Remote sensing is the process of collecting information about Earth’s surface from a distance, typically using sensors mounted on satellites, aircraft, or drones. In a GIS (Geographic Information System), this remotely captured imagery serves as one of the most important data sources, providing the raw spatial information that analysts layer, process, and interpret to study everything from crop health to urban climate risks. Where GIS is the analysis engine, remote sensing is the eyes.
How Remote Sensing Feeds Into GIS
A GIS combines many types of spatial data: road networks, property boundaries, elevation models, census figures. Remote sensing adds a continuous, image-based layer that covers large areas and can be updated on a schedule. A single satellite pass can capture land surface conditions across thousands of square kilometers, and that imagery flows into GIS software where it’s stacked with other datasets for analysis. An urban planner might overlay a thermal satellite image showing surface temperatures with a census layer showing elderly populations to build a heat vulnerability index. A farmer might combine a near-infrared drone image of a wheat field with soil survey data to decide where to apply fertilizer.
The integration goes both directions. GIS tools handle the heavy analytical lifting: classifying land cover types from satellite pixels, calculating vegetation indices, or running spatial statistics. Remote sensing provides the raw observations that make those analyses possible. Solving complex, multidisciplinary problems like modeling urban growth or tracking environmental change almost always requires both working together.
What Sensors Actually Measure
All remote sensing works by detecting electromagnetic energy, whether that’s visible light, infrared radiation, or microwave signals. Sensors fall into two categories based on where that energy comes from.
Passive sensors detect natural energy. They pick up sunlight reflected off Earth’s surface or heat radiated by the ground, water, and atmosphere. The power they measure depends on the surface’s composition, temperature, and roughness. Because they rely on existing energy, passive sensors can be sensitive to interference from other emitters on the ground, but spaceborne versions can still provide all-weather, day-and-night global observations.
Active sensors generate their own signal and measure what bounces back. Radar is the most common example. A precipitation radar sends out a pulse and reads the echo from rainfall to calculate rainfall rate. A cloud-profiling radar does the same to build three-dimensional maps of cloud structure. Because active sensors supply their own energy, they work regardless of sunlight or cloud cover, making them especially useful for monitoring weather and mapping terrain through dense vegetation.
Wavelength Bands and What They Reveal
Sensors don’t just take a single photograph. They capture data across multiple bands of the electromagnetic spectrum, and each band reveals different information about the surface below.
- Visible light bands produce imagery that looks similar to a normal photograph, useful for mapping land cover, water bodies, and built-up areas.
- Near-infrared bands are particularly valuable for vegetation studies because healthy plants reflect strongly in this range. Scientists use near-infrared data to measure vegetation health, detect plant stress, calculate canopy density, and monitor agricultural production trends.
- Thermal-infrared bands detect heat radiated from the surface. These bands let scientists measure land surface temperature, monitor urban heat islands, assess crop health through temperature variations, estimate soil moisture, and even support mineral and petroleum exploration by measuring how different materials retain heat.
- Specialized bands target narrow phenomena. NASA’s Landsat satellites, for instance, carry a cirrus band specifically designed to detect high, thin clouds that are difficult to spot in other parts of the spectrum.
The more bands a sensor captures, the more precisely analysts can distinguish materials on the ground. Hyperspectral sensors record hundreds of narrow bands. NASA’s AVIRIS instrument, for example, captures 224 spectral channels spanning wavelengths from 400 nanometers (visible light) to 2,500 nanometers (infrared), letting researchers identify materials with far greater specificity than a standard camera.
Four Types of Resolution
Resolution determines how much detail a remote sensing dataset contains, and it comes in four distinct flavors. Understanding each one helps you evaluate whether a particular dataset is appropriate for your project.
Spatial resolution is the size of each pixel on the ground. NASA’s MODIS instrument has bands at 250-meter, 500-meter, and 1-kilometer spatial resolution, meaning each pixel represents up to a 1 km × 1 km patch of Earth. That’s useful for continental-scale monitoring but far too coarse for inspecting individual farm fields. The Landsat OLI sensor captures data at 30 meters per pixel. Commercial satellites can reach roughly 30 centimeters per pixel, while drones routinely deliver centimeter-level detail.
Spectral resolution describes how finely a sensor slices the electromagnetic spectrum. More and narrower bands mean greater ability to distinguish between materials that look similar in broad wavelength ranges. A standard multispectral sensor might have 4 to 12 bands; a hyperspectral sensor like AVIRIS has 224.
Temporal resolution is the revisit time, how often a sensor images the same spot. MODIS revisits every 1 to 2 days because it sweeps a wide path. Landsat 8 has a narrower view and revisits every 16 days. Drones can fly on demand, making them the best choice when you need imagery on a specific date.
Radiometric resolution is the sensor’s sensitivity to differences in energy. It’s measured in bits per pixel. An 8-bit sensor records 256 possible brightness values per pixel. A 12-bit sensor records 4,096 values, capturing far more subtle differences in reflectance or temperature. Higher radiometric resolution helps distinguish features that would otherwise blend together.
Satellites vs. Drones
Satellites cover vast areas and return on predictable schedules, making them ideal for regional or global monitoring over time. Their spatial resolution ranges from about 30 centimeters per pixel on commercial platforms to 1 kilometer on free, publicly available instruments like MODIS. The tradeoff is that you’re locked into the satellite’s orbit and revisit cycle, and clouds can block optical sensors on any given pass.
Drones fill a different niche. They capture centimeter-level detail and fly whenever conditions allow, so you control both the timing and the area covered. The limitation is coverage: a drone survey might take hours to map a few hundred hectares, while a single Landsat scene covers 185 km × 180 km in one pass. For most GIS projects, the choice depends on whether you need high-frequency local detail or broad, repeatable coverage.
From Raw Image to GIS-Ready Data
Satellite or drone imagery doesn’t arrive ready for analysis. Several processing steps transform it into data that aligns accurately with other GIS layers.
Orthorectification corrects geometric distortions caused by terrain, sensor angle, and Earth’s curvature. Without it, features in the image won’t line up with GPS coordinates or existing map layers. GIS platforms like ArcGIS Pro include guided workflows for this, using control points and elevation models to produce geometrically accurate mosaics from drone, aerial, or satellite imagery.
Atmospheric correction removes the influence of haze, aerosols, and water vapor so that pixel values reflect actual surface conditions rather than atmospheric interference. This step is critical when comparing images captured on different dates or by different sensors.
Image classification is where remote sensing data turns into thematic maps. Analysts train algorithms to sort pixels into categories like forest, water, cropland, or urban area based on their spectral signatures. The classified output becomes a standard GIS layer that can be queried, measured, and combined with other datasets.
The finished data is typically stored in formats designed for geospatial work. GeoTIFF is the most common interchange format for georeferenced raster imagery and is an approved standard for NASA’s Earth science data. Cloud Optimized GeoTIFF (COG) reorganizes the same format for efficient streaming and processing in cloud environments. For multidimensional data, such as a time series of sea surface temperatures, NetCDF is widely used because it’s self-describing, portable, and allows efficient access to small subsets of very large datasets.
Precision Agriculture
Farming is one of the clearest examples of remote sensing and GIS working together. High-resolution satellite and drone imagery lets growers monitor fields at a level of detail that would be impossible on foot.
Vegetation indices calculated from near-infrared and red-edge bands reveal crop health, nutrient status, and water stress before problems become visible to the eye. In one study, researchers used hyperspectral drone imagery and machine learning to detect citrus canker with 96% accuracy, even during early stages of disease. Other work has shown that red-edge spectral indices outperform simpler vegetation measures when estimating nutrient levels and biomass in dense crop canopies, like late-season corn.
Yield prediction also benefits. Spectral indices measured during the growing season correlate with final grain yields, allowing farmers and agronomists to forecast production and adjust management. Satellite platforms like Sentinel-2 and IKONOS have both been used for this purpose, with red-based indices showing the highest correlation with corn yields across a range of fertilizer treatments.
Urban Heat Mapping
Thermal remote sensing plays a central role in understanding how cities trap and radiate heat. Satellites with thermal-infrared bands measure land surface temperature across an entire metro area in a single pass, revealing which neighborhoods, materials, and land cover types run hottest.
In a GIS, that thermal layer becomes the foundation for more targeted analysis. Analysts combine it with optical satellite imagery showing land use and land cover changes over time, then layer in socioeconomic data: population density, age demographics, income levels, health records. The result is a heat vulnerability index that identifies not just where temperatures are highest, but where people are most at risk. City planners use these indices to prioritize tree planting, cool-roof programs, and emergency response resources in the neighborhoods that need them most.
AI and Automated Analysis
The volume of remote sensing data flowing into GIS platforms has grown far beyond what human analysts can process manually. Deep learning models now handle much of the heavy lifting, particularly for tasks like land cover classification, object detection, and feature extraction from high-resolution and hyperspectral imagery. Attention mechanisms, a technique that lets a model focus on the most informative regions of an image, have improved classification precision by filtering out noise and irrelevant pixels. These automated workflows are increasingly built directly into GIS software, making it possible to process thousands of satellite scenes and extract meaningful spatial data without manually classifying each one.