# Observation

Observation means using space to learn something about Earth. Satellites can help measure weather, water, vegetation, ice, land use and activity at sea. The useful result is not always a picture. It may be a forecast, a map of flooded land or a record showing how a place changed over time.

The user usually needs to make a decision. To understand this part of the sector, follow the information from that decision back to the measurement that supports it.

## 1. Users: who needs to know something about Earth?

Observation serves people with very different jobs. A weather forecaster needs measurements of the atmosphere. A civil protection team needs to understand a disaster. A land manager needs to track changes in a large area. A farming adviser wants to identify fields that deserve closer attention.

The user, buyer and funder can be different. A publicly supported observation programme may provide data that a private company turns into a paid service. A public authority may also buy a finished map from a contractor. The existence of freely accessible data therefore does not mean every activity in the chain is free.

| User | The decision or task | What may be useful |
|---|---|---|
| Weather forecaster | Describe likely future weather | Atmospheric measurements used in forecast models |
| Civil protection team | Understand an emergency and plan response | Flood maps, exposure information and updates |
| Farmer or adviser | Decide where to inspect or change a practice | Field observations combined with local information |
| Land or environmental manager | Track change and assess conditions | Comparable measurements over time |
| Marine operator | Understand ocean conditions or plan activity | Ocean observations combined with navigation and weather information |

These are application categories, not a list of customers interviewed for this research. EUMETSAT describes weather, climate and environmental monitoring, while Copernicus emergency services support disaster decisions. [EUMETSAT: what it does](https://www.eumetsat.int/about-us/what-we-do), [Copernicus emergency service](https://emergency.copernicus.eu/about/)

## 2. Needs: what must the information tell the user?

“Better images” is not a complete need. Ask what the user must know, over which area, how often and before what deadline. A map arriving after a decision has been made may have little value for that decision, even if the map is accurate.

Three useful terms are **detail**, **frequency** and **delay**. Detail means how small a feature the observation can distinguish. Frequency means how often a location is observed. Delay means how much time passes between measurement and useful information reaching the user. These properties are different. A detailed image does not automatically arrive frequently or quickly.

A fourth question is confidence. Can the user tell what the measurement supports and what remains uncertain? A map showing likely flooding is not a guarantee that every road marked outside the flood area is safe. In practice, satellite information must be interpreted with local conditions and other evidence. This is the decision framework used in this guide, not an operating instruction for emergency response.

## 3. Services: from measurements to something people can use

There are several stages of service. A data provider can supply an image or a measured value. Another provider can process that material into a map. A specialist can interpret it for a sector, and software can place the result inside the user’s existing work.

Weather forecasting is a good example. A satellite does not directly transmit tomorrow’s finished forecast. Measurements are checked and combined with other observations and a model of the atmosphere. ECMWF describes this process as turning observations into predictions. [ECMWF: from satellite data to forecasts](https://www.ecmwf.int/en/about/media-centre/news/2026/satellite-data-weather-forecasts)

Copernicus Emergency Management Service is another real example. It provides early warning and monitoring, mapping on request and exposure information. Its products and services are free of charge. Activation of the on-demand mapping service is restricted to authorised users, typically civil protection institutions. Open access to a product is different from having authority to request a new emergency activation. [Copernicus: services and activation](https://emergency.copernicus.eu/about/)

A potential supplier needs to understand what is already available before proposing to sell a similar map.

## 4. Systems: how a measurement becomes information

Start with an instrument on a satellite. An optical instrument measures light; radar sends a signal and measures what returns. These are different ways of observing, with different strengths. ESA describes Sentinel-1 as providing radar imagery by day and night in all-weather conditions, and Sentinel-2 as providing optical imagery for land applications. Radar is useful through cloud, but its images require appropriate processing and interpretation. [ESA: the Copernicus space component](https://www.esa.int/Applications/Observing_the_Earth/Copernicus/Space_Component)

After collection, the data must reach Earth. Ground facilities receive it, check it, process it and make it available. EUMETSAT’s ground system includes teams controlling satellites and associated infrastructure around the clock. Observation therefore depends on continuing service operations, not only building and launching an instrument. [EUMETSAT: ground systems](https://www.eumetsat.int/ground-segment)

Further processing may turn those observations into a forecast, a change map or another product. A useful chain is: **measure → receive → check → process → interpret → deliver → decide**. This is a simplified teaching model. Real systems can contain several organisations and repeated checks at each stage.

## 5. Enablers: what makes the observations trustworthy and usable?

A measurement must be comparable with other measurements. Instruments need calibration: checking how their readings relate to known references. Analyses need validation: checking whether the result agrees with suitable independent evidence. NASA’s Landsat material explains the value of a continuing, consistent record and the use of its data by other satellite companies for calibration. [NASA: Landsat’s long observation record](https://science.nasa.gov/mission/landsat/)

Users also need access to storage, computing and clear documentation. The Copernicus Data Space Ecosystem provides access to data and processing capabilities, with opportunities for other providers to build complementary services. “Open data” does not mean unlimited computing or no conditions of use. The available capacities and terms still matter. [Copernicus: data and processing ecosystem](https://documentation.dataspace.copernicus.eu/Ecosystem.html)

Finally, someone must connect the result to the user’s work. A farming adviser may combine satellite observations with soil information and a field visit. The EU CAP Network identifies both the value of combining sources and the need for more usable, farmer-centred solutions. This is evidence of adoption challenges; it is not a measured market size for a new application. [EU CAP Network: sensing for farmers and advisers](https://eu-cap-network.ec.europa.eu/publications/factsheet-sensing-future-practical-applications-proximal-and-remote-sensing_en)

## Journey one: understanding a flood

Imagine an authorised civil protection organisation needs an overview of a flood. It requests mapping through the appropriate service. Satellite observations are processed into information about affected areas. Staff compare the product with local reports, roads, buildings and other information before deciding what to do next.

The user’s need is an overview they can use in time. The service is mapping and related information. The system includes observations, processing and delivery. Its enablers include data access, trained people, local reference information and an established activation process.

This journey explains the dependency chain; it is not a case study of one named flood. The real service behind it is documented by Copernicus. Its publication of maps does not by itself establish that another organisation would pay for an additional mapping service. [Copernicus: emergency mapping documentation](https://documentation.dataspace.copernicus.eu/Data/CopernicusServices/CEMS.html)

## Journey two: helping a farmer inspect the right field

A farmer may notice uneven growth across a large area. A satellite-based service can show differences that help an adviser decide where to inspect. A field visit and other measurements help determine what those differences mean. A visible change is not automatically a diagnosis or a recommendation to apply a particular treatment.

The paying customer could be a farm, an adviser, a cooperative or another organisation providing farm services. That buyer must be established through research. It cannot be inferred from the number of farms or the amount of available imagery.

This example also links two motivations. Observation can help identify **where attention is needed**. Positioning can help a machine return to **the intended place**. Their capabilities work together, but answer different questions. [EUSPA: space and agriculture](https://www.euspa.europa.eu/euspace-and-agriculture)

## Alternatives, Europe and the wider world

Alternatives include field visits, ground sensors, aircraft, drones and existing administrative records. They may provide more local detail or a different kind of evidence. Satellites are particularly useful for wide areas and repeated observation, but often complement those alternatives rather than replace them.

Europe provides an accessible base through Copernicus. Belgium also supports Earth observation research through BELSPO’s STEREO programme, which includes work with users and transfer of knowledge to industry and public bodies. These are practical institutions to investigate, not promises of funding for any proposed company. [BELSPO: STEREO](https://www.belspo.be/belspo/space/belspo_stereo_en.stm)

## What to investigate as a business

**Technical:** can the chosen data reveal the feature the customer cares about? **Operational:** does the information arrive often enough and before the decision? **Commercial:** does it improve a real workflow enough for someone to buy it? **Economic:** can the supplier cover processing, expert review, sales and support?

The sources support investigating integration and usability. Candidate questions include whether customers repeatedly combine data by hand, struggle to judge uncertainty or cannot fit existing products into their systems. These are hypotheses to test. A useful next interview asks for a recent decision, the evidence used, what went wrong and what the organisation already pays for.

The central idea is: **observation becomes useful when a measurement helps someone make a better-informed decision.**