Separate a satellite observation from a land-use claim
Can a changed pixel tell you why a neighborhood changed?
Your task: Create a legend for the supplied USGS image description, then draw the16-cell practice grid and label it fictional. Propose two GIS overlays and a field check before making a housing-change claim.
Read counts with their units. A percentage divides a part by its named whole and multiplies by 100. A model assumption is not an observed fact.
Remote sensing
Observation of an object or area from a distance using measured energy, often through airborne or satellite sensors.
Assumption
A condition accepted for a model or argument; it needs justification before applying the result elsewhere.
01 · TRY
Can a changed pixel tell you why a neighborhood changed?
Create a legend for the supplied USGS image description, then draw the16-cell practice grid and label it fictional. Propose two GIS overlays and a field check before making a housing-change claim.
Create a legend for the supplied USGS image description, then draw the16-cell practice grid and label it fictional. Propose two GIS overlays and a field check before making a housing-change claim. Use the supplied records and show one intermediate step.
Make a provisional claim
Loading saved evidence…
Inspect evidence and test an explanation
The investigation controls appear when your local evidence is ready.
04 · UNDERSTAND
Connect evidence and mechanism
Separate a satellite observation from a land-use claim
Question and evidence
Can a changed pixel tell you why a neighborhood changed?
USGS identifies Landsat 1 as launched July 23,1972 and describes a Dallas–Fort Worth false-color image from July 25,1972. In that image, red shades represent vegetation and gray/white surfaces may be urban or rocky. A separate fictional classroom grid has 16 cells, with 4 classified built-up initially and 6 later.
The institutional observation has its own date and context. Fictional numerical extensions are labeled separately and are not measurements from that source.
Mechanism
Remote sensing records reflected or emitted energy; false-color images map sensor bands into visible colors. GIS joins spatial layers such as roads, parcels and population. GPS supplies positioning observations, not automatically a social explanation. Classification needs validation and comparable acquisition conditions before interpreting change.
Worked calculation
The fictional classified share changes from 4/16=25% to 6/16=37.5%, a 12.5-point increase. That is not a measured Dallas change. To explain a real image, first inspect its date, band display, resolution and ground evidence.
Challenge and revision
The real source is an image description, not the classroom grid. Cloud, season, sensor resolution and classification error can affect comparisons. A built-up class does not establish who owns land, how people use a building or why development occurred.
Create a legend for the supplied USGS image description, then draw the 16-cell practice grid and label it fictional. Propose two GIS overlays and a field check before making a housing-change claim.
Revise your answer to “Can a changed pixel tell you why a neighborhood changed?” using both source IDs. Separate the checked calculation from a judgment, and identify evidence that could reverse your judgment.
Vocabulary
Remote sensing: Observation of an object or area from a distance using measured energy, often through airborne or satellite sensors.
Assumption: A condition accepted for a model or argument; it needs justification before applying the result elsewhere.
Independent extension
Was the change located correctly, or dated correctly?
A USGS study of LCMAP Collection1.0 reported82.5% overall per-pixel land-cover accuracy across years. That does not answer whether a change was detected in the correct year. Its published change-validation summaries below use different agreement definitions. These are historical study results, not current product accuracy.
The study used an independently collected reference sample interpreted from multiple imagery sources. Published percentages are rounded summaries; this activity does not reconstruct study counts, intervals or sampling weights from them. The selected definition corresponds to reported user/producer values13%/16%, not the toy results below. Widening agreement changes the validation question, not the map itself. A footprint over33 years does not verify the date of any particular change.
USGS publication record: Stehman and colleagues,2021, DOI10.1016/j.rse.2021.112646. The original study reports uncertainty for several estimates; no significance comparison is made here. Reference interpretation can also be imperfect.
Separate original timeline experiment
Seven fictional equal-area cells have at most one event in years1–25. Both map and reference records are supplied in full. A dash means no event during the interval, not an unknown value. The toy reference is stipulated for this exercise; it is not actual satellite validation evidence.
Circle: mapped event. Diamond: reference event, offset vertically to show coincident dates. Fixed year0–25 axis; rows are cell identities, not distances. Matching never crosses between cells.
Complete original records and current agreement
Cell
Mapped year
Reference year
Matched change under selected rule?
A
10
10
Yes
B
11
13
No
C
12
No event
No
D
No event
14
No
E
15
14
No
F
No event
No event
No
G
1
21
No
Matched changes 1; mapped events5; reference events5. Toy user's accuracy = matched/mapped = 1/5 = 20.0%. Toy producer's accuracy = matched/reference = 1/5 = 20.0%. Their denominators happen to be equal here; they answer different questions.
Unmatched mapped events:4; unmatched reference events:4. Cell F contributes neither a mapped nor reference event, so its no-change agreement does not inflate these change-specific ratios. A date mismatch can leave one event unmatched on each side.
Transfer between the three rules before and after changing B. Explain why a correct footprint can coexist with a wrong date. State which definition fits a claim about one year's development, and identify the reference evidence you would need. None of these labels establishes ownership, displacement or the cause of land use.
Compare your reasoning
Originally exact matching finds A only:1/5. A two-year window also matches B and E:3/5. The footprint also accepts G despite a20-year discrepancy:4/5. Moving B's reference to18 reduces window matching to2/5 but leaves exact and footprint unchanged. A larger score under a looser definition is not a repaired map. The seven-cell example does not estimate any published USGS accuracy or a real area's change.
Copy your temporal validation analysis
Temporary worksheet: reset restores exact-year/original dates and clears writing; guide/report exposure remains during this visit. Reload clears all temporary work. No mastery is awarded.
Independently extend the investigation “Separate a satellite observation from a land-use claim” with a documented real dataset or scholarly source. Preserve definitions, recreate one calculation or evidence table, compare a rival explanation and report why the teaching example cannot establish the real-world conclusion.
The fictional classified share changes from4/16=25% to6/16=37.5%, a12.5-point increase. That is not a measured Dallas change. To explain a real image, first inspect its date, band display, resolution and ground evidence.
