Mesopic Vision and Outdoor Lighting Design: S/P Ratio, CIE 191, and Energy-Efficient Street Lighting

1. Introduction: The Three Regimes of Human Vision

Human visual function operates across approximately 12 log units of ambient luminance, mediated by two distinct photoreceptor populations: cones (photopic, ≥ 3 cd/m²) and rods (scotopic, ≤ 0.001 cd/m²). The intermediate region—mesopic vision (0.001 to 3 cd/m²)—is where both rod and cone signals contribute to visual perception. This is precisely the luminance regime of outdoor lighting: typical road surfaces under street lighting produce luminances of 0.3–2.0 cd/m² (M- and ME-series lighting classes per EN 13201-2), well within the mesopic range.

Importantly, the spectral sensitivity of the visual system in the mesopic range is not constant; it shifts from the photopic V(λ) curve (peak 555 nm, dominated by L- and M-cones) toward the scotopic V'(λ) curve (peak 507 nm, dominated by rods) as luminance decreases. This Purkinje shift means that light sources with relatively higher short-wavelength (blue-cyan) content become more effective for peripheral target detection under mesopic conditions—a fact with profound implications for street lighting energy efficiency.

2. The S/P Ratio: A Practical Index

The scotopic/photopic (S/P) ratio of a light source is the ratio of luminous flux evaluated under the scotopic spectral luminosity function V'(λ) to that evaluated under the photopic V(λ) function:

S/P = [K'(λ_max) · ∫ E_λ(λ) · V'(λ) dλ] / [K(λ_max) · ∫ E_λ(λ) · V(λ) dλ]

where K'(λ_max) = 1700 lm/W (scotopic) and K(λ_max) = 683 lm/W (photopic). The S/P ratio is a source-specific parameter that can be reliably computed from the spectral power distribution (SPD) of any light source.

Light Source	CCT (K)	S/P Ratio	Relative Mesopic Efficacy Factor (at 0.3 cd/m²)*
High-pressure sodium (HPS)	2100	0.62	0.78 (baseline: 1.00 for HPS)
Warm white LED	3000	1.12–1.35	1.08–1.18
Neutral white LED	4000	1.50–1.80	1.25–1.40
Cool white LED	5000	1.80–2.20	1.40–1.55
Daylight LED	6500	2.10–2.50	1.45–1.60
Metal halide (ceramic)	4200	1.45–1.60	1.20–1.30
Low-pressure sodium (LPS)	1800	0.23	0.55

* Calculated per CIE 191:2010 unified luminance framework for an average road surface luminance of 0.3 cd/m² (ME4a/ME5 class). Higher values indicate greater visual effectiveness per photopic watt.

Key implication: A 4000 K LED (S/P ≈ 1.65) with a photopic output of 5000 lm is, under mesopic conditions at 0.3 cd/m², visually equivalent to an HPS lamp producing approximately 5000 × (1.35 / 0.78) = 8650 photopic lm. This means that the LED street light can be dimmed to approximately 58% of its photopic output while maintaining the same visual performance as the HPS lamp—representing a 42% reduction in energy consumption with no loss of visual function for the road user.

3. The CIE 191:2010 Unified System for Mesopic Photometry

CIE 191:2010 (recently updated as CIE S 026/E:2024 for α-opic metrology) defines a practical system for mesopic photometry based on the MES2 model, which features two fundamental innovations:

Unified luminance: A single-valued mesopic luminance L_mes (in cd/m²) that represents the visual effectiveness of a light source at mesopic levels, calculated as a weighted combination of photopic and scotopic luminances:

L_mes = [m(V') · L_s + (1 − m(V')) · L_p] / [m(V') · (V'/V)_max + (1 − m(V'))]

where L_s is the scotopic luminance, L_p the photopic luminance, and m(V') is the adaptation coefficient (0 for photopic, 1 for scotopic). The function m(V') depends on the adaptation luminance L_p of the surround.

Mesopic visual performance (MVP) factor: A correction factor applied to photopic illuminance values for outdoor lighting design, defined for each light source S/P ratio and each lighting class (M1–M6 per EN 13201-2).

4. Application to Real Street Lighting Retrofits

4.1 Energy Savings by Lighting Class

EN 13201 Class	Target L_avg (cd/m²)	HPS 150 W Lumens	4000 K LED (Same Visual Performance)	Power Saving
ME1 (highway)	2.0	22000 lm (155 W)	17500 lm (92 W)	41%
ME2	1.5	18000 lm (125 W)	13900 lm (74 W)	41%
ME3a/b	1.0	13500 lm (100 W)	9900 lm (56 W)	44%
ME4a/b	0.75	9500 lm (75 W)	6800 lm (42 W)	44%
ME5 (residential)	0.5	7200 lm (60 W)	4800 lm (31 W)	48%
ME6 (low-traffic)	0.3	5000 lm (45 W)	3000 lm (22 W)	51%

Note: The power savings diminish somewhat when practical constraints are accounted for: minimum luminaire spacing (typically 25–40 m), uniformity requirements (U₀ ≥ 0.4 for ME classes), and threshold increment (TI < 10% for veiling luminance control). Real-world retrofit savings from HPS-to-LED conversion are typically 30–50% for the same lighting class—significantly better than the 20–30% predicted by purely photopic calculations.

4.2 Glare Considerations: The S/P Paradox

While high-S/P sources improve peripheral target detection, they also produce greater disability glare under mesopic conditions. Disability glare (veiling luminance L_v) is proportional to the illuminance at the eye from the glare source (E_glare) divided by the square of the glare angle (θ) relative to the line of sight. Under mesopic conditions, the effective veiling luminance is amplified because scattered blue light (which rods are more sensitive to) contributes disproportionately to the retinal veiling luminance. This means that high-CCT street lighting (5000 K+) may achieve inferior glare control (higher TI) compared to neutral 4000 K sources at the same photopic luminance—a counterintuitive result that argues against the extreme cool-white (6500 K) street lighting seen in some retrofit installations. The optimal CCT for street lighting, considering both energy efficiency and visual comfort, appears to be in the 3500–4500 K range (S/P 1.4–1.8), consistent with guidance in CIE 115:2010 and ANSI/IES RP-8-21.

5. Adaptive Road Lighting Systems

Modern intelligent street lighting systems use real-time dimming based on traffic volume, time of night, and ambient moonlight. The mesopic benefit of high-S/P sources is particularly advantageous in adaptive systems: during low-traffic hours (00:00–05:00), when lighting can be dimmed to 30–50% of nominal levels, the effective luminance entering the mesopic range (0.1–0.5 cd/m²) means that the S/P ratio becomes even more impactful. A 4000 K LED dimmed to 40% photopic output (producing ≈ 0.25 cd/m²) still provides a mesopic equivalent luminance of approximately 0.35 cd/m²—equivalent to an HPS at 60% output but at half the power consumption. Adaptive systems using DALI-2 (IEC 62386-207: lighting control for street lighting) and LoRaWAN communication can achieve total energy savings of 55–70% compared to conventional HPS installations.

6. Standards and References

CIE 191:2010: Recommended System for Mesopic Photometry Based on Visual Performance. International Commission on Illumination.
CIE 115:2010: Lighting of Roads for Motor and Pedestrian Traffic.
EN 13201-2:2015: Road Lighting — Part 2: Performance Requirements.
ANSI/IES RP-8-21: American National Standard Practice for Roadway Lighting.
Rea, M.S. and Bullough, J.D. (2016). On the estimation of photometric and colorimetric parameters for mesopically scaled sources. LEUKOS, 12(1–2), 25–38.
Fotios, S. and Cheal, C. (2013). Obstacle detection: A pedestrian-facing approach to road lighting. Lighting Research & Technology, 45(4), 423–439.
Bullough, J.D. and Rea, M.S. (2004). Visual performance under mesopic conditions: Consequences for roadway lighting. Transportation Research Record, 1862(1), 89–94.
IES LM-50-21: Photometric Measurement of Roadway Lighting Installations.

7. Related Articles

Sources: CIE 191:2010 · CIE 115:2010 · EN 13201-2 · ANSI/IES RP-8-21 · Rea & Bullough 2016 · Fotios & Cheal 2013
Disclaimer: This article is for educational reference. Road lighting design should be performed by a qualified lighting engineer per applicable local codes.