Urban Heat Island Dynamics in Singapore

Objectives

To measure screen-level air temperature and humidity fields over Singapore.

Methodology

(i) Observations from car traverses which pass through various land-use types at fixed times (21:00-23:00 LT) during clear, calm nights

Singapore is a city-state located at the Southern tip of peninsular Malaysia. The island is separated from the mainland by the Straight of Johor which is 1 - 3 km wide. Most of the island is built up with either residential and commercial properties or industrial facilities. Extensive non-developed, green areas can be found in the center around Upper Selatar and Upper Pierce reservoirs and Bukit Timah Nature Reserve and in the North-West where vast cemeteries, agricultural farms and military areas are located. A major land reclamation project can be seen on the Eastern tip of the island adjacent to Changi International Airport (see image below). Main industrial areas are located in the West and on Jurong Island in the South-West which houses petrochemical facilities (hidden below clouds).

The heat island traverse route (red) leads through several distinct land-use types:

(1), (9) Institutional (start and end point) (2) Industrial (medium-size, low-rise factories with tree lined streets) (3) Low-density residential (attached, low-rise "terraces" houses) (4) Open, less-developed (5) High-density residential (HDB: < 20 storey apartment blocks) (6) Detached residential (single family dwellings with gardens and pools) (7) Commercial (Orchard road shopping district with tree-lined streets) (8) CBD (Financial district with up to 60 storey buildings)

The sensors are installed at a height of about 1.7 m on the left side of a pick-up truck with a semi-open canopy. Air temperature and relative humidity are measured with a shielded, aspirated fine-wire thermocouple (ASPTC) and a HUMICAP (HMP45C), respectively.

Signals are sampled at 1 Hz on a 21X data logger. 5-sec averages are continuosly downloaded to a laptop using PC208W for on-line quality control and further processing. Data are corrected for cooling during the traverse (usually < 0.5 ºC/2hr) and spurious points (e.g. due to stoppage at red-lkights, passing of buses or trucks) are removed with the help of a detailed log maintained during the traverse.

These UHI surveys constitute part of student assignments for a 3rd-year course "Microscale Climates" which is taught in the second semester (January - April). The traverses are therefore usually conducted in February/March which corresponds to the NE Monsoon season. Though the term monsoon implies heavy, sustained rains, this period is characterized by summy, dry conditions and most nights are surprisingly clear.

(ii) Observations from car traverses passing through limited range of land-use types repeated at hourly intervals during selected nights

This work was primarily carried out by Winston Chow (g0202394@nus.edu.sg) as his honours thesis project.

The same sensors used in (i) were employed to measure the temporal evolution of the temperature and humidity distribution in a limited part of Singapore (locations 2 - 5 on map). Traverses were conducted in hourly intervals starting after sunset until the early morning hours. Because of varying meteorological conditions during the observations, the data were divided into 2 groups representing clear/calm/ and windy/cloudy conditions, respectively.

Preliminary results

Temperature along traverses for 12 days at 22:00 LT. Note that lines connecting points are not meant to indicate temporal progression rather they are used to better demarcate individual land-use types. The commercial district (red)is generally warmest, followed by the CBD (blue) and HDB housing estates. The lowest temperature is always associated with the rural, open area (green).

Temperature differences between a range of land-use types and the rural reference. The largest differences are associated with the commercial district (red). The resulting urban heat island intensity is between 1.5 and 4.5 degC where larger values are associated with wet conditions (i.e. rain during the preceding afternoon).

Temporal development of rural and urban cooling rates and associate heat island intensity for two cities under clear/calm conditions. The rural cooling rate for Singapore (green) is much smaller than the one for Vancouver (light blue) which results in a smaller heat island intensity for the former (red). V- Vancouver, S - Singapore, UCR - urban cooling rate, RCR - rural cooling rate, UHI- urban heat island intensity.

Urban moisture excess (UME) for 2 weather categories. A small UME is established only well after midnight.