Changes per metric
Watching a dataset grow up.
Almost every movement across 42 releases is us getting the measurement right—admitting a zone, fixing a denominator, clipping to the right boundary—not the world changing. Each city's five headline controls—the share within 20 km of the CBD, the figure the Atlas leads with—are plotted release to release, with every explained move colour‑coded by why it happened.
Highly restricted
41–99% scaleSydney
80.9%
83% → 81%
Melbourne
43.0%
45% → 43%
Brisbane
86.6%
90% → 87%
Perth
86.5%
88% → 86%
Adelaide
91.8%
92% → 92%
Hobart
97.0%
97% → 97%
Canberra
74.4%
91% → 74%
Darwin
88.4%
88% → 88%
≤2 storeys
36–98% scaleSydney
77.2%
80% → 77%
Melbourne
37.3%
40% → 37%
Brisbane
85.6%
89% → 86%
Perth
85.9%
87% → 86%
Adelaide
90.8%
91% → 91%
Hobart
96.4%
97% → 96%
Canberra
74.3%
91% → 74%
Darwin
88.4%
88% → 88%
Low-density
13–76% scaleSydney
68.7%
69% → 69%
Melbourne
33.7%
36% → 34%
Brisbane
73.6%
75% → 74%
Perth
55.1%
56% → 55%
Adelaide
34.3%
34% → 34%
Hobart
14.2%
14% → 14%
Canberra
74.3%
74% → 74%
Darwin
73.3%
73% → 73%
Detached-only
0–75% scale · not measured before v0.11.0Sydney
3.1%
8% → 3%
Melbourne
2.0%
0% → 2%
Brisbane
66.3%
60% → 66%
Perth
0.0%
5% → 0%
Adelaide
0.0%
0% → 0%
Hobart
0.0%
0% → 0%
Canberra
0.0%
0% → 0%
Darwin
73.3%
73% → 73%
Heritage
0–15% scaleSydney
13.0%
14% → 13%
Melbourne
10.6%
11% → 11%
Brisbane
1.2%
1% → 1%
Perth
2.4%
1% → 2%
Adelaide
9.1%
9% → 9%
Hobart
3.5%
3% → 4%
Canberra
2.1%
2% → 2%
Darwin
0.0%
0% → 0%
Every share is the 20 km ex-rural cumulative—the same basis the rest of the site headlines, so Sydney's ~81% here is the ~81% on its city page. Each metric has its own vertical scale—heritage's inner-city share is far smaller than highly-restricted's ~80%. Dots mark releases with an explained move. Darwin joins at v0.7.0, when its ring data first reached 20 km; detached-only isn't measured before v0.11.0.
Every explained move
| Release | City | Δ pp (20km) | Explanation | Date |
|---|---|---|---|---|
| 1.9.0 | Melbourne | −1.4 | Real-world reformActivity Centres Program Stage 2 / Amendment GC285, gazetted 2026-07-22 — Prahran/South Yarra + 8 other centres' catchments rezoned to HCTZ1/HCTZ2. Residential area ~unchanged; restricted share drops 49.73%→49.26% GCCSA as HCTZ1/HCTZ2's height allowance is materially looser than what preceded it. Real-world policy | 2026-07-22 |
| 1.8.0 | Canberra | −0.1 | Coverage changeACT Territory Plan zones/roads/blocks refetched — small live-vs-local deltas, but no specific amendment was identified/dated; our local copy catching up to an already-current live service, not a new change in the law; heritage_scoped.geojson left untouched, no reproducible derivation | 2026-07-20 |
| 1.8.0 | Hobart | −0.3 | Coverage changeTAS zone/overlay/heritage layers refetched; legacy Interim scheme winding down as councils' Local Provisions Schedules take effect. Code Overlay fetch briefly picked up new non-heritage sub-types (Significant trees, archaeological potential, precincts) alongside Local heritage place, tripling the heritage-control share — Ethan called it back to Local-heritage-place-only (a methodology/scoping fix). Also: Hobart Commercial Zone under the HOB-S6.0 Specific Area Plan now counts as residential (Mixed Use, 4 storeys) — a real, pre-existing planning provision we'd been missing, not something that just changed. | 2026-07-20 |
| 1.8.0 | Adelaide | −0.1 | Real-world reformSA Planning & Design Code amendment refresh + rebuilt the derived LU-x-Zones layer, which had no prior committed script. | 2026-07-20 |
| 1.8.0 | Perth | −0.6 | Coverage changeR-Codes/CPS2/heritage layers refetched from live services — small live-vs-local deltas, but no specific amendment was identified/dated; our local copy catching up to an already-current live service, not a new change in the law | 2026-07-20 |
| 1.8.0 | Brisbane | +0.0 | Methodology refinementBCC/Moreton Bay zoning+heritage refetched. Fixed a ZONE_PREC_ field-collision bug that turned out to be a no-op (old export already resolved it the same way); the real bug was NULL-vs-zero handling on no-precinct zones, which briefly (mis)moved le-2-storey to 60.6% before Ethan traced it and the fetch script was fixed to zero-fill — final le-2-storey is back at 82.64%, unchanged from pre-#590. | 2026-07-20 |
| 1.8.0 | Melbourne | −0.3 | Real-world reformVIC zones/heritage overlay/register refetched; new PRZ zone classified Mixed Use / unknown-height rather than excluded. | 2026-07-20 |
| 1.8.0 | Sydney | −0.1 | Coverage changeNSW Zones/HOB/FSR/Heritage refetched from live services — small live-vs-local deltas, but no specific amendment was identified/dated; this reads as our local copy catching up to an already-current live service, not a new change in the law | 2026-07-20 |
| 1.6.0 | Brisbane | +0.1 | Methodology refinementPDA re-add clip extended to community-facilities/special-purpose base | 2026-07-13 |
| 1.5.0 | Brisbane | +0.4 | Methodology refinementPDA re-add clipped to exclude open-space/recreation base — golf course/parks no longer counted as unrestricted residential | 2026-07-13 |
| 1.4.0 | Sydney | −0.4 | Methodology refinementR2-in-TOD carved out of low_density — TOD R2 gets the SEPP apartment/22m uplift, no longer low-density; union −0.15pp, ≤2-storey unchanged | 2026-07-12 |
| 1.3.1 | Sydney | −0.0 | Methodology refinementHOB overlap precedence fix — overlapping HOB polygons are same-control precinct stacking, most-permissive wins; GCCSA HR ticks 86.9% → 86.8% on the rounding boundary, all moves sub-0.01pp citywide | 2026-07-12 |
| 1.3.0 | Sydney | ±0.0 | Methodology refinementSEPP dual-occ scoping fix — s.166 dual-occ is chapter-wide, not walking-catchment-gated; detached-only 7.2% → 4.5%, union unchanged | 2026-07-09 |
| 1.2.0 | Perth | −0.0 | Methodology refinementCadastre source switched Landgate Historical → PlanWA property-interest | 2026-07-07 |
| 1.1.0 | Adelaide | +0.0 | Methodology refinementSA DCDB cadastre refreshed, +2,021 parcels | 2026-07-07 |
| 1.1.0 | Perth | +0.2 | Coverage changeWA Heritage List (DPLH-090) added — 5,672 None→Local | 2026-07-07 |
| 1.0.0-rc.3 | Brisbane | −1.6 | Coverage changePriority Development Areas (PDA) re-included | 2026-06-27 |
| 1.0.0-rc.3 | Melbourne | −0.0 | Coverage changeCity-of-Melbourne SUZ6/7/8 added as Mixed use | 2026-06-27 |
| 1.0.0-rc.3 | Sydney | +0.7 | Methodology refinementC4 (Environmental Living) reclassified non-rural → enters the ex-rural denominator | 2026-06-27 |
| 0.20.0 | Perth | −0.7 | Methodology refinementRural-residential flagged rural, excluded from the ex-rural denominator | 2026-06-23 |
| 0.13.0 | Canberra | −16.8 | Real-world reformACT Missing Middle reforms — RZ2–RZ5 take multi-unit heights. Real-world policy (now gazetted/in effect; applied at v0.13.0 from the then-draft MPA, pre-gazettal). The first real-world change in the series (see also all of v1.8.0 and later Melbourne GC285 entries). | 2026-06-22 |
| 0.13.0 | Brisbane | −0.9 | Methodology refinementLMR2 uplift propagated across strata schemes | 2026-06-22 |
| 0.13.0 | Melbourne | ±0.0 | Methodology refinementClip to Greater Melbourne GCCSA, not a 60 km disc | 2026-06-22 |
| 0.12.0 | Brisbane | −1.2 | Methodology refinementLMR2 conditional 3-storey uplift near transit | 2026-06-20 |
| 0.7.0 | Darwin | — | Coverage changeCity of Palmerston council added, +11.8k lots | 2026-06-17 |
| 0.5.0 | Sydney | −0.5 | Methodology refinementRing origin → Central Station + 3-tier typology | 2026-06-17 |
| 0.4.0 | Adelaide | −0.2 | Methodology refinementAdelaide FC/UC zones admitted | 2026-06-16 |
| 0.4.0 | Perth | −0.0 | Methodology refinementPer-city admissions: commercial + R-Code | 2026-06-16 |
| 0.4.0 | Sydney | −2.0 | Methodology refinementSydney SEPP (Housing) height uplift, max(LEP,SEPP) | 2026-06-16 |
| 0.3.0 | Sydney | −0.2 | Methodology refinementSP5 Metropolitan Centre counted as residential | 2026-06-13 |