Written by 
Remote sensing has changed the way we understand our planet. Instead of relying solely on measurements collected by people working directly in the field, we can observe enormous areas of the Earth’s surface from aircraft, drones and satellites. Forests, cities, oceans, agricultural fields, glaciers and even atmospheric processes can all be studied using remotely sensed data.
But remote sensing is not simply about taking pictures of the Earth. The real value lies in turning measurements into information, and ultimately into knowledge that can support scientific research and practical decision-making.
What do we mean by remote sensing?
At its simplest, remote sensing involves collecting information about an object or environment without direct physical contact. ASPRS describes remote sensing as techniques used to gather and process information about an object without direct physical contact.
The technology is based on electromagnetic radiation. Different materials interact with electromagnetic energy in different ways: they absorb, reflect or emit energy according to their physical and chemical properties.
Human vision is sensitive to only a small portion of this electromagnetic spectrum, approximately 400–700 nanometres in the visible range. Remote-sensing instruments can operate beyond human vision, including in the infrared and microwave portions of the spectrum.
This is one of the fundamental advantages of remote sensing. A healthy plant, for example, may appear simply green to our eyes, but its interaction with near-infrared and red wavelengths contains information about vegetation condition. Similarly, radar can provide information that optical imagery cannot, including observations through cloud and in darkness.
Passive and active sensors
Remote-sensing systems can broadly be divided into passive and active sensors.
Passive sensors measure naturally occurring energy, usually sunlight reflected from the Earth’s surface or thermal radiation emitted by objects. Multispectral satellite missions such as Sentinel-2 are examples. Sentinel-2 carries a multispectral instrument with 13 spectral bands, providing observations at spatial resolutions down to 10 metres.
Active sensors generate their own energy and measure the response from the surface. Radar is a particularly important example. Because radar supplies its own illumination, it can operate day and night and, depending on wavelength and conditions, can acquire observations through cloud.
This distinction matters because choosing a sensor is essentially choosing what physical properties we want to measure.
Resolution is more than just pixels
One of the most commonly misunderstood aspects of satellite imagery is resolution.
Spatial resolution describes the size of the area represented by each pixel. Spectral resolution describes the ability of an instrument to distinguish between different wavelengths. Temporal resolution concerns how frequently an area can be observed, while radiometric resolution relates to how finely differences in measured energy can be represented.
These characteristics involve trade-offs.
Sentinel-2, for example, combines a 290 km swath with 13 spectral bands and frequent revisit capability, making it particularly valuable for monitoring land and vegetation over large areas.
A very high-resolution image may show individual buildings or vehicles, but cover a relatively small area or require more resources to acquire and process. A coarser-resolution sensor may not distinguish individual objects but can provide consistent observations across continents.
There is therefore no universally “best” remote-sensing dataset. The appropriate data depend on the question being asked.
From raw measurements to information
The journey from sensor measurement to useful information usually involves several stages.
First, the data need to be calibrated and geometrically corrected. Atmospheric effects may need to be considered for optical observations. Images may then be aligned, mosaicked, classified or transformed into derived products.
Indices can provide another layer of interpretation. Vegetation indices, for example, exploit differences in spectral response to provide information related to vegetation properties.
The result can be a map showing forest cover, crop condition, water extent, urban expansion or burn severity rather than simply a collection of pixels.
This distinction between data and information is important.
A satellite does not directly “see” a forest disease, drought or urban expansion. It measures electromagnetic responses. Scientists and analysts then develop methods for relating those measurements to physical phenomena.
The importance of time
Perhaps one of the greatest strengths of Earth observation is repetition.
A single image tells us what a landscape looked like at one particular moment. A time series can tell us how that landscape is changing.
The Landsat programme is an exceptional example. Landsat has provided continuous observations of Earth’s land surface since 1972. Research reviewing its first 50 years identified major contributions to agricultural mapping and water use, climate-change research, ecosystem monitoring and understanding the changing human footprint.
The availability of consistent historical observations means researchers can move beyond asking “what is here?” to asking “what has changed, when did it change, and how quickly?”
That shift is fundamental to Earth observation.
From observations to decisions
Remote sensing becomes particularly powerful when combined with other information.
Satellite observations can be integrated with field measurements, meteorological records, GIS databases, digital elevation models, socio-economic information and models. Increasingly, machine learning is also being used to extract patterns from large datasets.
The objective is not simply to produce increasingly impressive maps. It is to produce information that answers meaningful questions.
Where has flooding occurred?
Which crops are experiencing stress?
How quickly is a glacier retreating?
Where has forest cover been lost?
Which areas are vulnerable to wildfire?
These are questions where remote sensing becomes part of a wider evidence system.
Looking ahead
The future of remote sensing is likely to involve greater integration between sensors, platforms and analytical methods. Optical imagery, radar, LiDAR, hyperspectral data, photogrammetry and other sources can increasingly be combined.
At the same time, cloud computing and artificial intelligence are changing how enormous archives can be processed.
For the remote-sensing community, the challenge is therefore no longer simply acquiring more data. It is understanding what those measurements mean, quantifying uncertainty and ensuring that derived information is scientifically robust.
Remote sensing gives us the ability to observe our planet at extraordinary scales. The real scientific challenge and opportunity lies in turning those observations into reliable knowledge.