PSLE Science Reality Lab Vol No.285 | “Tree-Cover Loss = 500 ha” — Does That Mean 500 ha Was Permanently Deforested?

Series ID: PSLE-SCI-REALITY-0285

Wait, What? A Satellite Can See Trees Disappear Without Knowing the Final Land Use Yet

A fictional forest dashboard reports: “500 hectares of tree-cover loss.” A student rewrites it as: “500 hectares were permanently deforested.” The two statements sound similar because both involve missing trees. Scientifically, they are not interchangeable.

Tree-cover loss is a remote-sensing observation about removal or mortality of tree canopy within a defined mapping system. Deforestation is a narrower land-use conclusion: natural forest is converted to a different, non-forest use in a way that is expected to be permanent. Tree cover can disappear because of permanent agricultural clearing, but also because of timber harvest, wildfire, storms, insects or other disturbances. Some of those losses can later regrow.

The learner job is therefore not to weaken a serious environmental result. It is to name the result accurately before deciding what it means.

Quick Answer

Tree-cover loss means mapped tree canopy was removed or died according to a stated satellite product and baseline.
It can occur in natural forest, planted forest, plantations, tree crops or other areas with tree cover, depending on the dataset.
It can be caused by clearing, logging, fire, storms, disease, flooding, insects and other disturbances.
Some tree-cover loss is permanent; some is temporary.
Deforestation usually means human-caused permanent conversion of natural forest to another land use.
A tree-cover-loss map alone does not automatically reveal the cause or whether trees will return.
To infer deforestation, use extra evidence about the original land cover, driver of loss and what happened afterward.

The Exact Learner Job This Volume Owns

This Reality Lab owns one bounded evidence-transfer job: how to evaluate a real-world satellite tree-cover-loss map or headline without automatically converting a detected canopy-loss event into permanent deforestation.

It does not re-own forest ecology, photosynthesis, remote sensing, land-use policy, climate science or sampling. Those remain with existing owners. Here the communication object is the map itself: what does the coloured loss pixel actually support?

Why This Is a PSLE Science Evidence Problem

The 2026 PSLE Science syllabus assesses learners’ ability to interpret and analyse information, evaluate observations, information and methods, and communicate explanations and reasoning. The 2023 Primary Science syllabus also promotes healthy scepticism and openness to more than one plausible explanation.

A tree-cover-loss map creates exactly that challenge. The map contains evidence. The learner must separate the observation from the cause and from the long-term land-use conclusion.

Rebuild the Evidence Object

satellite images → define baseline tree cover → detect major reduction or loss of canopy in a pixel → map the loss event → investigate cause → check later land cover → decide whether the event represents permanent deforestation, temporary disturbance or another kind of tree-cover loss

The most important arrows are the last two. A satellite change-detection product can be very good at showing that tree cover disappeared. Determining why it disappeared and whether the land remains non-forest requires additional evidence.

Observed, Classified and Inferred

Layer	Example	What it supports
Observation	Tree canopy present in baseline imagery	Mapped tree-cover condition at the starting point
Detected change	Canopy is lost within a mapped pixel	Tree-cover-loss event under the product definition
Driver classification	Fire, logging, agriculture, storm or another driver	A proposed or mapped cause
Long-term land-use check	Area becomes cropland, road, mine or regrowing forest	Evidence about permanence and final use
Unsupported leap	Every tree-cover-loss pixel equals permanent deforestation	Not justified without additional evidence

The Baseline Matters

Global Forest Watch and World Resources Institute explain that tree-cover-loss statistics are calculated against a defined baseline. A commonly used global layer begins with tree cover in the year 2000 and applies a canopy-density threshold. Change is then detected relative to that mapped starting condition.

This means a number such as “500 ha lost” is not simply “500 ha where no trees exist today.” It is a result produced by a specific baseline, canopy definition, time period and change-detection method.

Tree Cover Is Not Exactly the Same as Forest

A satellite can detect tall woody vegetation without knowing whether the trees are an old natural forest, a timber plantation, an orchard, an agroforestry system or another tree-covered land use. That distinction matters because a tree plantation being harvested is not the same ecological event as primary rainforest being converted permanently to a road or field.

World Resources Institute explicitly warns that tree-cover-loss datasets can include natural forests, planted forests, tree crops and other areas with tree cover. A responsible headline should therefore state what baseline or forest class it is actually discussing.

Permanent Versus Temporary Loss

Some tree-cover loss is clearly permanent. If natural forest is cleared and a settlement, mine or permanent cropland remains, regrowth to the former forest is not part of the near-term land use. Other disturbances may be temporary: a plantation is harvested and replanted, a forest burns and later regenerates, or a storm knocks down trees but the land remains forest land.

“Temporary” does not mean harmless. Recovery may take years or decades, species composition may change, carbon can be released and habitats can be damaged. The scientific point is narrower: temporary tree-cover loss is still not identical to permanent deforestation.

Worked Case 1: A Timber Plantation Is Harvested

A satellite detects 200 ha of tree-cover loss in a plantation scheduled for harvest. Two years later, young trees are visible in the same blocks.

Evaluation: The original tree-cover-loss detection can still be correct. However, calling the event permanent deforestation would be misleading if the land remained a tree plantation and was replanted. The observation and the land-use conclusion answer different questions.

Worked Case 2: Natural Forest Becomes Cropland

Satellite imagery shows natural forest in Year 1, clearing in Year 2, and persistent cropland through Years 3 to 6.

Evaluation: This evidence chain supports a much stronger deforestation inference because the original forest, clearing event and persistent conversion to a non-forest land use are all documented.

Worked Case 3: Wildfire Removes Canopy

A severe wildfire causes large areas of canopy mortality. A tree-cover-loss layer maps the event. A headline immediately calls the full area “deforested.”

Evaluation: Fire-driven tree-cover loss is real, but deforestation depends on what happens afterward. If forest regenerates, the event is not the same as permanent conversion. If the burned land is then converted to another use, the long-term interpretation changes.

Worked Case 4: A Storm Damages a Forest

A cyclone snaps and uproots many trees. The satellite map records canopy loss. There is no evidence of clearing, farming or construction afterward.

Evaluation: The strongest claim is storm-related tree-cover loss, not human-caused deforestation. Cause matters.

Worked Case 5: A Map Pixel Contains More Than One Story

A coloured grid cell covers a mixed landscape: natural forest, a small plantation and a road edge. The product reports a dominant loss driver for the cell.

Evaluation: A dominant-driver classification is useful, but it does not mean every square metre experienced the same process. Current WRI driver datasets explicitly caution that a 1 km cell can contain more than one driver at smaller scales. The map is a modelled summary, not a photograph of one uniform event.

Worked Case 6: Gross Loss Without Counting Gain

A country dashboard reports 10,000 ha of tree-cover loss over a period. Elsewhere, 4,000 ha of new tree cover appears. A reader subtracts and says the loss dataset must really be 6,000 ha.

Evaluation: Many tree-cover-loss statistics are gross loss measures. They report mapped loss events separately from gain. The 10,000 ha loss figure does not become mathematically wrong because gain happened elsewhere. Gross loss and net change are different quantities.

Representation Check: A Red Pixel Is Not a Land-Use Biography

Maps compress complicated evidence into symbols and colours. A red loss pixel may tell us that canopy was removed or died within that mapped unit during a stated year. It usually does not tell the entire land-use history by itself.

Read the legend carefully. Is it showing annual tree-cover loss, primary-forest loss, fire-driven loss, a near-real-time disturbance alert, or a classified driver? Similar-looking maps can answer different questions.

Comparison and Baseline Check

Was the same canopy-density threshold used?
Was the same baseline year used?
Are both maps describing all tree cover or only natural/primary forest?
Are both statistics gross loss, or is one a net-change estimate?
Are the spatial resolutions comparable?
Did the product methodology change between years?

Without these checks, a dramatic difference between two maps can partly reflect a different measurement frame rather than a change in the forest alone.

What Evidence Strengthens a Deforestation Claim?

The baseline is confirmed as natural forest rather than any tree-covered land.
Satellite change detection clearly shows canopy removal.
Driver information indicates permanent agriculture, mining, infrastructure or settlement expansion.
Later images show persistent non-forest land use.
Independent land-cover maps, field observations or official records agree.
The spatial scale of the claim matches the spatial scale of the evidence.
The report distinguishes uncertainty and classification limits.

What Weakens an Over-Broad Claim?

Every tree-cover-loss pixel is called deforestation.
Plantation harvest is mixed with natural-forest conversion without explanation.
Fire, storms or insect damage are ignored as possible drivers.
The starting land cover is unknown.
No later observation checks whether trees regrew.
Gross loss is described as net forest change.
A 1 km driver class is treated as the exact cause at every point inside the cell.

Alternative Explanations for a Loss Patch

A new patch of mapped tree-cover loss could result from permanent agricultural clearing, rotational harvest, wildfire, storm damage, a landslide, insects, disease or another disturbance. The job is not to list alternatives forever. It is to identify which alternatives are plausible and ask what later evidence would distinguish them.

For example, persistent crop rows support a different conclusion from young forest regrowth. A burn scar plus later regeneration supports a different conclusion from a new road network and buildings.

How Far Can the Conclusion Travel?

A careful statement might be: “The satellite tree-cover-loss product detected 500 ha of canopy loss within its stated baseline and method. Additional driver and land-use evidence is needed to determine how much of that loss represents permanent deforestation.”

That sentence does not minimise the loss. It simply refuses to claim a land-use history that the first layer of evidence does not yet contain.

Tempting but Invalid Reasoning

“Tree-cover loss equals deforestation.” Some loss is temporary or occurs in planted tree systems.
“If trees disappeared, humans must have cut them.” Fire, storms, insects and other disturbances can cause loss.
“If loss is temporary, it does not matter.” Temporary disturbances can still have major ecological consequences.
“A map pixel tells the exact cause everywhere inside it.” Classification scale and mixed landscapes matter.
“Gross loss already subtracts regrowth.” Gross loss and net change are different measures.
“One year’s loss map tells us the final land use.” Later observations are often needed to establish permanence.

PSLE-Style Transfer Case

Area	Year 1	Year 2	Year 5
P	Tree plantation	Tree-cover loss	Young plantation regrowth
Q	Natural forest	Tree-cover loss	Permanent cropland
R	Natural forest	Fire-driven canopy loss	Natural regeneration

Question 1: Which areas experienced tree-cover loss? P, Q and R.

Question 2: Which area most clearly supports permanent deforestation? Q, because natural forest was converted persistently to cropland.

Question 3: Why is P different? The land remained a plantation and tree cover returned after harvest.

Question 4: Why is R not automatically deforestation? The loss was caused by fire and natural tree cover later regenerated.

Explained Practice

1. What does a tree-cover-loss product directly tell us? That mapped tree canopy was removed or died under the product’s definition.

2. What extra evidence helps establish deforestation? The original forest type, likely driver, and later persistent conversion to a non-forest use.

3. Why can a harvested plantation appear as tree-cover loss? Satellite canopy detection responds to the removal of trees; it does not automatically classify the land-use history as natural-forest deforestation.

4. Why check later imagery? It helps distinguish regrowth or temporary disturbance from persistent conversion.

5. What is the core habit? Separate detected change from cause and permanence.

Delayed Independent Return

Tomorrow, draw three imaginary satellite panels of the same place: before loss, immediately after loss and five years later. Make one version end in cropland, one in regrowing forest and one in a replanted timber plantation. For each, write exactly which year first supports “tree-cover loss” and which later evidence is needed before you use the word “deforestation.”

Useful eduKateSengkang Routes

Parent and Tutor Teaching Guide: Make the Timeline Visible

Give the child three simple drawings of a patch of land. Picture 1 has trees. Picture 2 has no canopy. Picture 3 is initially hidden. Ask, “What can we say after Picture 2?” The safe answer is tree cover was lost. Then reveal three possible Picture 3 endings: regrowth, plantation replanting, or permanent buildings/crops.

The learner should discover that the same middle picture can lead to different long-term interpretations. This makes the distinction between observation and later land-use inference concrete without requiring advanced remote-sensing theory.

Authoritative Sources

Current WRI guidance is explicit that tree-cover loss is not always deforestation. It can include natural and planted tree cover, human and natural disturbances, and temporary as well as permanent losses. That distinction is the evidence boundary this volume asks pupils to preserve.

The Quiet Rule to Keep

A map can show change very clearly while leaving the cause unfinished. When you see “tree-cover loss,” first accept the observation for what it is. Then ask the next scientific questions: What kind of trees were there? Why were they lost? What happened afterward? Only then decide whether the evidence supports the stronger word deforestation.

Small Group Tutorials