Urban heat island in a dense tropical CBD

The heat island is present by day but is stronger at night: the difference is 2.7 K at 14:00 (34.1 – 31.4) but 4.2 K at 02:00 (28.2 – 24.0). Any explanation must account for both the existence of the difference and its timing.

Daytime:

• The CBD has 85% impervious cover, so less energy is used in evapotranspiration and more is stored in concrete, asphalt and brick.
• Dark surfaces absorb shortwave radiation, and the low sky-view factor of 0.4 means less longwave radiation escapes to the sky.
• Anthropogenic heat from traffic, air-conditioning and commerce adds energy.

Nighttime:

• The stored heat is released slowly from the high thermal mass of the CBD, so A cools more slowly after sunset.
• The low sky-view factor traps outgoing longwave radiation in the street canyons.
• Wind is weaker at A (1.2 m s⁻¹) than at B (2.1 m s⁻¹), so there is less turbulent mixing and less advection of cooler air.
• The park at B cools more effectively through evapotranspiration and a larger sky-view factor; its higher humidity at 02:00 (81% vs 62%) is consistent with active evapotranspiration and less runoff.

Judgement: the daytime difference is mainly surface albedo, impervious cover and storage; the larger nighttime difference is mainly release of stored heat, canyon geometry and suppressed ventilation. The clear, calm conditions maximise the contrast because there is little cloud or wind to mix the boundary layer.

Cool roofs and pale pavements are likely to be more effective at 02:00, but they are constrained by the building stock; trees are better for daytime comfort.

Cool roofs/pale pavements:

• They raise albedo and reduce the heat stored during the day, so there is less heat to release at night. That directly targets the 4.2 K nighttime difference.
• They can be applied to large commercial roofs and roads without taking land.
• However, many conserved shophouses have pitched tiled roofs that cannot easily be converted; heritage rules may prevent reflective coatings. Narrow streets limit the area of pavement that can be treated.

Linear parks and street trees:

• Trees provide shade and evapotranspiration, improving daytime thermal comfort and reducing surface temperatures.
• At 02:00, transpiration is low, and a dense canopy can actually reduce sky-view cooling, so the effect on the 02:00 air temperature difference may be modest.
• Narrow streets, underground utilities and heavy traffic make planting difficult and expensive; trees need water and maintenance.

Judgement: neither option alone will remove the heat island. Cool roofs on the large commercial buildings plus targeted tree planting where space allows would address both day and night. If the single criterion is the 02:00 difference, cool roofs and pale surfaces are the stronger choice because they reduce stored heat; if the criterion is human comfort during the day, trees and shade are stronger. The S$40 million should not be split evenly without checking roof eligibility and street width.

At 02:00, station B is cooler (24.0°C) but much more humid (81%) than station A (28.2°C, 62%). Temperature alone therefore overstates the comfort advantage of the park.

• Evaporative cooling depends on the humidity gradient. In humid air, sweat evaporates more slowly, so the body cannot lose heat as effectively. The park's higher humidity partly offsets its lower air temperature.
• Heat stress indices combine temperature and humidity. A cooler but saturated environment can produce a similar apparent temperature to a warmer but drier one. Without calculating a heat index or wet-bulb temperature, the comparison is incomplete.
• The humidity difference also explains the energy balance. The park has active evapotranspiration and more soil moisture; the CBD has runoff and less vegetation, so its air is drier. That is why the two stations differ in both temperature and moisture.

Judgement: the park is still likely to be more comfortable because 24.0°C with 81% humidity is generally less stressful than 28.2°C with 62%, but the difference is smaller than the raw temperatures suggest. Any mitigation appraisal should use a heat-stress metric, not temperature alone.