Between satellites and conservation lies a missing middle — people who return, observe, verify, and translate. We have never been able to watch the Earth more than we do today. Yet one question remains unanswered.
Why does having more data not necessarily give us a greater ability to protect the Earth?
Satellites circle the planet every day. AI can detect change at global scale. Drones, sensors, and automated systems are expanding observation capacity at a pace once unimaginable.
Yet field sampling, long-term observation, local interpretation, action verification, community feedback, and data governance still depend on scattered people and institutions — a fragile, underinvested human layer.
The GEO In-Situ Data Strategy recognises that both satellite-based and in-situ observations are the foundation of Earth Intelligence, and directly addresses the challenges of field data access, integration, and use.
"Between observation and action lies a missing middle. This is a story about the people who return."
— Ground Truth, Earth LinCBetween satellite observation and meaningful conservation action lies a gap: the humans who return to the same place, verify what sensors detect, and translate data into decisions that protect living systems.
Entre l'observation satellitaire et l'action de conservation concrète, il manque un maillon humain : des personnes qui reviennent sur place, vérifient ce que les capteurs détectent, et transforment les données en décisions de protection.
在卫星观测与实际保护行动之间，存在一个缺口：那些回到同一地点、核实传感器所探测到的信号、并将数据转化为保护决策的人。这才是真正的Ground Truth。
Three signals from the evidence — three specific, measurable absences at the heart of Earth observation.
UNEP-WCMC and UNDP studied 365 indicators of the Kunming-Montréal Global Biodiversity Framework. Citizens, Indigenous peoples, and local communities can improve data collection for 51% of indicators — and for 30%, they can also participate in analysis and interpretation.
A global surface albedo validation study compiled only 416 field sites from a "sparsely distributed global observation network." A single ground point cannot represent a km-scale satellite pixel in heterogeneous terrain. Combining temporal and spatial data reduced RMSE by 6.04% globally — and 17.09% in complex landscapes.
Research on three London biodiversity recording platforms found most volunteers contribute few records and are active for only a single day. Data shows clear species, place, and time bias — common, easy-to-reach species accumulate records; rare and remote ones do not. Continuity, not participation, is scarce.
"The real gap is not the absence of people willing to join a BioBlitz. It is the absence of people willing to return — continuously supported, following a protocol, maintaining quality, and able to interpret change."
— Inspired by the long records of Concord and Broadbalk: one observation is common; continuity is rare.Remove any link and the chain breaks: signal without verification, data without context, knowledge without a path into decisions.
The world is not simply missing field workers. It is missing the full human capability chain — from verified observation through interpretation, to feedback that confirms whether action made a difference.
Il ne manque pas simplement des observateurs de terrain. Il manque toute la chaîne humaine : de l'observation vérifiée jusqu'à l'interprétation et au retour d'information confirmant si l'action a produit un effet.
世界缺少的不只是现场人员，而是一条完整的能力链：从经过验证的观察，到解释，再到能够确认行动是否有效的反馈回路。
Progress varies enormously — from near-national career systems to regions where the coordinator layer barely exists.
From volunteer monitoring to formal employment. Over 320 ranger groups, 159 organisations. AUD 1.3B+ government investment 2021–2028. Traditional Knowledge and technical monitoring combined.
Connect ranger data with EO, fire, coast, biosecurity, and climate models while ensuring Indigenous data sovereignty. Monitoring Country is building standardised tools.
240+ First Nations, Inuit, and Métis Guardian initiatives since 2018. ~1,500 culturally meaningful jobs. Guardians are officially described as the "eyes and ears" on lands, waters, and ice.
Convert short-term contribution agreements into long-term career stability. $8M/year for community-based climate monitoring. Youth entry, data sovereignty, cross-community networks.
Large volunteer monitoring platforms, national species recording traditions. EEA uses volunteer observations for policy indicators. Digital infrastructure and training relatively mature.
Methods inconsistent across projects. Rare species under-recorded. Local data struggles to enter official systems. Volunteer retention is fragile. Need paid coordinators, validators, taxonomy specialists.
Kenya Bird Map, regional training, local manager networks emerging. Citizen Science for Conservation in Africa trained 26 managers from 9 countries in protocol design, bias identification, and policy translation.
WMO: African ground station density far below global standards. The bottleneck is not observer count but the coordinator layer — people who design protocols, train participants, analyse data, and move results into decisions.
In one Philippine study, 156 interventions emerged from community monitoring — 22% targeted local resource-use rules. After one community reported declining fish, local government restricted gear; seven months later fish abundance increased.
Monitoring and local governance are not separate. Don't treat communities as cheap sensors — involve them in problem definition, interpretation, rule-making, and results use. Indigenous Navigator operates in 11 countries.
Existing forces distributed across nature reserves, forest/grassland monitoring, river-lake chief systems, nature education, university field stations, citizen science, and NGO projects.
The opportunity is not to create a new organisation, but to connect existing roles: protected areas + research stations + NGOs + schools + communities + farmers — linked by a shared, non-burdensome local protocol and long-term open archives.
Six layers every credible observation system needs — and three pathways to build it.
Help ordinary people find an entrance — a first observation, a species identification, a rain gauge reading. But volunteers cannot carry the full institutional burden. Entry matters; continuity is the harder design problem.
Real payment for continuous observation, equipment maintenance, data management, and verification. Canada and Australia have already demonstrated this is not imagination — it is infrastructure. Paid stewardship is the structural backbone.
People who can move between Earth Observation, AI, field science, community knowledge, and decision-making — translators across layers. The EnviroDNA eDNA Field Technician role is one concrete example already in the market.
AI is becoming fast. Sensors are spreading. But the thread between observation and action — the person who returns, verifies, interprets, and feeds back — is not yet built into the architecture of how we govern this planet.
Ground Truth is a distributed human infrastructure. It is not built yet. That is the work.
La vérité terrain est une infrastructure humaine distribuée. Elle n'est pas encore construite. C'est le travail qui reste.
Ground Truth 是一套分布式的人类基础设施。它尚未建立。这，正是接下来要做的事。