using Earth observation satellite data. These include details and use cases of utilizing the following systems and initiative. GSMaP: Real-time global rainfall monitoring using multiple satellite observations. GSMaP has multi-satellite global precipitation map under the Global Precipitation Measurement (GPM) Mission, by using Dual-frequency Precipitation Radar (DPR) onboard GPM core satellites, other GPM constellation satellites, and Geostationary satellites. This website offers hourly global precipitation information Today’s Earth: Satellite-driven hydrological modeling for flood and water-cycle monitoring. Today\'s Earth (TE)\" is JAXA\'s terrestrial hydrological simulation system developed under the joint research with University of Tokyo. Various hydrological products relating to the conditions of land surface and river are obtained as the results of numerical simulation. GLInT: The GSMaP Local Integration Tool with gauges (GLInT) is a tool that improves GSMaP satellite-based precipitation estimates by incorporating ground-based rain gauge observations. By combining satellite and rain gauge data, GLInT provides more accurate and reliable quantitative rainfall information, supporting applications such as local weather warnings and flood management. Sentinel Asia: International cooperation framework that provides satellite-based information for disaster management across the Asia-Pacific region. As of October 2026, Sentinel Asia are composed of 130 member organizations from 30 countries/regions and 21 international organizations including the Pacific Community (SPC), NDRMO Fiji, NDRMO Tonga, and UNOCHA\'s Office of the Pacific (OoP) from the Pacific. When disasters occur, member organizations can activate Sentinel Asia and receive not only satellite data but also action-ready information for their disaster activities. Members are variety from space organizations, universities, disaster management organizations and international organizations in the Asia-Pacific. Note: If possible, we would appreciate having two separate presentation slots (20 minutes each) rather than a single combined session for 20 mins. In addition, due to our travel schedule, we will arrive in Suva on the morning of 3 December, therefore we would be grateful if our presentations could be scheduled for the afternoon on 3 December, if possible. Thanks.
09:00 – 09:20
Ngima Chhiri Sherpa
Glacial Lake Evolution and the 2024 Thame GLOF in the Nepal Himalaya: A Multi-Source Remote Sensing Assessment
Disaster
The accelerating retreat of glaciers under climate change is driving the rapid expansion of glacial lakes across High Mountain Asia, progressively increasing the frequency and magnitude of glacial lake outburst floods (GLOFs) and the risk they pose to downstream communities. Despite this, conventional hazard assessment frameworks remain predominantly focused on larger lakes, leaving smaller yet potentially dangerous water bodies systematically underrepresented. This study investigates the 2024 Thame GLOF event in the Khumbu region, Nepal, using multi-temporal satellite remote sensing to reconstruct long-term glacial lake evolution, outburst dynamics, and downstream impact across five adjacent lakes within a single catchment. We investigated the cascading breach mechanism whereby GL5, underwent dam failure in 2024 and triggered a secondary breach of the hydrologically connected GL4, generating a compound outburst substantially exceeding the destructive potential of either lake independently. Prior to the GLOF event, GL5 expanded more than 12 times its area from approximately 0.008 km² to 0.1 km² at an average annual growth rate of 18% between 2015 and 2024. This rapid expansion has likely increased hydrostatic pressure on the moraine dam, potentially contributing to its failure and subsequent outburst. This represents one of the few directly observed cascading multi-lake GLOF events documented in HMA. The flood impacted approximately 130 Ha of the catchment, reached a maximum inundation width of ~400 m at Thame village, destroying 36 households and 10 Ha of farmland, and resulting in an estimated economic loss of 4.13 million USD. Our study underscores the critical inadequacy of size-based lake prioritization in GLOF hazard frameworks and advocates that consequence-based, catchment-scale assessments explicitly incorporating downstream geological conditions, community exposure, and socio-economic vulnerability are essential. Finally, we recommend that future GLOF hazard assessments adopt localized approaches that can incorporate potentially dangerous smaller lakes and the produced scientific findings to be actively translated into proper implementations safeguarding vulnerable mountain communities.
09:20 – 09:45
Coffee break
09:45 – 10:05
Keanno Fausto | Jonelle Sayama
The University of Guam Drone Corps – Fostering A Network of Skilled Drone Pilots & A Repository of Geospatial Data for Micronesia
Mapping
Pacific islands in Micronesia are increasingly vulnerable to a range of environmental threats, including natural disasters, coastal erosion, invasive species, and wildfires. While traditional data collection methods (e.g. satellite imagery, crewed aerial surveys, and field-based observations) remain essential tools for assessing environmental changes and hazards, each presents limitations in spatial resolution, accessibility, and operational costs that are particularly pronounced in remote island settings. Because many of these risks are rapidly evolving and site-specific, resource managers often require timely, high-resolution, and localized data to accurately assess their impacts and inform response efforts. Drones have emerged as a cost-effective and flexible solution that can augment traditional approaches by delivering efficient data that can address the unique needs of resource managers in Pacific island communities. In response to these needs, the University of Guam (UOG) Drone Corps has built a network of Federal Aviation Administration (FAA) Part 107b-licensed drone pilots who are trained in remote operations, local airspace regulations, and environmental considerations. Since its launch in 2021, the program has certified 64 drone pilots. This includes students from UOG and the Guam Community College. In 2025 and 2026, the program expanded its membership to university staff, faculty, and partners at natural resource agencies. Together, these members represent a cadre of experienced pilots ready to mobilize for various surveying needs, including vegetation health assessments, coral reef monitoring, and post-disaster imagery. This presentation will showcase the UOG Drone Corps model, highlighting its course structure, key mapping campaigns, and collaborative approach. Through partnerships with local agencies, remote pilots from the program participate in missions to gain training experience in acquiring geospatial data (e.g., RGB, NDVI, LIDAR, GHG emissions). The gathered datasets are processed, documented with metadata, and published in an online repository to supply resource managers and researchers with data that can guide long-term environmental monitoring. Beyond Guam, the UOG Drone Corps fosters regional capacity by sharing its expertise with institutions in Saipan, Palau, and Pohnpei to support the development of their own drone training programs. Because the UOG Drone Corps operates primarily within the U.S. FAA framework, the presentation will also explore how its training model can be adapted to non-U.S. Pacific island contexts. Finally, the presentation will conclude with an outlook on its future, including plans to increase capacity by introducing new technologies and broadening its collaborations with partners beyond Micronesia.
10:05 – 10:25
Hayato FUKUOKA
Building Sustainable Geospatial Societies: Lessons from Japan\'s Geospatial Information Infrastructure and International Cooperation
Mapping
This presentation examines the relationship between the geospatial information (GI) infrastructure developed and promoted by the Japanese government and the growing utilization of geospatial information in private-sector activities. Japan\'s GI infrastructure, including surveying regulations, Earth observation satellites, positioning satellites and the Continuously Operating Reference Station (CORS) network, digital topographic maps, and geospatial information platforms, has provided a foundation for a wide range of applications. These applications extend from conventional public-sector uses such as urban development, infrastructure management, and disaster risk reduction to emerging fields including digital twins and autonomous driving. The presentation also introduces examples of how Japanese experience and knowledge have been shared with other countries through Official Development Assistance (ODA) and related international cooperation programs. Based on these experiences, the presentation explores possible foundations for achieving a sustainable geospatial information society and promoting the broader utilization of geospatial information
10:25 – 10:45
Malakai Tuinasau Tadulala
Geospatial Analysis of Microplastic and Sigatoka River Quality Assessment in 1950 - 2026, Using Artificial Neural Network (ANN), Convolutional Neural Network (CNN), and Random Forest (RF).
Mapping
The introduction of cutting-edge geospatial technology has emerged as one of the most prominent, distinguished, and groundbreaking areas of research. Hence, an open-source platform is provided, along with source code for widely used GIS libraries: GeoPandas, Matplotlib, NumPy, Xarray, Folium, and Python. Moreover, the introduction of new geospatial public-domain code for Machine Learning sensors extended prediction analysis to longer time horizons by leveraging secondary, large-scale public data from Epoch AI Data models (AI data, AI chip sales and AI chip owners) and big data analytics in ArcGIS Pro and QGIS geospatial databases. Developed an integrated remote sensing Landsat 8 and Sentinel 2A and machine learning workflow using Google Earth Engine (GEE), `geemap`, and Python to analyzed, predicted, and mapped water quality (turbidity, chlorophyll-a) and model microplastic proxy distribution along the Sigatoka River in Fiji across five historical and future temporal eras; (Era-1)1650-2002, (Era-2)2002-2025, (Era-3) 2026-2030, (Era-4)2030-2050, and (Era-5)2050-2100. Implemented, trained (70%-30%), and compared models from RF, ANN and CNN on extracted spectral indices (NDWI, NDVI, and turbidity indices) from Landsat and Sentinel-2 imagery, and generated predictive time-series trend plots from 1950 to 2100. As a result, changes in the spectral indices (NDWI, NDVI, and NDTI) were used to simultaneously predict proxy concentrations of Turbidity, Chlorophyll-a, and Microplastics. Hence, the models analysed Mean Squared Error (MSE) for (1) Turbidity: 8.94 (RF), 5.18 (ANN), and 5.83 (CNN); (2) Chlorophyll-a: 6.28 (RF), 3.11 (ANN), and 3.13(CNN); and (3) Microplastics Proxy: 26.40 (RF), 25.40 (ANN), and 28.09 (CNN). Whilst the R2 validation scores for (1) Turbidity: 0.95 or 95% (RF), 0.98 or 98% (ANN), and 0.97 or 97% (CNN); (2) Chlorophyll-a: 0.98 or 98% (RF), 0.99 or 99% (ANN), and 0.99 or 99%(CNN), Microplastics Proxy   0.98 or 98% (RF), 0.99 or 99% (ANN), and 0.99 or 99%(CNN). In conclusion, the following key findings were identified: distinctive historical & predictive eras. The study analysed and forecasted water quality trends from 1950 to 2100 across five key temporal segments. It is recommended that Integrated Coastal Planning for Estuarine management and runoff-control policies mandatorily address mitigation of agricultural and urban discharges during these peak temporal windows. The sharp rise in predicted microplastics and turbidity during the 2002–2025 and 2026–2030 eras summarised a critical period of environmental stress.  
10:45 – 11:05
Sree Juwel Kumar Chowdhury | Chan-Su Yang
Coastline Mapping Using Sentinel-2 Imagery with Support from Sentinel-1 SAR in Tongatapu, Tonga
Mapping
Accurate coastline information plays an important role in environmental monitoring, coastal management, and natural hazard assessment, particularly for Pacific Island countries such as Tonga. However, automatically extracting coastlines from optical satellite images is often difficult due to complex coastal environments, including sandy beaches, mangroves, shallow waters, man-made structures, and cliffs. These varying surface characteristics can produce similar spectral responses from land and water, making coastline identification challenging. In addition, frequent cloud coverage limits the effectiveness of optical imagery acquired on a single date. To address these challenges, this study proposes an automated coastline extraction approach that primarily uses Sentinel-2 Level-1C imagery, supplemented by Sentinel-1 synthetic aperture radar (SAR) observations. The proposed method combines superpixel segmentation, spectral information, adaptive thresholding, and selected SAR-based analysis. Initially, Sentinel-2 images are divided into spatially connected superpixels using the simple linear iterative clustering (SLIC) algorithm, minimizing pixel-level variations while preserving the spatial characteristics of complex coastal areas. Land and water regions are subsequently distinguished using superpixel-level spectral statistics based on the normalized difference vegetation index (NDVI). Cloud-contaminated areas are identified and masked using the Sentinel-2 scene classification layer. Where cloud coverage prevents reliable coastline extraction, Sentinel-1 ground range detected (GRD) imagery acquired close to the Sentinel-2 observation date is used to supplement the missing information. The SAR data are processed through radiometric calibration, speckle filtering, and terrain correction. Land–water boundaries derived from Sentinel-1 are then incorporated exclusively within cloud-affected areas, allowing missing coastline sections to be reconstructed while maintaining the Sentinel-2-derived coastline in cloud-free regions. By combining the complementary capabilities of optical and SAR observations, the proposed approach minimizes cloud-related discontinuities and improves the consistency of coastline delineation. The findings indicate that this framework provides a more reliable approach to automated coastline extraction across Tonga\'s diverse coastal conditions.
11:05 – 12:05
Lunch
12:05 – 12:25
Jessica Pollock
Hold the Line: An AI-Verified Evidentiary Registry for Pacific Maritime Baselines
New Technologies (AI, Deep Learning, Cloud)

Pacific Island states have deposited fixed maritime baselines with the UN Division for Ocean Affairs and the Law of the Sea, locking in coordinates derived from coastlines as they stood at a point in time — for many states, around 2012. These baselines anchor the outer limits of the territorial sea, exclusive economic zone, and continental shelf, and with them, rights over fisheries and seabed resources worth far more than the atolls themselves.

In July 2025, the International Court of Justice issued an advisory opinion confirming that Pacific states face no legal obligation to update a deposited baseline as sea levels rise. On its face, this settled the fear driving years of Pacific advocacy: a state\'s maritime zones will not shrink simply because its physical coastline does. But the ruling did something else, too. By fixing the baseline in place, it fixed the coastline evidence behind that baseline as the permanent legal reference point — the record a tribunal would actually examine if any feature within it were ever contested.

Overnight, an old satellite pass, a colonial-era aerial photograph, a national survey plan sitting in an unindexed lands office — each one became load-bearing legal evidence rather than historical curiosity. That shift is why this matters now, and why it matters specifically to this room: if you hold shoreline imagery, drone captures, or survey data for a Pacific coastline, some of what is already sitting on your server may be the evidence a future case turns on, whether or not anyone has filed that case yet.

The South China Sea Arbitration (Philippines v. China, 2016) shows what a tribunal actually demands of that evidence once it\'s tested. To classify Scarborough Shoal, Mischief Reef, and a dozen other disputed features as islands, rocks, or low-tide elevations, the Permanent Court of Arbitration worked from satellite imagery, navigational charts, and historical surveys — and where China had since reclaimed land on top of a feature, the tribunal deliberately relied on evidence of its original, natural condition rather than its current, artificially altered state. A single striking image is not enough. What survives cross-examination is provenance: who captured it, when, under what conditions, with what independent corroboration, and with what quantified uncertainty rather than a confident-looking but unexplained line on a map. A defence team facing a challenge to a Pacific state\'s baseline needs exactly that — not one good picture, but a reproducible, dated, independently corroborated chain of evidence with its error margins stated up front, because an opposing counsel\'s first move will be to attack whatever isn\'t. A deposit is a claim, not proof, and neither is a single satellite scene.

For low-lying atoll states, this is precisely what does not yet exist: the baseline is legally frozen, but the evidentiary record supporting it is neither actively maintained nor built to the standard a tribunal will apply — even as the coastlines themselves keep visibly changing.

This presentation introduces a prototype geospatial pipeline built to close that gap — not by replacing a state\'s deposited baseline, but by building a corroborating evidentiary layer alongside it, closer in spirit to a Torrens-style title register than a new survey.

Using AI, the proposed pipeline fuses multiple independent data sources across a six-stage deterministic process: ingest, segment, reduce, fuse, seal, and emit. Multispectral optical imagery (Sentinel-2, Landsat) is segmented using a calibrated per-pixel water-probability model, rather than a hard threshold — a deliberate choice, because standard water-index thresholding fails on Pacific reef flats, where shallow, clear water sits over bright carbonate sand and produces false water/land boundaries. ICESat-2 photon altimetry is applied at controlling basepoints, anchored to a GNSS benchmark in the current global terrestrial reference frame, to solve the vertical-elevation gap that optical imagery alone cannot close on low-relief atolls. Tidal reduction draws on an ensemble of independent global tide models rather than a single source, so that model disagreement becomes an explicit, quantified part of the error budget rather than a hidden assumption. Every output record is hash-chained into an auditable, tamper-evident register, with a human reviewer in the loop on any contested or low-confidence feature.

We will share early results from a pilot shoreline reconstruction, the technical lessons learned building a pipeline for a tribunal audience rather than a research audience — where every design decision has to survive cross-examination, not just look accurate on average — and an open discussion of the governance model, data-sharing arrangements, and regional partnerships needed to move this from prototype toward a tool the Pacific geospatial community can use and maintain. We\'re especially keen to hear from anyone in the room already holding shoreline monitoring, survey, or lands data for their own jurisdiction — because whether or not it was built for this purpose, that archive may already be evidence.

12:25 – 12:45
Wilfred Moi
CCOP Geoinformation Sharing Infrastructure (GSi) Project and the PNG Geoscience Web Portal
New Technologies (AI, Deep Learning, Cloud)
The CCOP Geoinformation Sharing Infrastructure (GSi) Project provides a regional platform for integrating and disseminating geoscience data across East and Southeast Asia through web-based GIS technologies. As part of this initiative, Papua New Guinea (PNG) has developed a national GSi Portal to strengthen geospatial data management and interoperability within its geological and mineral resource sectors. This presentation introduces the technical framework and implementation of the PNG GSi Portal, focusing on its architecture, data integration processes, and GIS functionalities. The portal utilizes standardized geospatial services, enabling visualization, access, and basic spatial analysis of datasets including geological maps, mineral occurrences, and geohazard information. Emphasis is placed on system interoperability, data harmonization, and alignment with regional CCOP standards for geoinformation exchange. The PNG GSi Portal enhances institutional capacity in managing geoscience data and supports operational workflows in mineral exploration, land-use planning, and hazard assessment. It also contributes to regional data sharing by enabling seamless integration with the CCOP GSi network. https://geohazards-info.gsj.jp/gsi/pngwebgisportal/index.php
12:45 – 13:05
David Borges | Andiswa Mlisa | Nicholas Metherall | Sanjay Gowda | Brian Terry | Tommy Thomas | Oguz Yetkin
Digital Earth Pacific-GPT: Conversational Super Intelligence for Pacific Earth Observation
New Technologies (AI, Deep Learning, Cloud)
Digital Earth Pacific GPT (DE Pacific-GPT) is a collaboration between Digital Earth Pacific (DE Pacific) and the CEOS Systems Engineering Office (SEO) exploring how conversational Super Intelligence (SI) can make Earth observation data and analysis more accessible and useful for Pacific applications. DE Pacific provides analysis-ready Earth observation data and capabilities supporting decision-making across the Pacific. DE Pacific-GPT builds on this foundation by integrating EO-GPT, an open science framework that combines large language models with Earth observation data access, geospatial analysis, process orchestration, and interactive visualization. The goal is to allow users to move from a natural language question to relevant data, analysis, and interpretable results through a conversational interface. The pilot develops and demonstrates a set of Pacific-relevant analytical capabilities, including turbidity assessment, algal bloom analysis, change detection, and coastline analysis and querying existing Digital Earth Pacific and broader earth observation datasets. These capabilities can be invoked and combined through natural-language interactions. In parallel, the pilot advances the underlying architecture to support user accounts, multiple concurrent users, and scalable analysis of larger geographic areas. DE Pacific-GPT demonstrates a model in which SI does not replace established Earth observation tools or scientific methods but provides a more intuitive interface for accessing and orchestrating them. This approach has the potential to lower barriers to Earth observation analysis while retaining transparent, reusable analytical workflows. The presentation will introduce the DE Pacific-GPT architecture, demonstrate the pilot, discuss lessons learned, and explore opportunities to expand the approach to additional Pacific use cases and community-developed capabilities.
13:05 – 13:25
Gavin Jeter
Applicable AI: Scanned PDF to Geolocated Database
New Technologies (AI, Deep Learning, Cloud)
a { text-decoration: none; color: #464feb; } tr th, tr td { border: 1px solid #e6e6e6; } tr th { background-color: #f5f5f5; }
Across the Pacific, governments, utilities, researchers, and local authorities possess decades of valuable information locked inside scanned PDFs, handwritten forms, permits, service records, and archival documents. While many of these records have been digitized, they often remain inaccessible for search, analysis, and spatial decision-making. Recent advances in AI Vision models have transformed our ability to unlock this information. Modern multimodal AI can accurately interpret handwriting, forms, tables, and complex document layouts, making it possible to convert historical records into structured datasets at a scale that was previously impractical. This presentation showcases a practical workflow for transforming scanned PDF archives into geolocated databases using AI Vision APIs and FME. Using early 1900s New Zealand Motor Vehicle Registration Cards as the primary example, attendees will see how handwritten records can be converted into structured data and enriched for spatial analysis. A second example from UK Power Networks demonstrates how the same approach can be scaled to process large volumes of historical service record cards through automated workflows orchestrated by FME Flow. The session will explore how AI Vision models, structured outputs, and low-code/no-code data integration combine to create repeatable document-processing pipelines. Attendees will see how FME acts as an \"Any Data, Any AI\" platform, enabling organizations to integrate AI services, perform quality assurance, spatially enable extracted information, and automate processing from thousands to millions of documents. For organizations facing growing archives of scanned records, this presentation provides a practical roadmap for converting dormant document collections into searchable, structured, and spatially enabled information assets.
13:25 – 13:50
Coffee break
13:50 – 14:10
Andiswa Mlisa
Earth observations for UNCCD and SDG reporting and implementation in Pacific Small Island Developing States
Land Environment

The 2026 UNCCD reporting cycle has reinforced the value of Earth observations, geospatial data and national knowledge in supporting credible reporting on land degradation neutrality. For Pacific Small Island Developing States, this requires data that is fit for small island contexts and validated against national realities, including land cover change, coastal systems, food gardens, forests and restoration areas.

This presentation shares early insights from SPC’s role as the UNCCD Support Hub for the Pacific, supporting 14 Pacific countries during the 2026 reporting process. It will highlight how SPC is helping countries access and assess Earth observation datasets, use tools such as QGIS, Trends.Earth and PRAIS4, and combine global data with national datasets, field observations and local knowledge.

The presentation will also discuss practical support mechanisms, including regional training, bilateral consultations, recurring webinars, technical helpdesk support and guidance on dataset selection and quality assurance. A central message is that Earth observations can support more than compliance reporting: they can inform restoration and sustainable eland management priorities, strengthen policy coherence across the Rio Conventions, and build evidence for targeting investment in land, climate and biodiversity action.

By sharing SPC’s Support Hub experience, this presentation contributes to the Pacific GIS and Remote Sensing User Conference theme of strengthening and expanding geospatial networks in the Pacific Islands, and invites discussion on how regional platforms can turn Earth observation data into trusted, reusable information systems for national reporting and long-term resilience.

14:10 – 14:30
Bradley Eichelberger
Spatial ecology and remote sensing to estimate the population of globally endangered species: The Nosa Luta in the Northern Mariana Islands.
Land Environment
The Nosa Luta/Rota White-eye (Zosterops rotensis) is a critically endangered forest bird species found only on the island of Rota, Northern Mariana Islands. Given the high conservation status of the Nosa Luta, the Commonwealth of the Northern Mariana Islands (CNMI) Division of Fish and Wildlife (DFW) has been highly invested in continued population monitoring of the species to prevent its extirpation and facilitate its recovery. The CNMI DFW conducts biannual point count surveys in the species habitat (n=51) but also quarterly surveys at roadside stations (n=50) across the island and recently conducted a full scale island survey in 2022 (n=313). For each survey, an observer records individual birds that are seen or heard at various distance classes away from the station. By integrating data from the different surveys into hierarchical distance sampling models, we can leverage information from repeated site visits and distance bins to account for imperfect detection at both temporal and spatial scales. Models used precipitation, temperature, time and length of survey as predictors of availability and detectability; and vegetation and elevation characteristics derived from LiDAR. We selected the top models based on AIC and goodness of fit and present this as the updated global estimate for the species. Leveraging information from different surveys provided a robust population estimate useful for improving species population assessments and the CNMI DFW will continue to use hierarchical distance sampling to model abundance for all forest bird species in the future.
14:30 – 14:50
Nicholas Metherall
The impact of tropical cyclones Pam, Harold, Winston and Yasa on tree cover loss in Vanuatu and Fiji
Forestry
Many Pacific Small Island Developing States are vulnerable to Tropical Cyclones (TCs) leading to an estimated average annual loss of USD 1.08 billion. The study quantifies the impacts of tropical cyclones on tree cover and associated ecosystem services, beginning with coastal protection and the loss of carbon, for inclusion in Post Disaster Needs Assessment (PDNAs), Nationally Determined Contributions (NDCs), catastrophe risk insurance payments and loss and damage accounting. The study focuses on the impacts of tree cover loss resulting from four separate category five tropical cyclones in Fiji and Vanuatu: Pam, Harold, Winston and Yasa. Compared to national average annual tree cover losses between 2000 and 2023, TCs Pam and Harold increased tree cover loss 4.6 and 5.2-fold in Vanuatu and TCs Winston and Yasa increased tree cover loss 3.6 and 3.1-fold in Fiji, respectively. The resulting loss of carbon storage adds an estimated 23.4–25.0% in economic losses based on IPCC Tier II emissions factors and 37.2% for IPCC Tier I emissions factors to the Vanuatu and Fiji PDNA economic loss estimates, respectively. The focus on carbon emissions is a first step towards a quantification of the loss of ecosystem services in countries whose people depend on natural resources for daily subsistence. The study makes a case for inclusion of environmental damage in both PDNA and loss and damage estimates to justify additional financial investments in disaster recovery.
14:50 – 15:10
Elenoa Cibaiwasawasa | Vasemaca Naraukawa | Fayth Tikotani
GIS-Based Assessment of Coastal Hazard Vulnerability in Samoa Using Coastal Geomorphology and Land Use/Land Cover (LULC) Indicators
Remote Sensing data and software
Samoa is highly vulnerable to climate-induced coastal hazards, including tropical cyclones, storm surges, coastal erosion, sea-level rise, and coastal flooding, which have caused significant loss and damage to coastal communities, infrastructure, ecosystems, and livelihoods. This study assesses coastal hazard vulnerability and historical climate-induced loss and damage in Samoa from 2010 to 2020 using Remote Sensing and Geographic Information Systems (GIS). The focus is on vulnerable coastal communities and settlements located along Samoa\'s shoreline, particularly low-lying areas exposed to coastal hazards. Multi-temporal Landsat 8/9 and Sentinel-2 satellite imagery  integrated into digital earth pacific are used to map and analyse Land Use and Land Cover (LULC) changes within coastal zones, while historical shoreline datasets are applied to assess shoreline retreat, coastal erosion, and land loss. Site-level assessments examine geomorphological changes and the degradation or loss of natural protective ecosystems, including coastal forests, wetlands, mangroves, and coastal vegetation. GIS-based techniques such as shoreline change detection, buffer analysis, and hotspot mapping are employed to identify areas of elevated vulnerability. The study produces coastal erosion hotspot maps, LULC change maps, and a Coastal Hazard Vulnerability Map for Samoa. Findings will support climate adaptation planning, ecosystem-based adaptation, sustainable coastal management, and future loss and damage assessments.
18:00 – 22:00
Closing cocktail
2026-12-04
07:00 – 08:00
Registration
08:00 – 10:00
Gavin Jeter
Geospatial Data Integration for Disaster Response with FME and QGIS
Workshop
This workshop introduces practical geospatial data integration and automation through a realistic Pacific Islands disaster response scenario. Participants will explore how data from multiple sources and formats can be combined, transformed, validated, and automated using FME Form and FME Flow, then visualized and used for decision-making in QGIS. Through guided demonstrations and hands-on exercises, attendees will work with disparate datasets to develop a more complete operational picture of a disaster response situation, highlighting common challenges in geospatial data integration and the value of repeatable automated workflows. Designed for participants with little or no prior experience with FME or QGIS, the workshop focuses on practical problem-solving rather than software-specific expertise and demonstrates how modern geospatial technologies can help organizations turn disparate data into actionable information.
10:15 – 12:15
Md Sohan Khan | B. M. Refat Faisal
Assessment of sediment yield at lower Teesta River Basin of Bangladesh using geospatial technique
Land Environment

Sediment yield assessment is essential for understanding erosion and sediment transport in monsoon-dominated river systems, where strong seasonal hydrological variability controls sediment redistribution. This study evaluates seasonal soil erosion and sediment-delivery potential in the Lower Teesta River Basin, Bangladesh, a low-gradient floodplain reach of a transboundary Himalayan-fed river influenced by upstream sediment supply, floodplain storage, and channel processes. A semi-automatic geospatial framework integrating the Revised Universal Soil Loss Equation (RUSLE) and Sediment Delivery Ratio (SDR) was implemented in Google Earth Engine to estimate seasonal soil loss and sediment yield during 2018–2025. Multi-source remote sensing and geospatial datasets, including NASA DEM, Sentinel-2 imagery, OpenLandMap soil properties, and CHIRPS rainfall, were automatically processed to derive model parameters and seasonal erosion maps. Seasonal cover-management factors were estimated from Sentinel-2 NDVI composites, while soil erodibility and support-practice factors were derived from soil properties and land-use information. Mean modeled soil loss ranged from 1.02–1.71 t ha⁻¹ season⁻¹ during the pre-monsoon and 0.93–2.58 t ha⁻¹ season⁻¹ during the post-monsoon, with corresponding sediment yields of 0.37–0.61 and 0.34–0.93 t ha⁻¹ season⁻¹. Post-monsoon sediment yield was generally higher after 2020 and peaked in 2021. Evaluation against 20 Bangladesh Water Development Board observations from July–October 2019–2023 showed a positive temporal association between modeled and observed sediment yield (r = 0.778; R² = 0.605), although magnitude was substantially underestimated (NSE = −0.954; PBIAS = +59.58%). The framework therefore captures seasonal variability in sediment delivery but does not represent the complete fluvial sediment budget. Results should be interpreted as hillslope erosion and sediment-delivery potential, given that RUSLE and a basin-wide SDR do not explicitly simulate channel erosion, sediment storage, remobilization, hydraulic routing, or transport capacity. The approach provides a reproducible remote sensing framework for seasonal sediment assessment in data-scarce river basins.

13:15 – 15:15
David Borges | Brian Terry | Tommy Thomas | Sanjay Gowda | Andiswa Mlisa | Nicholas Metherall
Digital Earth Pacific-GPT: Transforming Earth Observation with Super Intelligence
Workshop

Discover how generative super intelligence (SI) is transforming the way we access, analyze, and visualize Earth observation data. This workshop will introduce Digital Earth Pacific GPT (DEP-GPT), an SI-powered platform that enables users to explore satellite imagery and perform geospatial analysis through natural language, making powerful Earth observation capabilities more accessible without requiring specialized programming or remote sensing expertise.

Through live demonstrations and real-world examples, participants will see how DE Pacific-GPT can support environmental monitoring and decision-making across the Pacific, including coastal change detection, water quality assessment, and algal bloom monitoring. The session will highlight the potential of AI to simplify complex geospatial workflows, accelerate scientific discovery, and expand the use of Earth observation data to address the Pacific region\'s environmental challenges.

2026 Conference Sponsors:



Event Partners