PSLE-SCI-REALITY-0451
Wait, What? A satellite can see a strip 185 km wide without each pixel being 185 km wide.
A student opens a satellite information card and sees two numbers printed only a few lines apart: swath width: 185 km and multispectral spatial resolution: 30 m. The student circles 185 km and says, “That must be the size of one pixel. The satellite sees one huge block of land at a time.”
The numbers are both real, but they describe different parts of the observation. Swath width tells us how wide a strip of Earth the instrument can observe as the spacecraft moves along its path. Spatial resolution tells us about the ground size represented by the image’s sampling or pixels for a stated band or product. Mixing the two turns a coverage number into a detail number.
This is a useful PSLE Science evidence problem because the current 2026 PSLE Science assessment frame asks learners not only to know science but also to interpret and analyse information, evaluate observations, information and methods, and communicate explanations and reasoning. A real scientific specification often contains several correct numbers. The difficult job is deciding what each number is evidence for.
Quick Answer
- Swath width is the width of the strip of ground observed during a pass.
- Pixel size or spatial resolution describes a much smaller ground sampling or detail scale for a stated band or product.
- A 185 km swath does not mean one pixel is 185 km across.
- A wide swath does not automatically mean poor detail, and a narrow swath does not automatically mean excellent detail.
- To compare two sensors, keep coverage, spatial detail, spectral bands, revisit timing, calibration and the scientific question separate until you know which feature matters for the job.
The Exact Learner Job This Reality Lab Owns
This article owns one real-world evidence-transfer job: how to read a satellite or airborne-sensor specification that gives a swath width without mistaking that coverage width for the size of one image pixel or the instrument’s spatial resolution.
It does not re-teach the general PSLE Science skills of reading units, interpreting diagrams, comparing quantities, evaluating models, checking scale or drawing conclusions from evidence. Those skills already have owners in the eduKateSengkang Science estate. Here, they are applied to a specific scientific communication object: an Earth-observation specification sheet.
For the broader evidence skill, use the Primary 6 Science Learning Guide: Data, Graphs, Diagrams & Evidence. For the separate question of what one satellite pixel represents, route to Reality Lab Vol No.062. This page does not replace either owner.
Rebuild the Evidence Object: The CoastWatch-8 Card
Imagine an original school research card for a fictional Earth-observation mission called CoastWatch-8:
| Specification | Value |
| Orbit altitude | about 700 km |
| Swath width | 185 km |
| Blue, green, red and near-infrared bands | 30 m ground sampling |
| Panchromatic band | 15 m ground sampling |
| Repeat cycle | 16 days |
A student is asked which number tells us how much territory can fit across the strip observed during a pass. The answer is 185 km. Another student is asked which number is relevant when deciding whether two small features tens of metres apart may be represented separately in a particular band. The relevant starting number is the spatial sampling or resolution for that band, such as 30 m or 15 m, not 185 km.
The same card contains all the numbers, but the questions point to different evidence. Scientific reading is not “find the biggest number.” It is “match the quantity to the claim.”
Observed, Claimed and Inferred
Observed from the specification: the instrument is described as collecting a strip about 185 km wide, while particular bands are described with ground sampling on the order of tens of metres.
Reasonable claim: during a pass, the instrument can cover a very wide cross-track strip while still dividing the recorded image into many much smaller spatial samples.
Unsupported inference: each individual pixel covers 185 km, or the 185 km number alone tells us the smallest object the sensor can distinguish.
That separation matters because the language of a datasheet can feel compressed. “Swath”, “scene”, “pixel”, “ground sample distance”, “spatial resolution”, “band” and “repeat cycle” can all appear together. The learner’s job is to slow down and attach each number to the correct noun.
A Strip Is Made of Many Samples
Imagine a long roll of graph paper moving under a scanner. The scanner can cover the full width of the paper while recording many small cells across that width. The width of the whole roll under the scanner and the width of one cell are not competing answers. They are measurements at different levels of the system.
Earth-observation instruments work in a far more sophisticated way, but the separation is similar. The spacecraft moves along its orbital path. The instrument views a cross-track width on Earth. Within that coverage, detector geometry and processing create many spatial samples. A wide footprint of observation can therefore contain a very large number of much smaller image elements.
If a simplified strip were 180 km wide and samples were 30 m across, then 180 km is 180,000 m. Dividing 180,000 m by 30 m gives about 6,000 sample widths across the strip. That does not claim a real satellite image always has exactly 6,000 usable pixels across every product; geometry, band design, resampling, edges and processing matter. The calculation merely shows why “wide swath” and “small ground sample” can coexist.
Do Not Confuse Scene Size With Pixel Size Either
Satellite documentation may also provide a scene size. A scene is a packaged image region or data product footprint. The scene can be roughly swath-wide and many kilometres long. It contains many pixels. A scene is not one pixel, just as a full class photograph is not one camera sensor element.
A careful learner can therefore build a simple hierarchy:
- Pass or orbit: the spacecraft travels along a path.
- Swath: the cross-track strip the sensor can observe.
- Scene or product tile: a packaged section of the collected strip.
- Pixel or spatial sample: a much smaller element inside the scene.
The hierarchy protects the evidence. If a claim is about total coverage, the swath is relevant. If a claim is about local detail, the spatial sampling and actual resolving performance matter. If a claim is about how often a place is seen, revisit or repeat timing matters. One number cannot silently do all three jobs.
Representation Check: Draw the Specification Before You Interpret It
When specifications become abstract, convert them into a simple diagram. Draw a long vertical arrow for the satellite’s direction of travel. Across it, draw a broad horizontal rectangle labelled “swath”. Then divide a tiny part of the rectangle into a grid of small cells labelled “spatial samples”.
Now ask: Which label belongs to the whole width? Which belongs to one cell? This representation check often repairs the mistake before any calculation is needed.
The diagram is not a scale drawing unless you deliberately make it one. It is a model for relationships. A good scientific model can simplify shape while preserving the distinction that matters.
Comparison Check: Wider Is Not Automatically Better
Suppose two fictional sensors are being considered for a student project:
| Sensor | Swath width | Spatial sampling | Project |
| RiverView | 300 km | 100 m | map broad regional flooding |
| StreetEye | 20 km | 2 m | inspect small urban features |
If the job is to cover a huge region quickly, RiverView’s wide swath may be valuable. If the job is to distinguish small structures, StreetEye’s much finer spatial sampling may be more useful. It would be poor reasoning to declare RiverView universally better because 300 km is the larger number, or StreetEye universally better because 2 m is the smaller number.
Science comparisons need a purpose. “Better” without a task is often an unfinished sentence.
Baseline Check: What Does “More Coverage” Mean?
A marketing graphic might say, “Our sensor covers twice the width.” That is not yet the same as “collects twice the useful area every day.” Why? Coverage can also depend on orbital path, overlap between passes, clouds for optical sensors, acquisition scheduling, sensor availability, usable edge quality and the definition of “covered”.
If two otherwise identical swaths are 100 km and 200 km wide, the wider one can cover more cross-track ground in one pass. But a real comparison should still check whether the same conditions, resolution, band set and usable-data rules apply. A single specification line is evidence, not the whole system.
Method and Variable Check
If someone claims that Sensor A is “better for mapping forests” because its swath is wider, identify the variable that was actually compared: swath width. Then ask what forest-mapping performance depends on. The answer may include band wavelengths, signal quality, spatial scale, cloud conditions, revisit timing, calibration and the kind of forest question being asked.
Swath width can support a statement about coverage. It cannot, by itself, prove better species identification, more accurate biomass estimates, finer tree-level detail or stronger trend detection. Those are different outcome variables that need their own evidence.
Alternative Explanations: Why a Wide Swath May Not Produce More Useful Data
Imagine a week of cloudy weather. A wide optical swath passes over a region repeatedly but much of the land is obscured. A narrower sensor on a different day captures a clear view of the target. “Wider swath” was true, yet “more useful evidence for this week’s project” may be false.
Or imagine a large region where only one narrow coastline is relevant. A wide swath may collect much more surrounding land and ocean than the question requires. The coverage advantage exists, but it may not be the limiting factor for the scientific job.
Alternative explanations do not make swath width meaningless. They stop one correct fact from expanding into an unsupported universal claim.
Worked Case 1: The Island Map
A fictional island is 60 km from east to west. Sensor A has a 185 km swath and 30 m spatial sampling. A student says, “The island fits inside one pixel because 185 km is wider than 60 km.”
Repair: the 185 km number describes the whole observation strip, not one pixel. A 60 km island could span many thousands of 30 m samples across its width. The correct first conclusion is that the island may fit comfortably within the swath, not that it collapses into one pixel.
Worked Case 2: The Two Cameras
Sensor B has a 50 km swath with 5 m spatial sampling. Sensor C has a 200 km swath with 50 m spatial sampling. The task is to map a very large agricultural province and then identify narrow irrigation channels.
Sensor C may cover the province with fewer strips, but Sensor B may represent narrower features more clearly. The best answer may involve combining evidence, using different sensors for different tasks, or accepting a trade-off. “Largest swath wins” fails because the project has more than one requirement.
Worked Case 3: The Cropped Image
A website displays only a 10 km by 10 km crop from a satellite scene. A learner says, “The swath must be 10 km because that is all the webpage shows.”
The displayed crop is not evidence for the full instrument swath unless the documentation says it is. A data portal can cut a small window out of a much larger collected scene. Always separate the display window from the acquisition geometry.
Worked Case 4: Overlap Is Not Wasted Evidence
Two neighbouring satellite passes overlap. A student sees the overlap and says, “Those kilometres were wasted because they were photographed twice.”
That conclusion is too quick. Overlap may support continuity, repeated viewing, mosaicking or coverage at high latitudes. It can also mean that simple multiplication of swath width by number of passes overstates unique new ground covered. The evidence job is to distinguish gross coverage from unique coverage.
What Evidence Would Strengthen a Coverage Claim?
- An authoritative mission specification defining swath width.
- A map showing the actual acquisition footprint.
- Orbit or flight-path information explaining where the strip falls.
- Clear distinction between gross swath width and usable scene width.
- Information about overlap, gaps and acquisition scheduling.
- Cloud or quality information when the scientific question requires visible surface observations.
- A consistent comparison in which competing sensors are evaluated for the same region and purpose.
What Evidence Would Weaken an Overclaim?
- The page gives only a giant coverage number but no spatial detail information.
- The displayed image is a crop, yet the crop width is called the sensor swath.
- The comparison changes both swath and resolution but credits all improvement to swath.
- The claim ignores clouds, missing acquisitions or unusable edge data.
- A scene dimension is silently renamed “pixel size”.
- A wide swath is used as proof of accuracy, sensitivity or scientific validity.
- The sensor is judged “best” without defining the scientific task.
How Far Can the Conclusion Travel?
From a verified swath-width specification, you can make a bounded statement about the width of ground coverage under the stated acquisition geometry. You cannot automatically infer the size of a pixel, the smallest resolvable object, the accuracy of a measurement, the number of useful cloud-free observations, the revisit time, the spectral sensitivity or the scientific quality of the final conclusion.
The scientific habit is not to distrust specifications. It is to use them for the job they were designed to describe.
Tempting but Invalid Reasoning
- “185 km is the biggest number, so it must be the resolution.” Quantity names matter more than numerical size.
- “A 185 km swath means a 185 km pixel.” The swath contains many spatial samples.
- “A wider swath always makes a better sensor.” Better depends on the scientific job.
- “A narrow swath means poor technology.” A narrow field may accompany very fine detail or a specialised measurement.
- “If an island fits within the swath, it fits in one pixel.” Whole-strip coverage and pixel scale are different.
- “The scene shown on screen equals the entire swath.” Portals can crop, tile, mosaic or resample data.
- “Two sensors have the same swath, so they collect the same evidence.” Bands, resolution, calibration, timing and conditions can differ.
Model and Measurement Limits
Even spatial resolution is not a magical promise that every object larger than the stated number will be perfectly visible and every smaller object will vanish. Real detectability depends on contrast, shape, signal, sensor response, processing and the way the feature falls across samples. The point here is narrower: swath width is not the same quantity as spatial resolution.
Likewise, the edges of a swath can differ geometrically from the centre for some sensors, and mission products may use resampling or map projections. Students do not need specialist remote-sensing mathematics to reason correctly. They need to keep the quantities attached to their definitions and avoid extending a number beyond its evidence boundary.
PSLE-Style Transfer Case
A fictional satellite has a swath width of 120 km and a spatial resolution of 20 m in one band. A news-style infographic says, “Each satellite image measures 120 km, so small farms cannot be distinguished.” Evaluate the claim.
Reasoned answer: the 120 km figure describes the width of the ground strip observed, not the size of each image element. The 20 m spatial-resolution figure is more relevant to the scale of detail represented in that band. Whether a particular farm can be distinguished also depends on its size, contrast, band choice, image quality and processing. Therefore the infographic uses the swath number for the wrong evidential job.
Explained Practice
Practice A. A sensor has a 250 km swath and 500 m spatial sampling. Which number is evidence about cross-track coverage? 250 km.
Practice B. A different sensor has a 20 km swath and 1 m spatial sampling. Can you call it worse because 20 is smaller than 250? No. The units and scientific job differ. It covers a narrower strip but can represent much finer spatial detail.
Practice C. A webpage shows a 5 km crop from a mission with a 185 km swath. Is the mission swath now 5 km? No. Cropping changes what is displayed, not the original instrument coverage specification.
Practice D. Two satellites both have 185 km swaths. One uses 30 m multispectral bands; another uses different wavelengths and resolutions. Are they scientifically identical? No. Equal swath width answers only one comparison question.
Practice E. A student says, “This 185 km swath means the sensor can see an object 185 km long.” What is missing? The claim confuses total coverage with the ability to distinguish a feature. Feature detection depends on spatial scale and other measurement properties.
Delayed Independent Return
Several days later, without rereading this article, take any unfamiliar scientific sensor specification and answer:
- Which number describes total coverage?
- Which number describes local sampling or detail?
- Which number describes time between observations?
- Which number describes what wavelengths or quantities are sensed?
- Which claim is each number allowed to support?
- Which conclusion would require another piece of evidence?
If you can keep those quantities separate even when they appear on one attractive infographic, the evidence-transfer habit has become portable.
Parent and Tutor Teaching Guide
Start with two pieces of paper. Use one whole A4 sheet to represent the swath. Draw a grid of tiny squares on it to represent spatial samples. Ask the learner: “Does the width of the paper equal the width of one square?” The physical contrast makes the distinction obvious before any remote-sensing terminology appears.
Next, give three fictional sensor cards. Change only one feature at a time: first swath width, then spatial sampling, then repeat interval. Ask a different question after each change. This teaches the child to match evidence to a claim rather than ranking every specification on one imaginary “better” scale.
Finally, mix the features. Give a wide-coarse sensor, a narrow-fine sensor and a medium sensor with a useful wavelength band. Ask which is best for mapping a continent, locating narrow roads and detecting a vegetation property. The correct answer can change with the task. That is not indecision; it is scientific fitness for purpose.
Authoritative Reality Check: Landsat 8
NASA’s current Landsat 8 mission information provides a real example of why these quantities must be separated. It lists a swath width of about 185 km while also listing 30 m spatial resolution for visible, near-infrared and shortwave-infrared data, 15 m for the panchromatic band and coarser thermal sampling. USGS likewise describes 15 m panchromatic and 30 m multispectral spatial resolutions along a 185 km swath. Those figures coexist because they describe different levels of the observing system.
Sources: NASA Science — Landsat 8; U.S. Geological Survey — Landsat 8; NASA Science — Operational Land Imager; Singapore Examinations and Assessment Board — PSLE.
Quiet Return: Ask What the Number Measures
A scientific specification can be completely accurate and still be misused by a reader. The repair is rarely “memorise more numbers”. It is to ask a calmer question: What physical or observational quantity does this number actually describe?
For a swath, the answer is coverage width. For a pixel-scale or spatial-resolution statement, the answer is local spatial sampling or detail. Keep those jobs separate, and a dramatic 185 km number stops swallowing the rest of the evidence.
