How much of the Andrews is old growth? Checking the 40% against two open maps

HJ Andrews says about 40% of the forest was old growth before the 2020 fires. Neither open map gives 40% inside the forest boundary: the 1988 Landsat stand-age map puts 46.1% in its old-conifer class, close to the 45% a 1984 introduction gave for trees over 400 years, while the Forest Service stand-origin map puts 30.0% in stands that began before 1600 and 59.3% in stands that began before 1800. Counted as 400 years or more by 2023 instead of by 2000, the same map gives 49.6%, so 40% falls between the two reference years.

How we did Matches except 0 of 5 exact, 1 within tolerance, 4 explained all studies
HJ Andrews Fast Facts: about 40% of the forest old growth before the 2020 fires · tolerance 2.5 percentage points (the page says 'we estimated' and gives a round 40%)
  • Old conifer class, 1988 Landsat TM map (by cell centre)40.0% → 46.1% +6.1 ptsexplained why
  • Stands that began before 1600, Forest Service stand-origin map (area-weighted)40.0% → 30.0% −10.0 ptsexplained why
  • Stands 400 years or more by 2023 (origin 1623 or earlier), context only: the page does not say which year it counts from40.0% → 49.6% +9.6 ptsexplained why
1984 introduction: dominant trees over 400 years cover about 45 percent · tolerance 2.5 percentage points ('about 45 percent')
  • Stands that began before 1600 (the same definition: 400 years or more by 2000)45.0% → 30.0% −15.0 ptsexplained why
  • Old conifer class, 1988 Landsat TM map (a spectral class with no age threshold)45.0% → 46.1% +1.1 ptswithin tolerance
46.1%old conifer in the 1988 Landsat map, inside the boundary (46,986 of 101,910 cells); Fast Facts says 40%
[5]
30.0%in stands that began before 1600 (over 400 years old by 2000), Forest Service stand-origin map
[28]
49.6%in stands that began in 1623 or earlier (400 years or more by 2023), the same map
[31]
170 of 170steps replay identically in a fresh session

Numbers in brackets, like [14], are step numbers in this study's run.json: the NativeGIS call that returned the figure. NativeGIS 0.3.0, recorded 2026-10-31 15:57 UTC.

Map of the HJ Andrews Experimental Forest coloured by the 1988 Landsat stand-age classes, from pale mixed open to dark old conifer, with the forest boundary in black
Stand-age classes from the 1988 Landsat TM image (TV06105) inside the forest boundary (GI006): old conifer (dark green) covers 46.1% of the forest. Rendered by gis_render in the recorded run.

Old conifer in 1988 against stand origin

The 1988 old-conifer class (green) under the TV06104 stands that began before 1600 (red) and in 1600–1799 (blue), both as published. The polygons sit about 175 m east of the grid: about 94 m of that is TV06104's own offset and about 80 m the grid's; see Why any numbers differ.

Old conifer (1988) and stand origin (TV06104): map of the 1988 old-conifer class in green with the before-1600 and 1600-1799 stand-origin polygons outlined in red and blue
Old conifer (1988) and stand origin (TV06104)

Click or tap a map to open it full screen, then zoom and pan.

What was published

HJ Andrews' Fast Facts page says that before the fires since 2020 "we estimated that 40%" of the Experimental Forest was covered in old growth. Its "Old-growth trees area" page describes old growth by 400- to 500-year-old emergent Douglas-fir and a complex vertical structure. Neither page names the map or the year behind the 40%. A 1984 introduction to the forest gives a different figure: stands whose dominant trees are over 400 years old cover about 45 percent.

Web page and a 1984 introductionHJ Andrews Experimental Forest, "Fast Facts" and "Old-growth trees area" (andrewsforest.oregonstate.edu); an introduction to the H.J. Andrews Experimental Forest dated 1984 (scanned copy, author not given), hosted for a Western Oregon University course.

Fast Facts (HJ Andrews) · Old-growth trees area (HJ Andrews) · 1984 introduction (scanned PDF)

Published method. Not given. The question here is which open map gives the figure, and under which definition of old growth.

The data and where it came from

All of it is open: downloaded without an account or a form.

DataWhereLicence
Stand-age classification of a 1988 Landsat TM image (TV06105; Cohen & Fiorella, 1994): grid tmveg88, 25 m cells, ArcInfo E00 export, NAD83 / UTM zone 10N. Classes from tv06105_fgdc.xml: 1 ice snow, 2 water, 7 mixed open, 8 mixed semi-open, 9 mixed closed, 10 young conifer, 11 mature conifer, 12 old coniferHJ Andrews LTER, 524 kB zip, no form or account
https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/tv06105.zip
HJ Andrews data policy (CC BY, acknowledgement requested)
Willamette National Forest vegetation polygons clipped to the Andrews (TV06104, hjaveg8.shp; compiled July 2000, published as the 2002 classification), with YR_ORIGIN, the year of origin of natural and managed standsHJ Andrews LTER, 1.9 MB zip, no form or account
https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/tv06104.zip
HJ Andrews data policy (CC BY, acknowledgement requested)
The 2008 version of the same stand polygons (TV06101, hja_veg2008.shp), used only to check where TV06104 sits: its managed stands carry real origin years, its natural stands the placeholder 1111HJ Andrews LTER, zip, no form or account
https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/tv06101.zip
HJ Andrews data policy (CC BY, acknowledgement requested)
HJ Andrews Experimental Forest boundary (GI006, boundary83.shp)HJ Andrews LTER, zip, no form or account
https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/gi00601.zip
HJ Andrews data policy (CC BY, acknowledgement requested)
Gauged watershed boundaries from a 10 m DEM (HF014-04, wsheddem.shp), used only to check how the two maps line upHJ Andrews LTER, zip, no form or account
https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/hf01404.zip
HJ Andrews data policy (CC BY, acknowledgement requested)

The input files the run read, pinned by checksum

File (relative to the data folder)Bytessha256
samples/hja-open/tv06105/x/tv06105.e003,879,13793b1638e9dbb7c97…
samples/hja-open/gi00601/x/boundary83.shp10,5400774de0a9ebd9693…
samples/hja-open/tv06104/x/hjaveg8.shp2,754,280153dac389d063426…
samples/hja-open/tv06101/x/hja_veg2008.shp3,098,488ef83a267c95a7d1f…
samples/hja-open/hf01404/x/wsheddem.shp149,8766cbafbe316d2fa08…

Full checksums are in run.json. Replay refuses an input whose checksum differs.

Method

  1. Open the 1988 stand-age grid straight from its E00 export (gis_open; the CRS, EPSG:26910, comes from the file's PRJ section), and the forest boundary.
  2. Count each class inside the boundary by cell centre with one zonal class tally over the boundary (gis_process zonal classes:true): cells, and the share of each class.
  3. Cross-check by area: polygonize the grid into regions of one class, clip the regions to the boundary, measure them and summarize the area by class (polygonize, clip, measure, summarize).
  4. Stand origin: clip the TV06104 polygons to the boundary (area-weighted), bin YR_ORIGIN with a field calculation (before 1600, 1600–1799, 1800–1899, 1900–1949, 1950 or later, and no year), add each piece's area and summarize area by bin, and by single origin year for the shares at other cutoffs (clip, fieldCalc, summarize).
  5. Cross the two maps: count old-conifer cells and open-to-young cells (classes 7–10) in each origin bin (zonal over the binned polygons, then summarize by bin). Repeat with the polygons moved 100, 175 and 250 m west and 25 m south (translate) to find the shift that lines the two maps up.
  6. Which map is off: (1) the centroids of TV06104's stands over 2 ha against the same stands in TV06101, the 2008 version of the map (centroid, listed with stand number and area; the page pairs stands present once in each with areas within 5% and takes the median offset); (2) the share of three watersheds clearcut whole between 1962 and 1975 (WS1, WS6, WS10, HJ Andrews' DEM boundaries) in stands of 1950 or later, for each stand map where published and moved 94 and 175 m west; (3) the 1988 grid's open-to-young share in TV06101's stands of 1950 or later, with the stands moved 0 to 120 m west.
  7. Render the 1988 classes inside the boundary, and the old-conifer class against the stands of origin before 1600 and 1600–1799.

The run made 170 NativeGIS calls: gis_open × 5, gis_query × 34, gis_process × 73, gis_close × 54, gis_render × 4.

Published numbers next to ours

The 1988 stand-age classes inside the forest

1988 TM class (TV06105)Cells (25 m)HectaresShare by cell centreShare by clipped area
12 old conifer46,9862,93746.1%46.1%
11 mature conifer17,0251,06416.7%16.7%
10 young conifer15,87099215.6%15.6%
9 mixed closed5,1043195.0%5.0%
8 mixed semi-open13,87786713.6%13.6%
7 mixed open3,0401903.0%3.0%
0, 1, 2 unclassified, ice and snow, water80.50.01%0.01%
Inside the boundary101,9106,369100%6,369 ha

Cell counts from one zonal class tally over the boundary [4, 5] (a cell counts when its centre is inside); clipped areas [14]. Published: Fast Facts 40% old growth before the 2020 fires; 1984 introduction about 45% in stands with dominant trees over 400 years. Units: 25 m cells (0.0625 ha) and hectares in NAD83 / UTM zone 10N.

Stand origin, and how much of it was old conifer in 1988

Stand origin (TV06104 YR_ORIGIN)Polygon piecesHectaresShare of the forestOld conifer in 1988, as publishedOld conifer in 1988, polygons moved 175 m W
before 1600251,91330.0%48.0%58.4%
1600–1799381,86729.3%59.9%69.5%
1800–18995170711.1%40.1%35.8%
1900–19497711.1%33.3%29.6%
1950 or later1581,52223.9%31.8%11.1%
no year (rock, meadow, water, buildings)852894.5%38.2%22.5%

Areas [28] (polygons clipped to the boundary, area-weighted); old-conifer share of each bin's 25 m cells [39, 64]. Open-to-young share of the 1950-or-later stands with the polygons moved 0, 100, 175 and 250 m west of the grid: 57.1%, 71.6%, 82.2%, 74.4% [39, 51, 64, 77]. The 175 m is the two maps' offsets added together; see below.

Shares by origin cutoff

Stands of TV06104 that began…Over 400 years old inShare of the forest
before 1600200030.0%
before 1620202032.9%
in 1623 or earlier2023 (400 years or more)49.6%
before 1700–51.6%
before 1800–59.3%

Area by origin year inside the boundary [29, 31], as a share of the clipped total. Origin years recorded from 1600 to 1623: 1610, 1620, 1623; the 1623 cohort alone is 16.7% of the forest. Context, not a match: Fast Facts does not say which map, year or threshold its 40% used.

How the maps line up

Registration testWhere publishedMoved 94 m W, 18 m SMoved 175 m W, 25 m S
TV06101 (2008): share of WS1, WS6, WS10 in stands of 1950 or later94.6%91.1%85.9%
TV06104 (2000): the same89.0%94.8%93.5%
1988 grid: open-to-young cells (classes 7–10) in…Where publishedMoved 40 m WMoved 80 m WMoved 120 m WMoved 80 m W, 20 m S
TV06101's stands of 1950 or later70.5%77.2%83.4%82.6%83.4%
WS1, WS6, WS1074.1%–75.2%––

Stand centroids of TV06104 and TV06101 over 2 ha [88, 89, 90, 98, 99, 100]: 459 stands present once in each with areas within 5%; TV06104 minus TV06101, median 93.8 m east and 17.9 m north. Watershed shares [143, 153, 164, 113, 123, 134]; grid shares [172, 176, 181, 186, 191, 195, 198]. Watershed boundaries from HJ Andrews' 10 m DEM (HF014-04).

Why any numbers differ

Replay it yourself

The run file records every call, its arguments, the checksums of the files it read and the numbers it returned. Replaying it repeats every call in a fresh NativeGIS session and checks that each answer comes out the same.

Verified reproducible on 2026-10-31: 170/170 steps identical. The run file was replayed in a fresh NativeGIS session (0.4.0 (source 3692d6e)) against a folder holding nothing but the public downloads below, fetched with these exact commands.
  1. Make an empty data folder and fetch the inputs into it exactly as below (bash; the folder layout is the one the run file expects):
    cd /path/to/an/empty/folder
    mkdir -p samples/hja-open/tv06105 && curl -fL -o samples/hja-open/tv06105/tv06105.zip https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/tv06105.zip && unzip -o samples/hja-open/tv06105/tv06105.zip -d samples/hja-open/tv06105/x
    mkdir -p samples/hja-open/tv06104 && curl -fL -o samples/hja-open/tv06104/tv06104.zip https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/tv06104.zip && unzip -o samples/hja-open/tv06104/tv06104.zip -d samples/hja-open/tv06104/x
    mkdir -p samples/hja-open/tv06101 && curl -fL -o samples/hja-open/tv06101/tv06101.zip https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/tv06101.zip && unzip -o samples/hja-open/tv06101/tv06101.zip -d samples/hja-open/tv06101/x
    mkdir -p samples/hja-open/gi00601 && curl -fL -o samples/hja-open/gi00601/gi00601.zip https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/gi00601.zip && unzip -o samples/hja-open/gi00601/gi00601.zip -d samples/hja-open/gi00601/x
    mkdir -p samples/hja-open/hf01404 && curl -fL -o samples/hja-open/hf01404/hf01404.zip https://andrewsforest.oregonstate.edu/sites/default/files/lter/data/studies/spatialdata/hf01404.zip && unzip -o samples/hja-open/hf01404/hf01404.zip -d samples/hja-open/hf01404/x
  2. Download run.json (637,823 bytes, sha256 462c3ae4bbb105954bb60c6efcedf2a2485df0096a6d30f6e6a832dc62b10be7).
  3. Ask your assistant to replay it with NativeGIS, or call the tool directly:
gis_replay {"action": "replay",
            "file": "<absolute path to the run.json you downloaded>",
            "inputRoot": "<your data folder>",
            "outputRoot": "<a new, empty folder>"}

Paths in the run file are relative to the data folder (inputRoot), and outputs go under outputRoot, so nothing depends on where the original run lived.

Files

Limits

NativeGIS reproductions · every NativeGIS number on this page is read from the recorded run file it links to.