Official statistics
Detected loss and official harvest figures
Satellite-detected forest loss beside the official area harvested in each province, year by year.
Compare with official harvest statisticsWitness Tree shows forest loss in four Canadian provinces, from satellite images and public records.
Four provinces
Lost at least once
44,975,298.15 ha
22.69% of the 198,250,714.71 ha of forest mapped in 1984. A place counts once, however many times it was cleared.
Yearly losses added together
70,030,574.55 ha
A place cleared in two different years counts twice, so this is larger. It has no percentage, because it can count the same forest more than once.
Both are minimums. The satellite source covers only Canada’s forest regions, so 46,424,717.91 ha of these provinces have no data. Loss there counts as unknown, not zero.
Every figure on this site is marked with the kind of evidence behind it.
How much forest satellites detected as lost in each province, 1984–2022.
How much forest was detected as lost
Satellite observation
18,028,529 ha
25.97% of the forest mapped in 1984 was lost at least once. In 1984, 47% of the province was forest. The satellite source didn’t cover 15.05% of the province: mostly in the far north, beyond where dense forest ends; the rest is in the settled south. Any loss there counts as unknown, not zero.
15.05% of the province has no satellite data: mostly in the far north, beyond where dense forest ends; the rest is in the settled south
18,028,529.19 ha detectedSource and limits
Satellite observation
10,823,353 ha
19.57% of the forest mapped in 1984 was lost at least once. In 1984, 60% of the province was forest. The satellite source didn’t cover <0.01% of the province: mostly along the shoreline, where different versions of the provincial boundary disagree. Any loss there counts as unknown, not zero.
<0.01% of the province has no satellite data: mostly along the shoreline, where different versions of the provincial boundary disagree
10,823,352.93 ha detectedSource and limits
Satellite observation
9,280,647 ha
19.07% of the forest mapped in 1984 was lost at least once. In 1984, 50% of the province was forest. The satellite source didn’t cover 9.03% of the province: in the settled south, below 52° north, outside the forest areas the source maps. Any loss there counts as unknown, not zero.
9.03% of the province has no satellite data: in the settled south, below 52° north, outside the forest areas the source maps
9,280,647.27 ha detectedSource and limits
Satellite observation
6,842,769 ha
27.52% of the forest mapped in 1984 was lost at least once. In 1984, 39% of the province was forest. The satellite source didn’t cover 24.02% of the province: mostly in the prairies and the settled south, outside the forest areas the source maps; the rest lies between 52° and 56° north. Any loss there counts as unknown, not zero.
24.02% of the province has no satellite data: mostly in the prairies and the settled south, outside the forest areas the source maps; the rest lies between 52° and 56° north
6,842,768.76 ha detectedSource and limits
Why these areas were not mapped, and what we know about them
The record covers these four provinces only.
A satellite can see that trees are gone, but not why. These pages bring in other sources: official harvest statistics, British Columbia’s timber volumes, and Natural Resources Canada’s yearly maps of harvest and fire.
Official statistics
Satellite-detected forest loss beside the official area harvested in each province, year by year.
Compare with official harvest statisticsHarvest and fire
Natural Resources Canada’s maps of forest cleared by harvest and by fire, each year. Choose provinces and years, then download the chart or its numbers.
Build a chartTimber volume
How much timber British Columbia billed each year, next to the amount the province allowed to be cut.
See BC harvest volumeThe share of each riding’s mapped forest detected as lost at least once from 1984 to 2022. Only ridings mapped in full, with at least 50,000 ha of forest, are ranked; the rest are left out, never counted as zero.
A share can be high because of harvest, fire or both; a satellite can’t tell why trees are gone. Detected loss is a minimum.
The share of each city’s mapped forest detected as lost at least once from 1984 to 2022. A city here is a place Statistics Canada lists as a city (City, Ville or Cité). Only cities mapped in full, with at least 5,000 ha of forest, are ranked; the rest are left out, never counted as zero.
A city’s boundary often takes in far more land than its built-up area, so most of this forest can lie well outside town. A satellite can’t tell whether the trees went to harvest, fire or building. Detected loss is a minimum.
How to read it
Each figure is marked with its kind of evidence, how sure we are and how much was mapped.
See the marks and labelsMethods
What counts as forest, how the figures are worked out, and how sure we are.
Read the methodsData
Download the province figures for 1984 to 2022, with their sources and limits.
Go to downloads and sourcesWitness Tree reports only what its sources record and what satellite images detect.
A satellite can see that trees are gone, but not why. How the methods work.
These figures for 1984 to 2022 are an early preview, not the final release. The loss patches on the Explore map are for looking at, not adding up, and no expert has reviewed them. Data, sources and licences.
Context source: Statistics Canada 2021 cartographic boundaries; province aggregate derived from Natural Resources Canada VLCE2. Source catalogue.