The temporal variability of the canopy-level urban heat island (UHI) of Singapore is examined for different temporal scales on the basis of observations during a 1-year period. Temperature data obtained from different urban areas (commercial, Central Business District (CBD), high-rise and low-rise housing) are compared with 'rural' reference data and analysed with respect to meteorological variables and differences in land use. The results indicate that the peak UHI magnitude occurs 3-4 h (>6 h) after sunset in the commercial area, (at other urban sites). Higher UHI intensities generally occur during the southwest monsoon period of May-August, with a maximum of ∼7°C observed in the commercial area under ideal meteorological conditions. Variations in seasonal precipitation explain some of the differences in urban-rural cooling. No clear relationship between urban geometry and UHI intensity can be seen, and intra-urban variations of temperature are also shown to be influenced by other site factors, e.g. the extent of green space and anthropogenic heat. Lastly, results from the present study are compared with UHI data from other tropical and mid-latitude cities.

No abstract

No abstract

No abstract

ABSTRACT:Observations of local-scale urban surface energy balance (SEB), which include fluxes of net all-wave radiation (Q*), and eddy covariance measurements of sensible (Q H ) and latent heat (Q E ) were collected in an arid Phoenix, AZ suburb from January to December 2012. We studied diurnal variations in SEB partitioning over four distinct seasons: winter, equinoxes, and summer; the latter period is further subdivided into (1) months prior to and (2) months occurring during the North American Monsoon. Largest flux densities were observed in summer, with most available energy partitioned into Q H . Much less energy is partitioned into Q E , but this term is strongly affected by monsoonal precipitation, where greater-than-average Q E can be discerned for several days after storm events. The presence of a positive daily flux residual (RES) [i.e. Q* − (Q H + Q E )] for most of the summer indicates that anthropogenic heat (Q F ) from residential cooling is likely a significant factor influencing SEB. Analysis of hourly ensemble SEB fluxes during all seasons also indicates that RES is largest in the morning, but Q H dominates in the afternoon. Results of SEB trends and magnitudes from Phoenix were also compared with other urban sites, especially in (sub)tropical cities. When normalized with net radiation terms, a consistent diurnal hysteresis between ensemble Q H and RES occurs, suggesting a robust parameterization of this relationship for model development during clear-sky conditions. SEB dynamics also appear to be affected by local surface characteristics, with regular nocturnal negative Q H associated with a high urban sky-view factor. Measured Q E fluxes during dry seasons were larger than expected based on the small proportion of irrigated plan area vegetated surfaces. A probable explanation could be an enhanced micro-scale advective forcing of evapotranspiration arising from leading-edge effects over patchy residential lawns, which has possible implications for modelling evapotranspiration in hot arid cities.