What Is Remote Sensing? Research Areas, Funding, and Career Paths

Remote sensing is the science of obtaining information about an object, surface, or process without being in physical contact with it, usually by measuring electromagnetic radiation (or, in some cases, sound) that the target reflects, emits, or scatters back to a sensor. In practice the term most often means Earth observation: instruments on satellites, aircraft, and drones that record how the land, oceans, ice, and atmosphere look across the spectrum, and the analysis that turns those recordings into maps, measurements, and time series. A single satellite image is a grid of numbers, one per pixel and per spectral band. Everything the field does, from tracking crop health to measuring the height of a forest canopy, is a way of turning those numbers into something a researcher can defend.

What Remote Sensing Actually Studies

Remote sensing is better understood as a measurement approach than as a body of subject matter. The subject matter belongs to the fields that use it, and the discipline’s own questions are about the measurement: what physical quantity does a sensor actually respond to, how does the atmosphere distort the signal, how finely can the sensor distinguish objects (its resolution), and how can a measured value be trusted enough to publish. The field is usually organised around a few recurring ideas.

• Electromagnetic interaction. Surfaces reflect, absorb, and emit energy differently at different wavelengths. Healthy vegetation, for example, absorbs strongly in red light and reflects strongly in the near-infrared, which is the physical basis for vegetation indices such as the Normalized Difference Vegetation Index (NDVI).
• Passive and active sensing. Passive sensors record energy that already exists, mostly reflected sunlight or thermal emission from the surface. Active sensors, such as radar and lidar, send out their own pulse and time or measure the return, so they work at night and, for radar, through cloud.
• The four resolutions. Spatial resolution (the ground size of a pixel), spectral resolution (how many bands and how narrow), temporal resolution (how often a place is revisited), and radiometric resolution (how finely brightness is quantised). Every sensor design trades these against each other, so no instrument is best at all four.
• Calibration and correction. Raw sensor values must be converted to physical units and corrected for atmosphere, viewing geometry, and terrain before images from different dates or sensors can be compared.
• Validation. A remotely sensed product is an estimate. Its accuracy is assessed against reference data, often collected in the field, which is why remote sensing and fieldwork are partners rather than alternatives.

Platforms and Sensors: Optical, Radar, and Lidar

Sensors are classified by the part of the spectrum they use and by whether they supply their own energy. Three families account for most research use.

Optical and Multispectral Sensors

Optical sensors record visible, near-infrared, and shortwave-infrared light, and some also record thermal infrared emitted by the surface. A multispectral sensor records a handful of broad bands. A hyperspectral sensor (also called an imaging spectrometer) records dozens to hundreds of narrow, contiguous bands, which lets researchers identify minerals, vegetation traits, and water constituents from their spectral signatures. The main limitation of optical sensing is that it needs daylight and a clear view: clouds and haze block it.

Radar and Synthetic Aperture Radar

Radar sensors transmit microwave pulses and record the echo. Synthetic aperture radar (SAR) uses the motion of the platform to synthesise a much longer antenna than could be flown physically, which produces fine-resolution images. Because microwaves pass through cloud and do not depend on sunlight, SAR can image a place in any weather, day or night. Radar backscatter responds to surface roughness, moisture, and structure rather than colour, so it is used for flood mapping, sea-ice monitoring, forest structure, and, through interferometric SAR (InSAR), for detecting millimetre-to-centimetre changes in ground height from subsidence, earthquakes, and volcanoes.

Lidar

Lidar (light detection and ranging) fires laser pulses and measures the time for each to return, producing a three-dimensional point cloud. Airborne lidar is the standard method for building high-resolution terrain models, including under forest canopy, because some pulses reach the ground through gaps. Terrestrial and drone-mounted lidar serve smaller sites, and spaceborne lidar is used to measure ice-sheet elevation and forest height. The result is direct information about vertical structure that passive imagery can only infer.

Airborne, Drone, and Satellite Platforms

The platform sets coverage and detail. Satellites give repeated, consistent coverage of the whole planet. Crewed aircraft carry specialised or high-resolution instruments over chosen areas. Uncrewed aircraft systems (drones) have become a routine research tool for centimetre-scale mapping of a field plot, a glacier front, or an archaeological site, and a growing number of studies pair them with satellite data. Ground-based instruments, such as handheld spectrometers and tower-mounted cameras, provide the reference measurements that the aerial and orbital data are checked against.

Landsat and Sentinel: The Public Satellite Workhorses

Two public programmes supply a large share of the imagery used in published research, and both are verifiable, well-documented starting points for anyone new to the field.

• Landsat is a joint programme of NASA and the U.S. Geological Survey. Landsat 1, then called the Earth Resources Technology Satellite (ERTS), launched on July 23, 1972, and was the first Earth-observing satellite launched specifically to study and monitor the planet’s land. Landsat 9 launched on September 27, 2021, from Vandenberg Space Force Base in California. The result is a continuous record of the land surface stretching back more than five decades, which is what makes Landsat so valuable for studying long-term change. Landsat multispectral imagery has a 30-metre pixel size.
• Sentinel satellites are operated for the European Union’s Copernicus programme, with the European Space Agency developing the space segment. Sentinel-1A, a radar satellite, launched on April 3, 2014, and Sentinel-2A, an optical multispectral satellite, launched on June 23, 2015. Sentinel-1 provides C-band SAR; Sentinel-2 images the land in multiple bands at up to 10-metre resolution, which complements Landsat’s longer record with finer detail and frequent revisits.

Used together, the two programmes are a common basis for land-cover mapping, crop monitoring, and change detection. Many other missions matter as well, including NASA’s Terra, Aqua, and other Earth Observing System satellites, commercial constellations that sell high-resolution imagery, and national missions from agencies around the world. The Committee on Earth Observation Satellites (CEOS) coordinates civil Earth-observation programmes at the international level.

Image Analysis and Methods

Getting from a raw scene to a research result follows a fairly standard chain, whatever the application.

1. Pre-processing. Radiometric calibration to physical units; atmospheric correction to estimate surface reflectance; geometric correction so that pixels line up with map coordinates; and, for radar, speckle filtering and terrain correction. Many agencies now distribute analysis-ready data so that users can skip some of these steps.
2. Feature extraction and indices. Combining bands into indices (NDVI for vegetation, water indices, burn indices), computing texture, or deriving height from lidar and interferometric data.
3. Classification and mapping. Assigning each pixel or object to a class such as forest, cropland, or water. Approaches range from traditional supervised and unsupervised classification and object-based image analysis to machine learning and deep learning models that learn spatial patterns directly from imagery.
4. Time-series and change detection. Comparing images across dates to find deforestation, urban growth, flooding, or the greening and browning of a season. Dense time series from Landsat and Sentinel make it possible to model the trajectory of a pixel rather than just two snapshots.
5. Accuracy assessment. Comparing the map against independent reference data and reporting error rates, typically with a confusion matrix. A map published without an accuracy assessment is a picture, not a measurement.

Most of this work is done inside geographic information systems and scripting environments. Remote sensing and GIS are close partners: GIS stores and overlays the spatial layers, and remote sensing is one of the main ways those layers are produced. The processing is also increasingly done in the cloud, because the archives are now far larger than a single laptop can hold, and it leans heavily on data science methods. For the geographic framing that surrounds these tools, see CASRAI’s guide to geography.

Applications in Earth Science, Agriculture, and Beyond

Earth and Environmental Science

Remote sensing is a standard instrument across the Earth sciences. Geologists use multispectral and hyperspectral imagery to map rock and mineral units, and InSAR to measure ground deformation, which connects it to geology and geophysics. Hydrologists estimate snow cover, surface water extent, and soil moisture from space, which is the observational side of hydrology. Oceanographers use satellites to measure sea-surface temperature, ocean colour, and sea level, as covered in oceanography, and atmospheric scientists use them for clouds, aerosols, and trace gases, which supports atmospheric science and meteorology. Ecologists and conservation scientists map habitat, track land-cover change, and measure vegetation structure; see ecology.

Agriculture and Forestry

Agriculture is among the most visible applications. Time series of vegetation indices show crop emergence, vigour, and stress across a season; thermal and radar data help estimate water stress and soil moisture; and classified maps support crop-type inventories and yield modelling at regional scale. At the field scale, drone and tractor-mounted sensors support precision agriculture, where inputs such as fertiliser or irrigation are varied within a field. The research context is described in CASRAI’s agricultural science guide. Forest monitoring uses the same toolkit, with lidar and radar adding canopy height and biomass estimates; see forestry.

Hazards, Cities, and Climate

Optical and radar imagery are used to map floods, wildfires, landslides, and earthquake damage soon after an event, and to build the long baselines against which a hazard is later judged. Urban researchers map land use and the growth of built-up areas. Climate research relies on long, consistently calibrated records of temperature, ice extent, sea level, and vegetation. Cultural-heritage researchers use imagery and lidar to find and document sites that are invisible from the ground, a use that overlaps with archaeology.

A Short History of the Discipline

Remote sensing begins with photography from the air. Cameras were flown on balloons and kites in the nineteenth century, and even on pigeons in the early twentieth,, and aerial photography became a systematic mapping and military tool during the First and Second World Wars. Photogrammetry, the science of measuring from photographs, matured alongside it; the American Society of Photogrammetry was founded in 1934. The term “remote sensing” itself is generally credited to the 1950s and 1960s, as researchers began using non-photographic sensors such as radar and infrared scanners.

The space era moved the field from aircraft to orbit. Meteorological satellites came first; the launch of Landsat 1 in 1972 then gave scientists a dedicated land-observing system and a long-term archive. Radar and lidar matured from the 1970s onward, and hyperspectral imaging developed through airborne instruments before reaching orbit. A decisive shift for research came in 2008, when the U.S. Geological Survey made the Landsat archive available at no cost; the agency has reported that the number of scenes distributed per day rose from 53 to 5,775 after the change. Since then free and open policies, including Copernicus’s open data approach for Sentinel, together with cloud computing and machine learning, have turned remote sensing from a specialist activity into routine research infrastructure.

Societies, Journals, and Professional Organizations

• American Society for Photogrammetry and Remote Sensing (ASPRS). Founded in 1934 as the American Society of Photogrammetry and renamed in 1985. Its journal, Photogrammetric Engineering & Remote Sensing (PE&RS), is published monthly and covers photogrammetry, remote sensing, and related geospatial science.
• IEEE Geoscience and Remote Sensing Society (GRSS). The IEEE society for the field. It publishes IEEE Transactions on Geoscience and Remote Sensing, IEEE Geoscience and Remote Sensing Letters (short papers), and IEEE Geoscience and Remote Sensing Magazine.
• Other widely read journals include Remote Sensing of Environment, the ISPRS Journal of Photogrammetry and Remote Sensing, and the International Journal of Remote Sensing. The International Society for Photogrammetry and Remote Sensing (ISPRS) is the international scientific body for the field.

Earth-science preprints are commonly posted on EarthArXiv; see CASRAI’s explainer on EarthArXiv.

Who Funds Remote Sensing Research

Because remote sensing serves so many missions, funding is spread across agencies.

• NASA builds and operates many of the Earth-observing missions and funds the science that uses them, mainly through its Earth Science programmes and competitive solicitations. For early-career routes, see the NASA Postdoctoral Program and NASA FINESST, the graduate-student fellowship solicitation.
• U.S. Geological Survey (USGS) operates the Landsat archive with NASA and distributes data through its EROS Center. See the guide to the USGS ScienceBase repository.
• National Science Foundation (NSF) supports remote sensing methods and applications through its geosciences, engineering, and computing programmes, as described in the NSF dictionary entry. To prepare a proposal, see how to apply for an NSF grant, and for training grants, the NSF Graduate Research Fellowship and NSF REU.
• NOAA, the Department of Defense, and other mission agencies fund weather, ocean, and operational applications, and in Europe the European Space Agency and the European Union’s Copernicus programme fund missions and user-oriented research. Private foundations and companies fund applied work, especially in agriculture and climate.

Specific programme names, solicitation numbers, and eligibility rules change often, so check the current announcement before building a proposal around any of them.

Data Access and Open Data

Much of the data in this field is free, which is a large part of why it is so widely used. The main public sources are:

• NASA Earthdata and NASA’s Distributed Active Archive Centers (DAACs) distribute data from NASA’s Earth-observing missions and Earth science programmes. For a NASA domain repository CASRAI has covered in detail, see the NASA Planetary Data System.
• USGS EROS distributes the Landsat archive at no cost, following the 2008 free-and-open policy.
• Copernicus provides free and open Sentinel data; since January 24, 2023 the Copernicus Data Space Ecosystem has been the main portal for them. For Copernicus climate reanalysis and projection data, see the Copernicus Climate Data Store (C3S).
• NOAA archives climate, ocean, and weather observations; see NOAA NCEI.

Sharing the products you make is part of the job too. Derived maps and time series should carry metadata that records the source scenes, processing steps, coordinate reference system, and accuracy. The principles behind this are set out in the FAIR Data Principles and open data entries, and the choice of where to deposit is covered in how to choose an open data repository.

Training and Career Paths

There is no single undergraduate route. Remote sensing is taught within geography, geosciences, environmental science, forestry, agriculture, civil and electrical engineering, and computer science departments, and strong candidates usually combine a domain (such as hydrology or ecology) with quantitative skills: statistics, programming, and the physics of radiation. Many universities offer professional master’s degrees and certificates in GIS and remote sensing, and research master’s and PhD degrees are the usual route for academic and agency research positions. Typical working titles include remote sensing analyst, GIS analyst, geospatial data scientist, image scientist, and photogrammetrist, with employers in government agencies, environmental and agricultural consultancies, utilities, insurers, and technology and space companies. ASPRS offers certification for geospatial professionals; check its current requirements before planning around a specific credential.

Where Research Administration Comes In

Remote sensing projects generate a set of administrative questions that research offices see regularly.

• Data management and sharing. Imagery archives are large, and derived products need long-term storage and citation. Funders expect a plan; see the data management plan entry and NIH vs. NSF data management plans.
• Licensing. Public data such as Landsat and Sentinel are free, but commercial imagery often carries licences that limit redistribution, which can affect what can be deposited in a repository or published as supplementary material.
• Privacy and human subjects. Very high resolution imagery and drone surveys can capture people and private property. Projects that link imagery to individuals or households generally need review by an institutional review board.
• Drone operations. Uncrewed aircraft flights are regulated, and an institution typically sets its own approval and insurance requirements on top of national rules, so fieldwork plans should involve the research office early.
• Export controls and international work. Some sensors, high-resolution data, and collaborations abroad raise export-control and data-transfer questions that sponsored programs and compliance staff will want to review.

Frequently Asked Questions

What is remote sensing in simple terms?
It is measuring something from a distance, usually from a satellite, aircraft, or drone, by recording the energy that reaches a sensor from the target and then analysing those measurements to learn about the surface or atmosphere.

What is the difference between remote sensing and GIS?
Remote sensing produces spatial data by measuring from a distance. GIS stores, overlays, and analyses spatial data from any source. The two are used together, and remote-sensing imagery is one of the main inputs to a GIS.

What is the difference between optical, radar, and lidar sensing?
Optical sensors are passive and record reflected sunlight or thermal emission; they need daylight and clear skies. Radar sends microwave pulses and works through cloud and at night. Lidar sends laser pulses and measures return time to build a three-dimensional picture, especially of terrain and vegetation structure.

Is Landsat and Sentinel data free?
Yes. USGS distributes Landsat at no cost under a policy announced in 2008, and Copernicus provides Sentinel data free and open through the Copernicus Data Space Ecosystem and related services.

How is remote sensing used in agriculture?
Mainly to monitor crop condition and extent: vegetation indices from optical imagery show vigour and stress, radar and thermal data help estimate moisture, and classified maps support inventories and yield models. At the field scale, drones guide variable-rate management.

How accurate is a remote-sensing map?
It depends on the sensor, the method, and the landscape, which is why an accuracy assessment against independent reference data is a standard part of a credible study.

Where Remote Sensing Fits Among the Sciences

For a broader map of how remote sensing relates to the other major disciplines, see CASRAI’s branches of science guide, the anchor index for this series of discipline overviews.