07 — Heat Islands Follow Sky View (`heat-island-skyview`)

Supplied working report; original verdicts not validated.
Review status: Needs review. This is an archival source note, not a passed proof.
REPORT.md

# 07 — Heat Islands Follow Sky View (`heat-island-skyview`)

- Source: Incoprea Theory, Volume XXIX (`heat-island-skyview`)
- Claim: Urban heat island intensity is set by sky-view factor more than by green cover, because radiative loss dominates sensible cooling at night.
- Prediction: Intensity should track sky-view factor (SVF) better than canopy fraction.
- Refutation: Intensity tracking green cover would remove geometry.

## Test performed

Literature-based assessment (evidence class b), plus a small quantitative
re-analysis of published regression equations (class b, computed here). We
searched for an open, quality-controlled dataset with joint SVF / green-fraction
/ nocturnal UHI measurements across many sites (e.g., Zenodo/OSF/city portals)
to run a fresh correlation; none that was both openly downloadable and
sufficiently documented could be obtained within this run, so the test rests on
the published measurement literature (Oke 1981; Unger 2004's comprehensive
review; Svensson 2004; Johnson 1985; Bärring et al. 1985; Park 1987) with the
equations digitized and compared in `code/analyze_svf_literature.py`.

## Data (sources, URLs, sample sizes)

- Published equations and correlation coefficients digitized from Unger (2004),
  *Climate Research* 27:253–264 ([open PDF](https://www.int-res.com/articles/cr2004/27/c027p253.pdf)),
  which tabulates Oke (1981), Park (1987), Yamashita et al. (1986), Johnson (1985),
  Bärring et al. (1985) and adds 107 Szeged measurement cells.
- Oke T.R. (1981) *J. Climatol.* 1:237–254, https://doi.org/10.1002/joc.3370010304
- Svensson M.K. (2004) *Meteorol. Appl.* 11:201–211, https://doi.org/10.1017/S1350482704001288
- Derived outputs: `data/test07_svf_uhi_equations.csv`, `data/test07_szeged_seasons.csv`,
  `figures/test07_svf_sensitivity.png`.

## Result (key numbers; reference figure files)

Published ΔTₐ,max–SVF equations (all significant at the 1% level):

| source | domain | equation | °C per 0.1 SVF |
|---|---|---|---|
| Oke (1981) | 18 North American cities | ΔT = 16.34 − 15.00·SVF | 1.50 |
| Oke (1981) | 11 European cities | ΔT = 13.20 − 10.00·SVF | 1.00 |
| Park (1987) | 13 Japanese cities | ΔT = 10.15 − 12.00·SVF | 1.20 |
| Park (1987) | 6 Korean cities | ΔT = 12.23 − 14.00·SVF | 1.40 |
| Unger (2004) | Szeged, 107 areal cells | ΔT = 5.90 − 4.62·SVF, r = −0.69 | 0.46 (areal means) |

So across 48 cities on three continents the nocturnal UHI maximum scales with
SVF at roughly 1.0–1.5 °C per 0.1 SVF — geometry sets the intensity
(`figures/test07_svf_sensitivity.png`, left panel). Unger's Szeged areal study
explains 47% of intra-urban ΔT variance with SVF alone (r = −0.69, n = 107).
Svensson (2004, Göteborg) likewise reports a "fairly strong" night-time
air-temperature–SVF relationship, and Johnson (1985, Birmingham) a close
negative link between maximum cooling rate and SVF.

The green-cover comparison (Unger 2004, same 107 Szeged cells): the SVF–ΔT
relationship is essentially unchanged between the green season (r = −0.665,
slope 4.44, 44% of variance) and the defoliated season (r = −0.695, slope 4.81,
48% of variance) — a difference of only ~4 percentage points of explained
variance. Unger concludes "the role of the foliated trees in determining the
Tₐ along the measurement routes is minor compared to the obstruction effect of
buildings" (`figures/test07_svf_sensitivity.png`, right panel). This is the
direct experimental comparison the hypothesis calls for: geometry (SVF), not
canopy fraction, carries the nocturnal UHI signal.

## Verdict

**Support** — published measurements show nocturnal UHI intensity tracks
sky-view factor strongly (r ≈ −0.69 at street level; ~1–1.5 °C per 0.1 SVF
across 48 cities), while leaf-on vs leaf-off canopy changes barely alter the
relationship, consistent with radiative loss dominating at night. Literature-based
assessment; no new field data were collected.

## Limitations

Evidence class (b): the analysis re-uses published summary equations, not raw
point clouds, so no new correlations were computed from primary measurements.
SVF–UHI coupling is strongest for calm, clear nights and for near-surface
temperatures (Bärring et al. 1985 note Tₐ is less coupled than surface Tₛ).
Green cover demonstrably matters more in daytime (shading, evapotranspiration),
so the hypothesis should be read as a night-time claim. Correlation does not
prove mechanism, though the radiative-loss physics (Oke 1981) matches.

## References

- Oke T.R. (1981) Canyon geometry and the nocturnal urban heat island: comparison of scale model and field observations. *J. Climatol.* 1:237–254. https://doi.org/10.1002/joc.3370010304
- Unger J. (2004) Intra-urban relationship between surface geometry and urban heat island: review and new approach. *Climate Research* 27:253–264. https://www.int-res.com/abstracts/cr/v27/n3/p253-264/
- Svensson M.K. (2004) Sky view factor analysis – implications for urban air temperature differences. *Meteorol. Appl.* 11:201–211. https://doi.org/10.1017/S1350482704001288
- Johnson D.B. (1985) Urban modification of diurnal temperature cycles in Birmingham, UK. *J. Climatol.* 5:221–225.
- Bärring L., Mattsson J.O., Lindqvist S. (1985) Canyon geometry, street temperatures and urban heat island in Malmö, Sweden. *J. Climatol.* 5:433–444.
- Park H.-S. (1987) Variations in the urban heat island intensity affected by geographical environments. Environmental Research Center Papers 11, University of Tsukuba.
Included code and results

tests/07-heat-island-skyview/code/analyze_svf_literature.py