Search

Interpreting the similarities and dissimilarities in spatiotemporal variations and various controls between surface and canopy urban heat islands (UHIs) is critical for a better understanding of their vertical structure. Preceding comparisons of the surface UHI (SUHI) and canopy UHI (CUHI), however, remain mostly restricted either in a single city or over a few cities within limited background climates; therefore, the associated similarities and dissimilarities between the SUHI and CUHI under different climates, especially at a global scale, remain largely unknown. Based on both satellite and in situ data, we simultaneously investigated the spatiotemporal patterns of the SUHI intensity (SUHII) and CUHI intensity (SUHII) of 366 global cities within various background climates. We further investigated the different impacts of several controls (e.g., vegetation coverage, population size, precipitation) on SUHII and CUHII. Our results indicate the following: (1) For the selected 366 cities, the annual mean SUHII is higher than CUHII by 1.1 ± 1.9 °C (mean ± Std) during the day and 0.3 ± 1.5 °C (mean ± Std) at night. The SUHII and CUHII in the equatorial, warm temperate, and snow climates are generally consistent with the above characteristics (i.e., SUHII > CUHII), however, in arid regions SUHII is lower than CUHII by 0.8 °C during the day. (2) The annual mean day–night difference in SUHII is positive (i.e., 0.6 ± 1.8 °C (mean ± Std)), while the difference in CUHII becomes negative (i.e., −0.2 ± 1.6 °C (mean ± Std)), indicating that urbanization increases the diurnal temperature range (DTR) based on land surface temperature, but it decreases the DTR based on surface air temperature. (3) Despite the high correlation between vegetation coverage and impervious surface percentage (ISP), their impacts on SUHII and CUHII were not consistent. The urban–rural difference in ISP exerts an insignificant impact on both SUHII and CUHII during the day and a greater impact on CUHII than on SUHII at night, whereas the urban–rural difference in vegetation coverage has a greater impact on SUHII than on CUHII during the day, while the opposite occurs at night. The impacts of population size on SUHII and CUHII are much greater during the night than on the day in which their impacts can be minimal. The relationship between annual mean precipitation and SUHII is positive during the day but negative at night, while for CUHII, their relationship is insignificantly negative both during the day and at night. These results can improve our understanding of the spatiotemporal patterns and controls of these two types of UHIs under various climates.

Estimating future temporal patterns of Surface Urban Heat Islands (SUHIs) on multiple time scales is an ongoing research endeavor. Among these time scales, estimation of next-day SUHIs is of special significance to urban residents, yet we currently lack a simple but efficient approach for making such estimations. In the present study, we propose a statistical strategy for estimating next-day nighttime SUHIs, based on incorporating various SUHI controls into a support vector machine regression (SVR) model. The majority of both the surface controls (including factors related to land cover and solar radiation) and meteorological controls (including temperature fluctuations, relative humidity, accumulated precipitation, wind speed, aerosol optical depth, and soil moisture) that have previously been found to account for daily SUHI variations were used as estimators, and we provide estimations for both the overall SUHI intensity (SUHII) and pixel-by-pixel Gaussian-based LSTs over 59 Chinese megacities. For the overall SUHII, the mean absolute error (MAE) is 0.67 K on average, and the mean absolute percentage error (MAPE) is no more than 25% for more than 90% of the cities. For the pixel-by-pixel LSTs, the associated MAE is less than 2.0 K in most scenarios. In addition, the contribution from each selected estimator to SUHII estimation is assessed comprehensively. Among all the estimators, the contribution from relative humidity is the greatest, followed by rural surface temperature and surface air temperature. Moreover, for nearly 78% of the cities, the estimators related to day-to-day SUHI variations make a larger contribution than those related to intra-annual SUHI variations. We conclude that our simple yet effective statistical approach for estimating next-day SUHIs can potentially help urban residents to better adapt to urban heat stress.

Understanding the diurnal dynamics of surface urban heat islands (SUHIs) is an indispensable step towards their full interpretation at multiple time scales. However, because of the tradeoff between the spatial and temporal resolutions of satellite-derived land surface temperature (LST) data, the climatology, variety, and taxonomy of diurnal SUHI (DSUHI) patterns remain largely unknown for numerous cities with different bioclimates. By combining daily MODIS LST data with a newly developed four-parameter diurnal temperature cycle (DTC) model, we selected 354 Chinese megacities located in different bioclimatic zones to examine the characteristics of the DSUHI descriptors and systematically investigate the prevalent DSUHI temporal patterns. The DSUHI variations demonstrate that both the daily maximum and minimum SUHI intensity (SUHII) can occur during most periods of the day, although these intensities are more likely to occur in the early morning and noon/afternoon. Our results also reveal that both strong SUHIs (SUHII > 3 K) and surface urban cool islands (SUCIs) (SUHII

Remotely sensed and accurate daily mean land surface temperature (Tdm) is valuable for various applications such as air temperature estimation and climate change monitoring. However, most traditional methods employed by the remote sensing community estimate Tdm by averaging the – usually few – observed cloud-free land surface temperatures (LSTs). Such estimates can have large sampling bias, especially for tandem polar orbiters, due to their sparse sampling of diurnal LST dynamics and the unavailability of under-cloud LSTs. To estimate accurate Tdm based on thermal observations from tandem polar orbiters, here we propose a simple yet robust framework that combines the annual temperature cycle (ATC) and the diurnal temperature cycle (DTC) models (termed the ADTC-based framework). The ATC model is used to reconstruct daily instantaneous under-cloud LSTs, based on which the DTC model is employed to establish diurnally continuous LST dynamics for estimating Tdm. The proposed framework is validated with geostationary LST observations and in-situ thermal measurements under both cloud-free and overcast conditions. The validations show that, under cloud-free conditions, the ADTC-based framework is able to reduce the positive sampling bias obtained with simple averaging (> 2.0 K) and yields a mean absolute error (MAE) of approximately 0.5 K. Under overcast conditions, the ADTC-based framework yields MAEs of 1.0 K and 0.5 K at the daily and monthly scales, respectively. Furthermore, a contribution analysis indicates that the ATC model reduces the MAE from around 4.2 K to 2.0 K while the DTC model reduces the MAE from around 2.0 K to 1.0 K. Based on our validation results and tests performed with MODIS data, the presented simple yet robust ADTC-based framework is able to accurately estimate large-scale spatiotemporally continuous Tdm from thermal observations of tandem polar orbiters. Therefore, the ADTC-based framework is a potentially valuable tool for many related applications.

Modelling of the Surface Urban Heat Island (SUHI) temporal variations have been a great concern. However, most previous studies only focused on modelling the SUHI variations in the past period, yet those for their future patterns remain rarely investigated. By incorporating various predictable meteorological variables in the SVM regression model, this study achieved an attempt to the prediction for the next-day SUHIs over Chinese main cities. Both the SUHI intensity (SUHII) and the pixel-based Gaussian-simulated LSTs were predicted. The averaged MAE of our predicted SUHII across Chinese megacities is 0.67 K; and the MAE for the LST is generally less than 1.5 K. The incorporation of meteorological variables was shown to greatly contribute to the predicted daily SUHIIs. We consider our study, by achieving an attempt to the SUHI prediction, can improve the understanding of the SUHI mitigation.

Separate impacts of building and tree on the urban thermal environment have been studied extensively, but their combined impacts, especially from both the horizontal (i.e., two-dimensional (2D)) and vertical (i.e., three-dimensional (3D)) perspectives remain largely unclear. Based on satellite thermal data and elaborate 2D and 3D urban morphology, herein we simultaneously investigate the separate and combined impacts of building and tree over Nanjing in China from both the 2D and 3D perspectives. We further examine the day–night contrast together with the sensitivity of such impacts to scale. Our results show that, when compared with urban structures from a single dimension, the combination of 2D and 3D structures is more capable of predicting urban land surface temperatures (LSTs) for both day and night. The assessments further illustrate that the separate and combined impacts of building and tree on LSTs are usually more significant when the spatial scale increases. As for the separate impacts of building and tree, 2D structure affects more urban thermal environment than 3D structure at all spatial scales during the day, but an opposite trend occurs at night. Moreover, for the combined impact of building and tree on LST across different scales, daytime and nighttime LSTs are respectively dominated by 2D and 3D building structures. Combining 2D and 3D structures improves the explained LST variation by 7.3%–11.1% and 25.3%–37.7% for day and night, respectively when compared to using 2D structures only. These findings emphasize the need to incorporate both 2D and 3D urban morphology to improve the urban thermal environment.

Most previous studies of surface urban heat islands (SUHIs) have focused solely on their controlling factors on a seasonal/annual timescale, while the controls on daily variations are largely unknown. By extracting the daily variations of nighttime SUHI features using the Gaussian model and investigating their correlations with various explanatory factors, we have attempted to determine the controls on SUHIs on a daily-basis over Chinese cities. Specific controls of weather conditions on the intensity, extent, shape, and centroid of the SUHIs were identified. Our results show that: (1) SUHI intensity (SUHII) was considerably more sensitive to weather conditions than the SUHI footprint (i.e., extent, shape, and centroid). (2) Meteorological variables including relative humidity, accumulated precipitation, and aerosol optical depth, had the greatest impact on SUHI intensity; whereas factors related to temperature fluctuations (day-to-day fluctuations of surface and air temperature) were the main factors influencing SUHI extent, shape, and the direction in which SUHI centroid varies. (3) Antecedent precipitation substantially impacted the subsequent SUHIs under clear-skies, changing both the SUHI itself and its sensitivity to other factors. Typically, the clear-sky SUHIs directly following rainfall showed a higher dependence on the relative humidity, soil moisture and aerosol, but were less affected by wind. (4) The meteorological contributions to the daily nighttime SUHIIs varied among Chinese cities with different bioclimatic conditions. In general, they were stronger in temperate zones than in subtropical zones. Our results provide an improved understanding of the controls on SUHIs on a daily timescale, as well as a foundation for predicting daily SUHIs based on the influencing meteorological variables.

The urban heat island (UHI) is a major topic in the study of urban climates. However, comprehensive research on the atmospheric UHI (UHIAtm), surface UHI (UHISurf), and subsurface UHI (UHISub) simultaneously has yet not been reported. Using the MODIS land surface temperature and atmospheric profile data during the 2010–2016 period, we investigated the diurnal, seasonal, and vertical variations of UHIAtm, UHISurf, and UHISub in Beijing. The major findings include but are not limited to the following: As the altitude increases from 1000 hPa (near the surface) to 700 hPa (~3 km), the daytime UHIAtm intensity in summer decreases piecewise linearly with a rapid decline above 850 hPa (~1.5 km), while in spring and autumn it decreases after a slight increase in the lower atmosphere. The nighttime UHIAtm intensity decreases approximately linearly in all seasons with a rapid (gradual) decline in winter (summer). As the depth increases from the surface to ~0.2 m, the daytime intensity from UHISurf to UHISub decreases in summer but increases in the other seasons, while the nighttime trends are opposite to the daytime ones. The diurnal (seasonal) variation of UHISub intensity converges to the daily (annual) mean as the depth reaches ~0.5 m (10 m). These new findings have theoretical and practical implications for in-depth understanding of the urban thermal environment from the boundary layer to the subsurface.

Satellite thermal remote sensing provides land surface temperatures (LST) over extensive areas that are vital in various applications, but this technique suffers from its sampling style and the impenetrability of clouds, which frequently generates data gaps. Annual temperature cycle (ATC) models can fill these gaps and estimate continuous daily LST dynamics from a number of thermal observations. However, the standard ATC model (termed ATCS) remains incapable of quantifying the short-term LST variations caused by synoptic conditions. By incorporating in-situ surface air temperatures (SATs) and satellite-derived normalized difference vegetation indexes (NDVIs), here we proposed an enhanced ATC model (ATCE) to describe the daily LST fluctuations. With Aqua/MODIS LST products as validation data, we implemented and tested the ATCE over the Yangtze River Delta region of China. The results demonstrate that, when compared with the ATCS, the overall root mean square errors of the ATCE decrease by 1.0 and 0.8 K for the day and night, respectively. The accuracy improvements vary with land cover types with greater improvements over the forest, grassland, and built-up areas than over cropland and wetland. The assessments at different time scales further confirm that LST fluctuations can be better described by the ATCE. Though with limitations, we consider this new model and its associated parameters hold great potentials in various applications.