Satellite microwave sensing uses wavelengths longer than visible and infrared light to investigate Earth’s surface and atmosphere. Passive instruments measure weak microwave energy naturally emitted by land, water, ice and air. Active instruments transmit pulses and measure the energy scattered back. Many microwave bands can operate at night and see through ordinary cloud, which makes them valuable for tracking storms, sea ice, soil moisture and ocean winds. They are not all-weather magic. Rain, vegetation, surface roughness and radio interference can alter the signal, while the long wavelengths often produce coarser images than optical cameras. Understanding whether a product comes from emission, backscatter, travel time or Doppler shift is the first step towards reading it correctly.
Key takeaways
- Passive sensors measure natural microwave emission; active sensors transmit a signal.
- Many microwave observations work through cloud and without sunlight.
- Frequency and polarisation determine what a sensor is sensitive to.
- Moisture, roughness, vegetation and viewing angle can produce similar changes.
- Longer wavelengths usually mean a trade-off in spatial resolution.
- Calibration, models and ground data turn measurements into useful products.
Why microwaves are useful from space
Microwaves occupy part of the electromagnetic spectrum between radio waves and infrared radiation. Their wavelengths range from millimetres to metres, although Earth-observing missions use selected frequency bands. The atmosphere is relatively transparent in some of these windows, allowing instruments to observe through clouds that block visible and thermal infrared sensors.
The wavelength also controls interaction with objects. A short microwave can respond to raindrops or a rough water surface, while a longer wavelength may penetrate farther into dry snow, vegetation or soil. Mission designers select bands for a scientific question rather than treating “microwave” as one uniform technique.
Passive microwave radiometers
A radiometer listens for natural microwave emission. Because the signal is weak, its antenna gathers radiation over a relatively broad footprint. Algorithms express the measured radiance as brightness temperature: the temperature a perfect emitter would need to match the observation. This is not automatically the physical temperature of the ground.
Water has different microwave emission from dry mineral soil, and liquid water strongly changes soil’s dielectric properties. Missions such as SMAP use this contrast to estimate near-surface soil moisture. Sea ice, snow water equivalent and atmospheric water vapour can also be inferred with appropriately chosen frequencies.
Active radar instruments
Active sensors supply their own energy. A radar transmits a pulse and records the echo’s strength, timing, phase or frequency. Travel time can provide distance, backscatter describes how strongly a target returns energy, and a Doppler shift can reveal motion towards or away from the sensor.
Synthetic aperture radar combines observations made as the spacecraft moves to create finer spatial detail than the physical antenna alone would provide. Scatterometers sacrifice some spatial detail to measure ocean-surface roughness across a wide swath. Radar altimeters point downwards and time echoes to estimate surface height.
Frequency, polarisation and incidence angle
Microwave missions are often described by bands such as L, C, X or Ku. These labels refer to frequency ranges, and the same landscape can look different in each. Polarisation describes the orientation of the electric field on transmission and reception. Comparing horizontal and vertical channels provides extra information about surface structure.
Incidence angle—the angle at which the beam meets the surface—also changes backscatter. A smooth surface may reflect energy away from the antenna at one angle, while waves or rough soil scatter it back. Comparisons should therefore use matched geometry or products that correct for it.
How soil moisture is estimated
Wet soil generally has a higher dielectric constant than dry soil and therefore changes microwave emission and backscatter. Retrieval algorithms combine the measured signal with assumptions about temperature, vegetation and roughness. The product commonly represents only the upper few centimetres and an area much larger than a garden or individual field.
Validation networks compare satellite estimates with probes at different depths. A farmer’s root-zone moisture may not equal the satellite surface value, but repeated regional patterns can still support drought monitoring and hydrological models. The distinction resembles the difference between surface and air temperature explained in satellite land-surface temperature.
Rain, snow and ice
Weather radar and spaceborne precipitation radar detect scattering by hydrometeors—water droplets, ice crystals and hail. Doppler methods add radial motion, as described in how Doppler weather radar works. Passive channels sense emission and absorption associated with atmospheric water and ice.
Sea ice concentration is mapped because ice and open water have contrasting microwave properties. Melt ponds, thin ice and coastal land can confuse retrievals. Snow depth is particularly difficult when grain size, liquid water and layered ice alter scattering in ways that resemble a change in snow mass.
Vegetation and forest structure
Leaves and stems absorb and scatter microwaves, sometimes hiding the soil signal. Longer wavelengths may penetrate more vegetation and respond to woody biomass, although dense forests eventually saturate the measurement. Radar interferometry and polarimetry can add information about canopy height and structure.
Optical indices such as NDVI vegetation measurements respond mainly to reflected sunlight. Combining optical and microwave evidence is powerful because clouds and moisture affect them differently, but their pixel sizes and observation times must be aligned.
Resolution, swath and revisit time
Antenna size, wavelength, orbit and processing determine spatial resolution. Passive microwave footprints can cover tens of kilometres, so coastlines and mixed land cover are challenging. Active radar can achieve metres to kilometres, depending on mode, while often covering a narrower swath.
These choices connect directly to satellite orbit and mission design. A product with daily global coverage may average away local detail; a fine-resolution radar image may revisit less often or require more data processing.
Sources of uncertainty
Radio-frequency interference from communications systems can contaminate passive observations. Rain can obscure ocean winds, frozen soil can resemble dryness, and topography changes local geometry. Retrieval algorithms may be calibrated for one landscape and perform differently elsewhere.
Quality flags are therefore part of the data, not optional decoration. Users should check units, footprint, overpass time, frozen-ground and rain flags, algorithm version and validation results. A smooth map may include interpolation between observations rather than direct measurements everywhere.
A practical interpretation workflow
Start by identifying sensor type, frequency, polarisation and measured quantity. Then separate the observation—brightness temperature or backscatter—from the retrieved variable such as soil moisture. Examine resolution and flags before comparing places or dates.
Use field measurements or an independent dataset to test the expected relationship. If several factors could explain a change, state them. Microwave sensing is strongest when treated as a physical measurement with a model, not as a camera that reveals an unambiguous property.
Worked comparison: flood mapping with two sensors
After a tropical storm, an optical image may be hidden by cloud. A synthetic aperture radar image can still distinguish many areas of smooth open water from rough land because calm water reflects much of the transmitted energy away from the antenna. Analysts compare the new image with a dry-season reference, mask permanent water and check roads, buildings and wind-roughened lakes that can produce confusing returns.
The resulting flood map is a classification, not a direct measurement of water depth. Forest canopy can hide water, urban walls create bright radar echoes and a different incidence angle changes appearance. Field reports and river gauges provide independent checks. The example shows why microwave access through cloud is valuable while interpretation remains dependent on geometry and local surface conditions.
Combining active and passive records
Passive instruments can provide frequent broad measurements of soil moisture or ice concentration, while active radar supplies finer spatial detail on selected passes. A hydrologist might use a coarse daily moisture series to identify regional drying and radar imagery to examine which fields or wetlands changed. The datasets cannot simply be overlaid without reconciling footprint, depth sensitivity and observation time.
Data fusion models can estimate variables at finer scale, but the output includes assumptions from every input and from the fusion method. Validation should occur at the resolution of the claimed product. A few ground probes beside roads do not establish accuracy across mountains, forests and farms.
Microwave data in long-term monitoring
Climate records require stable calibration across instruments and missions. Engineers monitor antenna performance, receiver noise and orbit, while teams compare overlapping satellites and well-characterised targets. An apparent trend can otherwise arise from a new sensor, frequency, overpass time or retrieval version.
Researchers distinguish a climate data record from a rapid near-real-time product. The first emphasises consistency and reprocessing; the second prioritises speed. Both are useful, but a near-real-time map should not automatically be joined to an older series for trend analysis.
How to evaluate a microwave-sensing claim
Begin with the measured quantity and band. “The satellite measured drought” is too vague: it may have measured brightness temperature and used a retrieval model for surface soil moisture. Check footprint, depth, overpass time, flags and whether the comparison used the same sensor and algorithm. Ask which ground sites validated the product and whether they represent the landscape.
Then examine alternative causes. Vegetation growth, frozen ground, surface roughness or interference can change a signal without the proposed environmental change. A strong claim should show how these factors were controlled, display uncertainty and explain where the method performs poorly. That sequence—from observation through model to conclusion—is more useful than judging a map by visual detail.
Conclusion
Satellite microwave sensing extends Earth observation beyond daylight and clear skies. Passive radiometers measure natural emission, while active radars measure echoes from transmitted energy. The resulting signals can reveal moisture, ice, rainfall, wind and surface structure, but only through algorithms that account for wavelength, polarisation, geometry and interfering conditions. Used with quality flags and ground validation, microwave products provide evidence that optical sensors cannot; read without those checks, they can turn several physical causes into one misleading colour scale.
Frequently asked questions
Can microwave satellites see through every cloud?
Many bands penetrate ordinary cloud, but heavy rain and atmospheric water can attenuate or alter the signal.
Is radar a microwave sensor?
Yes. Most Earth-observing radar instruments use microwave frequencies and actively transmit pulses.
Does passive microwave measure temperature?
It measures radiance often expressed as brightness temperature, which is influenced by physical temperature and emissivity.
How deep can microwaves see into soil?
Usually only the near surface for soil-moisture retrieval; depth varies with wavelength, moisture, texture and vegetation.
Why are passive microwave maps coarse?
Antenna physics and weak signals require a broad footprint unless very large antennas or specialised processing are used.
