Three-Dimensional Urban Thermal Effect across a Large City Cluster during an Extreme Heat Wave: Observational Analysis

极端热浪背景下超大城市群的三维热效应:观测分析

+ Author Affiliations + Find other works by these authors
  • Corresponding author: Ping LIANG, liangping1107@163.com
  • Funds:

    Supported by the Guangdong Major Project of Basic and Applied Basic Research (2020B0301030004), National Natural Science Foundation of China (42175056, 41790471), Natural Science Foundation of Shanghai (21ZR1457600), China Meteorological Administration Innovation and Development Project (CXFZ2022J009), and UK–China Research and Innovation Partnership Fund through the Met Office Climate Science for Service Partnership (CSSP) China as part of the Newton Fund

  • doi: 10.1007/s13351-022-1171-x

PDF

  • Given extensive and rapid urbanization globally, assessing regional urban thermal effects (UTE) in both canopy and boundary layers under extreme weather/climate conditions is of significant interest. Rapid population and economic growth in the Yangtze River Delta (YRD) have made it one of the largest city clusters in China. Here, we explore the three-dimensional (3D) UTE in the YRD using multi-source observations from high-resolution automatic weather stations, radiosondes, and eddy covariance sensors during the record-setting heat wave (HW) of July–August 2013. It is found that the regional canopy layer UTE is up to 0.6–1.2°C, and the nocturnal UTE (0.7–1.6°C) is larger than daytime UTE (0.2–0.5°C) during the HW. The regional canopy layer UTE is enhanced and expanded northwards, with some rural sites contaminated by the urban influences, especially at night. In the boundary layer, the strengthened regional UTE extends vertically to at least 925 hPa (~750 m) during this HW. The strengthened 3D UTE in the YRD is associated with an enlarged Bowen ratio difference between urban and non-urban areas. These findings about the 3D UTE are beneficial for better understanding of the thermal environment of large city clusters under HW and for more appropriate adaption and mitigation strategies.
    随着全球城市化水平的快速提升,极端热浪背景下冠层和边界层的区域性城市热效应(urban thermal effects,UTE)备受关注。长江三角洲(以下简称长三角)凭借其快速的经济人口增长已成为中国最大的城市群之一。本研究利用高密度的区域地面自动气象站、探空、涡动通量等多源观测资料,探讨了2013年7–8月破纪录的高温热浪期间长三角的三维UTE。在高温热浪期间,长三角冠层UTE高达0.6–1.2oC,且夜间UTE(0.7–1.6oC)高于白天(0.2–0.5oC)。由于周边乡村地区受到城市影响,长三角冠层的UTE在热浪期间存在加强并向北延伸的现象,在夜间尤为明显。在边界层,高温热浪期间增强的UTE可向上延伸至925 hPa(约750米)。三维UTE的增强与城市和非城市间波文比差异的扩大有关。上述关于三维UTE的研究结果有助于理解高温热浪背景下的城市群热环境以及评估适应减缓策略。
  • 加载中
  • Fig. 1.  Location of observation sites and distribution of impervious fraction (shaded) in the Yangtze River Delta (YRD) region. The large cities in the “Z-shaped” city cluster include: Nanjing (NJ), Yangzhou (YZ), Zhenjiang (ZJ), Changzhou (CZ), Wuxi (WX), Suzhou (SZ), Shanghai (SH), Hangzhou (HZ), Cixi (CX), and Ningbo (NB). The automatic weather stations (AWS, circles), national weather stations (NWS, triangles), and radiosonde sites (squares) are classified into urban (solid) and non-urban (hollow) types. The energy balance observation sites (XJH, SWEP) are in Shanghai (inset). Section 2.2 provides more details.

    Fig. 2.  Variations of urban characteristics through time for YRD: (a) population density (data from China Statistical Bureau 1983–2013) and (b) impervious fraction > 0.5 (impervious cover includes building roofs, driveways, and parking lots) in different time periods of first occurrence (color shaded; data from Gong et al., 2019a, b).

    Fig. 3.  Number of days in July–August 2013 in the Yangtze River Delta (YRD) for (a) daily Tmax,2m ≥ 35°C and (b) both daily Tmax,2m ≥ 35°C and Tmin,2m ≥ 25°C, based on the data from the automatic weather stations (AWS).

    Fig. 4.  Frequency distributions (0.5°C bins; %) during July–August 2013 of (a, d) Tmean,2m (°C), (b, e) Tmin,2m (°C), and (c, f) Tmax,2m (°C) at all urban (U, red) and non-urban (nU, blue) automatic weather station (AWS) sites over the (a–c) entire period and the period of (d–f) heat wave (HW, 34 days) and non-HW (nHW, 28 days) conditions. Values given within each plot are the medians.

    Fig. 5.  Temopral evolution of daily regional averaged canopy layer urban thermal effects (UTE) [Iav,2A; °C] for (a) Tmean,2m (°C), (b) Tmin,2m (°C), and (c) Tmax,2m (°C) temperature derived from the automatic weather station (AWS) data, using three areal extents [MSA radii (km): 200 (blue), 250 (yellow), and 300 (red)] during July–August 2013 in the YRD. For methods see Section 2.3.

    Fig. 6.  Interannual variation of the average urban thermal effect (UTE) [Iav,mean,2N; °C, black] and mean temperature (Tmean,2m; °C, orange) derived from the national weather station (NWS; Fig. 1) data during July−August of 1980–2017 in the YRD.

    Fig. 7.  Spatial distributions of the average urban thermal effect (UTE) [Ii,mean,2A; °C] calculated from the automatic weather station (AWS) data using a 200-km radius during (a–c) non-HW (nHW; 28 days) and (d–f) heat wave (HW; 34 days) conditions for (a, d) mean temperature (Ti,mean,2A; °C), (b, e) minimum temperature (Ti,min,2A; °C), and (c, f) maximum temperature (Ti,max,2A; °C) in the YRD.

    Fig. 8.  Diurnal differences (Δ = HW − nHW) in median surface energy balance fluxes: (a) net all-wave radiation (Q*; W m−2), (b) turbulent sensible heat (QH; W m−2), (c) turbulent latent heat (QE; W m−2), and (d) Bowen ratio (QH/QE) at an urban (XJH, U; red) and a grass (SWEP, nU; blue) site. For methods see Section 2.2. Note that the number of days for heat wave (HW) is 34 while that for non-HW (nHW) is 28.

    Fig. 9.  Vertical profiles of air temperature (°C) and urban thermal effect (UTE) in the Yangtze River Delta (YRD) during July–August 2013 under heat wave (HW, 34 days, solid) and non-HW (nHW, 28 days, dashed) periods from radiosonde (R, black) and ERA5 (E, gray) data for (a, d) urban (U), (b, e) non-urban (nU), and (c, f) U−nU difference [Iav,mean, °C] at (a–c) 0800 and (d–f) 2000 BT. For methods see Section 2.3.

    Table 1.  Notation used to distinguish the different aspects of the urban thermal effects (UTE)

    Height (m AGL)Data sourceMean (Tmean)Minimum (Tmin)Maximum (Tmax)
    ~2AWS–MSAImean,2AImin,2AImax,2A
    ~2NWSImean,2N
    ~50–100, 750, 1500, 3000RadiosondeImean,R
    ERA5Imean,E
    Note: MSA denotes moving spatial anomaly.
    Download: Download as CSV

    Table 2.  Analysis of the frequency curves (Fig. 6) of daily Tmean,2m (°C), Tmin,2m (°C), and Tmax,2m (°C) from AWS under HW and non-HW conditions for urban (U) and non-urban (nU) sites in the YRD (Fig. 1) during July–August 2013. For statistical methods see Section 2.3

    Tmean,2m (°C)Tmin,2m (°C)Tmax,2m (°C)
    UnUUnUUnU
    July–AugustSkewness−0.59−0.36−0.91−0.37−0.43−0.29
    Kurtosis0.490.102.470.38−0.10−0.11
    IQR3.13.42.63.34.24.8
    Median30.929.227.225.236.035.4
    HWSkewness−0.63−0.35−0.66−0.19−0.43−0.37
    Kurtosis0.34−0.010.39−0.170.490.18
    IQR2.83.22.63.43.64.2
    Median31.629.827.525.436.836.3
    non-HWSkewness−0.78−0.50−1.37−0.78−0.23−0.10
    Kurtosis0.670.232.941.08−0.59−0.29
    IQR2.93.02.72.94.24.7
    Median29.227.526.024.533.332.5
    Download: Download as CSV
  • [1]

    Ao, X. Y., C. S. B. Grimmond, D. W. Liu, et al., 2016: Radiation fluxes in a business district of Shanghai, China. J. Appl. Meteor. Climatol., 55, 2451–2468. doi: 10.1175/JAMC-D-16-0082.1.
    [2]

    Ao, X. Y., C. S. B. Grimmond, H. C. Ward, et al., 2018: Evaluation of the Surface Urban Energy and Water balance Scheme (SUEWS) at a dense urban site in Shanghai: Sensitivity to anthropogenic heat and irrigation. J. Hydrometeor., 19, 1983–2005. doi: 10.1175/JHM-D-18-0057.1.
    [3]

    Ao, X. Y., J. G. Tan, X. Zhi, et al., 2019: Synergistic interaction between urban heat island and heat waves and its impact factors in Shanghai. Acta Geographica Sinica, 74, 1789–1802. doi: 10.11821/dlxb201909007. (in Chinese)
    [4]

    Arnfield, A. J., 2003: Two decades of urban climate research: A review of turbulence, exchanges of energy and water, and the urban heat island. Int. J. Climatol., 23, 1–26. doi: 10.1002/joc.859.
    [5]

    Basara, J. B., P. K. Hall, Jr., A. J. Schroeder, et al., 2008: Diurnal cycle of the Oklahoma City urban heat island. J. Geophys. Res. Atmos., 113, D20109. doi: 10.1029/2008jd010311.
    [6]

    Chen, F., X. C. Yang, and W. P. Zhu, 2014: WRF simulations of urban heat island under hot-weather synoptic conditions: The case study of Hangzhou city, China. Atmos. Res., 138, 364–377. doi: 10.1016/j.atmosres.2013.12.005.
    [7]

    CMA, 2018: Technical Specifications for Maintenance of Regio-nal Automatic Weather Stations. QX/T 465–2018. . Available at http://cmastd.cmatc.cn/standardView.jspx?id=3076. Accessed on 5 May 2022. (in Chinese)
    [8]

    Copernicus Climate Change Service, 2018: ERA5 hourly data on pressure levels from 1979 to present. doi: 10.24381/cds.bd0915c6.
    [9]

    Cui, L. L., and J. Shi, 2012: Urbanization and its environmental effects in Shanghai, China. Urban Climate, 2, 1–15. doi: 10.1016/j.uclim.2012.10.008.
    [10]

    Dairaku, K., K. Kuraji, M. Suzuki, et al., 2000: The effect of rainfall duration and intensity on orographic rainfall enhancement in a mountainous area. J. Japan Soc. Hydrol. Water Resour., 13, 57–68. doi: 10.3178/jjshwr.13.57.
    [11]

    Du, Y., Z. Q. Xie, Y. Zeng, et al., 2007: Impact of urban expansion on regional temperature change in the Yangtze River Delta. J. Geogr. Sci., 17, 387–398. doi: 10.1007/s11442-007-0387-0.
    [12]

    Estévez, J., P. Gavilán, and J. V. Giráldez, 2011: Guidelines on validation procedures for meteorological data from automatic weather stations. J. Hydrology, 402, 144–154, doi: 10.1016/j.jhydrol.2011.02.031.
    [13]

    Fischer, E. M., S. I. Seneviratne, P. L. Vidale, et al, 2007: Soil moisture–atmosphere interactions during the 2003 European summer heat wave. J. Climate, 20, 5081–5099. doi: /10.1175/JCLI4288.1.
    [14]

    Gong, P., X. C. Li, and W. Zhang, 2019a: 40-year (1978–2017) human settlement changes in China reflected by impervious surfaces from satellite remote sensing. Sci. Bull., 64, 756–763. doi: 10.1016/j.scib.2019.04.024.
    [15]

    Gong, P., H. Liu, M. N. Zhang, et al., 2019b: Stable classification with limited sample: Transferring a 30-m resolution sample set collected in 2015 to mapping 10-m resolution global land cover in 2017. Sci. Bull., 64, 370–373. doi: 10.1016/j.scib.2019.03.002.
    [16]

    Grimmond, C. S. B., and T. R. Oke, 1999: Heat storage in urban areas: Local-scale observations and evaluation of a simple model. J. Appl. Meteor., 38, 922–940. doi: 10.1175/1520-0450(1999)038<0922:HSIUAL>2.0.CO;2.
    [17]

    Grimmond, C. S. B., H. A. Cleugh, and T. R. Oke, 1991: An objective urban heat storage model and its comparison with other schemes. Atmos. Environ., 25, 311–326. doi: 10.1016/0957-1272(91)90003-W.
    [18]

    Grimmond, C. S. B., T. R. Oke, and H. A. Cleugh, 1993: The role of “rural” in comparisons of observed suburban–rural flux differences. Exchange processes at the land surface of a range of space and time scales. Proceedings International Symposium, Yokohama, Japan. IAHS Publ., 212, 165–174.
    [19]

    Guo, Y. J., and Y. H. Ding, 2009: Long-term free-atmosphere temperature trends in China derived from homogenized in-situ radiosonde temperature series. J. Climate, 22, 1037–1051. doi: 10.1175/2008JCLI2480.1.
    [20]

    Guo, Y. J., P. W. Thorne, M. P. McCarthy, et al., 2008: Radiosonde temperature trends and their uncertainties over eastern China. Int. J. Climatol., 28, 1269–1281. doi: 10.1002/joc.1633.
    [21]

    Han, Y., Y. W. Zhou, J. P. Guo, et al., 2019: The characteristics of spatial and temporal variations in the PBL during the landfall of tropical cyclones across East China. J. Appl. Meteor. Climatol., 58, 1557–1572. doi: 10.1175/JAMC-D-18-0131.1.
    [22]

    Hausfather, Z., M. J. Menne, C. N. Williams, et al., 2013: Quantifying the effect of urbanization on U.S. Historical Climatology Network temperature records. J. Geophys. Res. Atmos., 118, 481–494. doi: 10.1029/2012JD018509.
    [23]

    Huang, B., G. H. Ni, and C. S. B. Grimmond, 2019: Impacts of urban expansion on relatively smaller surrounding cities during heat waves. Atmosphere, 10, 364. doi: 10.3390/atmos10070364.
    [24]

    Huang, Z., H. Chen, and H. Tian, 2011: Research on the heat wave index. Meteor. Mon., 37, 345–351. (in Chinese)
    [25]

    Ichinose, T., K. Shimodozono, and K. Hanaki, 1999: Impact of anthropogenic heat on urban climate in Tokyo. Atmos. Environ., 33, 3897–3909. doi: 10.1016/s1352-2310(99)00132-6.
    [26]

    Jiang, S. J., X. Lee, J. K. Wang, et al., 2019: Amplified urban heat islands during heat wave periods. J. Geophys. Res. Atmos., 124, 7797–7812. doi: 10.1029/2018JD030230.
    [27]

    Jiang, S. J., K. C. Wang, and Y. N. Mao, 2020: Rapid local urbanization around most meteorological stations explains the observed daily asymmetric warming rates across China from 1985 to 2017. J. Climate, 33, 9045–9061. doi: 10.1175/JCLI-D-20-0118.1.
    [28]

    Jiang, X. L., D. H. Wang, J. J. Xu, et al., 2017: Characteristics of observed tropopause height derived from L-band sounder over the Tibetan Plateau and surrounding areas. Asia-Pacific J. Atmos. Sci., 53, 1–10. doi: 10.1007/s13143-016-0035-7.
    [29]

    Lau, N. C., and M. J. Nath, 2012: A model study of heat waves over North America: Meteorological aspects and projections for the twenty-first century. J. Climate, 25, 4761–4784. doi: 10.1175/JCLI-D-11-00575.1.
    [30]

    Li, D., T. Sun, M. F. Liu, et al., 2015: Contrasting responses of urban and rural surface energy budgets to heat waves explain synergies between urban heat islands and heat waves. Environ. Res. Lett., 10, 054009. doi: 10.1088/1748-9326/10/5/054009.
    [31]

    Liao, J. B., T. J. Wang, Z. Q. Jiang, et al., 2015: WRF/Chem modeling of the impacts of urban expansion on regional climate and air pollutants in Yangtze River Delta, China. Atmos. Environ., 106, 204–214. doi: 10.1016/j.atmosenv.2015.01.059.
    [32]

    Loridan, T., and C. S. B. Grimmond, 2012: Characterization of energy flux partitioning in urban environments: Links with surface seasonal properties. J. Appl. Meteor. Climatol., 51, 219–241. doi: 10.1175/JAMC-D-11-038.1.
    [33]

    Lowry, W. P., 1977: Empirical estimation of urban effects on climate: A problem analysis. J. Appl. Meteor. Climatol., 16, 129–135. doi: 10.1175/1520-0450(1977)016<0129:EEOUEO>2.0.CO;2.
    [34]

    Lu, H. M., M. L. Zhang, W. W. Sun, et al., 2018: Expansion analysis of Yangtze River Delta urban agglomeration using DMSP/OLS nighttime light imagery for 1993 to 2012. ISPRS Int. J. Geo-Inf., 7, 52. doi: 10.3390/ijgi7020052.
    [35]

    Ma, S. M., T. J. Zhou, D. A. Stone, et al., 2017: Attribution of the July–August 2013 heat event in Central and Eastern China to anthropogenic greenhouse gas emissions. Environ. Res. Lett., 12, 054020. doi: 10.1088/1748-9326/aa69d2.
    [36]

    Mardia, K. V., 1970: Measures of multivariate skewness and kurtosis with applications. Biometrika, 57, 519–530. doi: 10.1093/biomet/57.3.519.
    [37]

    Miao, S. G., W. M. Jiang, P. Liang, et al., 2020: Advances in urban meteorological research in China. J. Meteor. Res., 34, 218–242. doi: 10.1007/s13351-020-9858-3.
    [38]

    National Bureau of Statistics of China, 2014: China Statistical Yearbook: 2014. China Statistics Press, Beijing, 23–45. (in Chinese)
    [39]

    National Bureau of Statistics of China, 2018: China Statistical Yearbook: 2018. China Statistics Press, Beijing, 29–51. (in Chinese)
    [40]

    Oke, T. R., 1976: The distinction between canopy and boundary-layer urban heat islands. Atmosphere, 14, 268–277. doi: 10.1080/00046973.1976.9648422.
    [41]

    Oke, T. R., 1982: The energetic basis of the urban heat island. Quart. J. Roy. Meteor. Soc., 108, 1–24. doi: 10.1002/qj.49710845502.
    [42]

    Oke, T. R., G. Mills, A. Christen, et al., 2017: Urban Climates. Cambridge University Press, Cambridge, 546 pp.
    [43]

    Peng, J. B., 2014: An investigation of the formation of the heat wave in southern china in summer 2013 and the relevant abnormal subtropical high activities. Atmos. Ocean. Sci. Lett., 7, 286–290. doi: 10.3878/j.issn.1674-2834.13.0097.
    [44]

    Piringer, M., C. S. B. Grimmond, S. M. Joffre, et al., 2002: Investigating the surface energy balance in urban areas—recent advances and future needs. Water Air Soil Poll. Focus, 2, 1–16. doi: 10.1023/A:1021302824331.
    [45]

    Ren, G. Y., Z. Y. Chu, Y. Q. Zhou, et al., 2005: Recent progresses in studies of regional temperature changes in China. Climatic Environ. Res., 10, 701–716. doi: 10.3878/j.issn.1006-9585.2005.04.01. (in Chinese)
    [46]

    Ren, Y. Y., and G. Y. Ren, 2011: A remote-sensing method of selecting reference stations for evaluating urbanization effect on surface air temperature trends. J. Climate, 24, 3179–3189. doi: 10.1175/2010JCLI3658.1.
    [47]

    Schatz, J., and C. J. Kucharik, 2015: Urban climate effects on extreme temperatures in Madison, Wisconsin, USA. Environ. Res. Lett., 10, 094024. doi: 10.1088/1748-9326/10/9/094024.
    [48]

    Schluenzen, K. H., C. S. B. Grimmond, and B. Alexande, 2021: Guidance to measuring, modelling and monitoring the ca-nopy layer Urban Heat Island. EMS Annual Meeting Abstracts, Vol. 18, EMS2021-301, doi: 10.5194/ems2021-301..
    [49]

    Shepherd, M., T. Andersen, C. Strother, et al., 2013: Urban Climate Archipelagos: A new Framework for urban impacts on Climate. Earthzine. Available online at https://earthzine.org/urban-climate-archipelagos-a-new-framework-for-urban-impacts-on-climate/. Accessed on 18 May 2022.
    [50]

    Stewart, I. D., 2011: A systematic review and scientific critique of methodology in modern urban heat island literature. Int. J. Climatol., 31, 200–217. doi: 10.1002/joc.2141.
    [51]

    Stewart, I. D., and T. R. Oke, 2012: Local climate zones for urban temperature studies. Bull. Amer. Meteor. Soc., 93, 1879–1900. doi: 10.1175/bams-d-11-00019.1.
    [52]

    Stone, B. Jr., 2012: The City and the Coming Climate: Climate Change in the Places We Live. Cambridge University Press, Cambridge, New York, 198 pp.
    [53]

    Sun, J. Q., 2014: Record-breaking SST over mid-North Atlantic and extreme high temperature over the Jianghuai–Jiangnan region of China in 2013. Chinese Sci. Bull., 59, 3465–3470. doi: 10.1007/s11434-014-0425-0.
    [54]

    Sun, T., S. Kotthaus, D. Li, et al., 2017: Attribution and mitigation of heat wave-induced urban heat storage change. Environ. Res. Lett., 12, 114007. doi: 10.1088/1748-9326/aa922a.
    [55]

    Tan, J. G., L. M. Yang, C. S. B. Grimmond, et al., 2015: Urban integrated meteorological observations: Practice and experience in Shanghai, China. Bull. Amer. Meteor. Soc., 96, 85–102. doi: 10.1175/BAMS-D-13-00216.1.
    [56]

    Tao, S. Y., and L. X. Chen, 1987: A Review of Recent Research on the East Asian Summer Monsoon in China. Oxford University Press, Oxford, 60–92.
    [57]

    United Nations, 2018: 2018 Revision of World Urbanization Prospects. United Nations Department of Economic and Social Affairs, 1–126. Available online at https://www.un.org/development/desa/publications/2018-revision-of-world-urbanization-prospects.html. Accessed on 12 May 2022.
    [58]

    Wang, J., Z. W. Yan, X. W. Quan, et al., 2017: Urban warming in the 2013 summer heat wave in eastern China. Climate Dyn., 48, 3015–3033. doi: 10.1007/s00382-016-3248-7.
    [59]

    Wang, J., Y. Chen, W. L. Liao, et al., 2021: Anthropogenic emissions and urbanization increase risk of compound hot extremes in cities. Nat. Clim. Chang, 11, 1084–1089. doi: 10.1038/s41558-021-01196-2.
    [60]

    Ward, H. C., and C. S. B. Grimmond, 2017: Assessing the impact of changes in surface cover, human behaviour and climate on energy partitioning across Greater London. Landsc. Urban Plan., 165, 142–161. doi: 10.1016/j.landurbplan.2017.04.001.
    [61]

    Wu, K., and X. Q. Yang, 2013: Urbanization and heterogeneous surface warming in eastern China. Chinese Sci. Bull., 58, 1363–1373. doi: 10.1007/s11434-012-5627-8.
    [62]

    Xia, J. J., K. Tu, Z. W. Yan, et al., 2016: The super-heat wave in eastern China during July–August 2013: A perspective of climate change. Int. J. Climatol., 36, 1291–1298. doi: 10.1002/joc.4424.
    [63]

    Yang, P., G. Y. Ren, and W. D. Liu, 2013: Spatial and temporal characteristics of Beijing urban heat island intensity. J. Appl. Meteor. Climatol., 52, 1803–1816. doi: 10.1175/JAMC-D-12-0125.1.
    [64]

    Zeng, Y., X. F. Qiu, L. H. Gu, et al., 2009: The urban heat island in Nanjing. Quatern. Int., 208, 38–43. doi: 10.1016/j.quaint.2009.02.018.
    [65]

    Zhai, P. M., and X. H. Pan, 2003: Trends in temperature extremes during 1951–1999 in China. Geophys. Res. Lett., 30, 1913. doi: 10.1029/2003gl018004.
    [66]

    Zhang, H. S., X. Y. Zhang, Q. H. Li, et al., 2020: Research progress on estimation of the atmospheric boundary layer height. J. Meteor. Res., 34, 482–498. doi: 10.1007/s13351-020-9910-3.
    [67]

    Zhang, K. X., R. Wang, C. C. Shen, et al., 2010: Temporal and spatial characteristics of the urban heat island during rapid urbanization in Shanghai, China. Environ. Monit. Assess., 169, 101–112. doi: 10.1007/s10661-009-1154-8.
    [68]

    Zhang, L. S., J. J. Xue, W. G. Wang, et al., 2014: Comparative analysis of extreme high temperature weather in the summers of 2013 and 2003. Atmos. Ocean. Sci. Lett., 7, 132–136. doi: 10.3878/j.issn.1674-2834.13.0073.
    [69]

    Zhang, N., Z. Q. Gao, X. M. Wang, et al., 2010: Modeling the impact of urbanization on the local and regional climate in Yangtze River Delta, China. Theor. Appl. Climatol., 102, 331–342. doi: 10.1007/s00704-010-0263-1.
    [70]

    Zhang, P. F., G. Y. Ren, Y. Qin, et al., 2021: Urbanization effects on estimates of global trends in mean and extreme air temperature. J. Climate, 34, 1923–1945. doi: 10.1175/JCLI-D-20-0389.1.
    [71]

    Zhong, S., Y. Qian, C. Zhao, et al., 2017: Urbanization-induced urban heat island and aerosol effects on climate extremes in the Yangtze River Delta region of China. Atmos. Chem. Phys., 17, 5439–5457. doi: 10.5194/acp-17-5439-2017.
    [72]

    Zittis, G., Hadjinicolaou, P., Fnais, M., et al., 2015: Projected changes in heat wave characteristics in the eastern Mediterranean and the Middle East. Reg. Environ. Change, 16, 1863–1876. doi: 10.1007/s10113-014-0753-2.
  • Ping LIANG and Yue MA.pdf

  • 加载中
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

Three-Dimensional Urban Thermal Effect across a Large City Cluster during an Extreme Heat Wave: Observational Analysis

    Corresponding author: Ping LIANG, liangping1107@163.com
  • 1. Key Laboratory of Cities’ Mitigation and Adaptation to Climate Change in Shanghai, Shanghai Regional Climate Center, China Meteorological Administration, Shanghai 200030, China
  • 2. Shanghai Jiading District Meteorological Bureau, Shanghai 201800, China
  • 3. Department of Meteorology, University of Reading, Reading RG6 6ET, United Kingdom
  • 4. Ocean College, Zhejiang University, Zhoushan 316021, China
  • 5. Jiangsu Meteorological Information Center, Jiangsu Meteorological Bureau, Nanjing 210008, China
  • 6. National Climate Center, China Meteorological Administration, Beijing 100081, China
Funds: Supported by the Guangdong Major Project of Basic and Applied Basic Research (2020B0301030004), National Natural Science Foundation of China (42175056, 41790471), Natural Science Foundation of Shanghai (21ZR1457600), China Meteorological Administration Innovation and Development Project (CXFZ2022J009), and UK–China Research and Innovation Partnership Fund through the Met Office Climate Science for Service Partnership (CSSP) China as part of the Newton Fund

Abstract: Given extensive and rapid urbanization globally, assessing regional urban thermal effects (UTE) in both canopy and boundary layers under extreme weather/climate conditions is of significant interest. Rapid population and economic growth in the Yangtze River Delta (YRD) have made it one of the largest city clusters in China. Here, we explore the three-dimensional (3D) UTE in the YRD using multi-source observations from high-resolution automatic weather stations, radiosondes, and eddy covariance sensors during the record-setting heat wave (HW) of July–August 2013. It is found that the regional canopy layer UTE is up to 0.6–1.2°C, and the nocturnal UTE (0.7–1.6°C) is larger than daytime UTE (0.2–0.5°C) during the HW. The regional canopy layer UTE is enhanced and expanded northwards, with some rural sites contaminated by the urban influences, especially at night. In the boundary layer, the strengthened regional UTE extends vertically to at least 925 hPa (~750 m) during this HW. The strengthened 3D UTE in the YRD is associated with an enlarged Bowen ratio difference between urban and non-urban areas. These findings about the 3D UTE are beneficial for better understanding of the thermal environment of large city clusters under HW and for more appropriate adaption and mitigation strategies.

极端热浪背景下超大城市群的三维热效应:观测分析

随着全球城市化水平的快速提升,极端热浪背景下冠层和边界层的区域性城市热效应(urban thermal effects,UTE)备受关注。长江三角洲(以下简称长三角)凭借其快速的经济人口增长已成为中国最大的城市群之一。本研究利用高密度的区域地面自动气象站、探空、涡动通量等多源观测资料,探讨了2013年7–8月破纪录的高温热浪期间长三角的三维UTE。在高温热浪期间,长三角冠层UTE高达0.6–1.2oC,且夜间UTE(0.7–1.6oC)高于白天(0.2–0.5oC)。由于周边乡村地区受到城市影响,长三角冠层的UTE在热浪期间存在加强并向北延伸的现象,在夜间尤为明显。在边界层,高温热浪期间增强的UTE可向上延伸至925 hPa(约750米)。三维UTE的增强与城市和非城市间波文比差异的扩大有关。上述关于三维UTE的研究结果有助于理解高温热浪背景下的城市群热环境以及评估适应减缓策略。
    • Human activities alter the material properties and structures of cities worldwide, with direct impacts on the surface energy balance and the overlying boundary layer. Urban areas are often warmer than their surroundings in response to these modifications. With the large and ever-increasing urban population (United Nations, 2018), the urban thermal environment is of great importance to many sectors of society (e.g., public health, traffic, power industry, and water resource management). The east coast of the Yangtze River Delta (YRD; Fig. 1) is now one of the largest economic zones in China, following recent decades of rapid economic development, with a densely populated cluster of cities in this region in a distinctive Z-shape (Fig. 1; Du et al., 2007).

      Figure 1.  Location of observation sites and distribution of impervious fraction (shaded) in the Yangtze River Delta (YRD) region. The large cities in the “Z-shaped” city cluster include: Nanjing (NJ), Yangzhou (YZ), Zhenjiang (ZJ), Changzhou (CZ), Wuxi (WX), Suzhou (SZ), Shanghai (SH), Hangzhou (HZ), Cixi (CX), and Ningbo (NB). The automatic weather stations (AWS, circles), national weather stations (NWS, triangles), and radiosonde sites (squares) are classified into urban (solid) and non-urban (hollow) types. The energy balance observation sites (XJH, SWEP) are in Shanghai (inset). Section 2.2 provides more details.

      The well-known canopy layer urban heat island (CL-UHI) effect is the air temperature difference in canopy layer (~2 m) between the urban area and its surrounding rural area (Stewart, 2011). To be representative, careful consideration needs to be directed to the site selected (Schluenzen et al., 2021). Although the CL-UHI is widely reported for individual cities in the YRD, such as Shanghai (Zhang K. X. et al., 2010; Cui and Shi, 2012), Nanjing (Zeng et al., 2009), and Hangzhou (Chen et al., 2014), the classification of urban and rural areas for CL-UHI is critical (Grimmond et al., 1993; Arnfield, 2003; Stewart, 2011). Many rural observation sites, despite being away from an urban area, are impacted by urbanization to some extent (Lowry, 1977; Huang et al., 2019; Zhang et al., 2021). These impacts may differ with synoptic conditions and mesoscale circulations (e.g., land/sea, Miao et al., 2020). Thus, when the rural reference is contaminated by urban impacts, more detailed analysis of the source area of the observations is required (Oke et al., 2017; Schluenzen et al., 2021). This can be challenging to identify (Miao et al., 2020; Zhang et al., 2021).

      Across the YRD region (219,000 km2; Fig. 1), individual city effects merge and compound each other (Shepherd et al., 2013; Huang et al., 2019). This makes the interpretation of the CL-UHI—as a measure of difference between an urban and rural area—very complex. The boundary layer UHI (BL-UHI) is affected by the larger spatial extent of the city (horizontally and vertically, and is dominated by local to mesoscale processes) (Oke, 1976). In this study, unlike previous YRD CL-UHI investigations, we consider the urban thermal effects (UTE), not in reference to a single rural area outside one city; rather we consider the urban region and its surroundings through examining both the canopy layer and boundary layer influences.

      While UTE may not pose stress to citizens under normal thermal conditions, these urban thermal effects can aggravate heat stress and cause severe hazards when compounded by stressful heat wave (HW) temperatures (Wang et al., 2021). Many studies have documented CL-UHI intensification during HWs, both in the daytime (e.g., Jiang et al., 2019) and at night (e.g., Beijing, Guangzhou, Shanghai, etc.) (Schatz and Kucharik, 2015; Ao et al., 2019; Jiang et al., 2019). With urban expansion in the YRD, the urban influences have become evident in climate signals (Du et al., 2007). Enhancing our understanding of the regional UTE in the YRD is important for the large city clusters in this region to mitigate and adapt to the heat stress.

      In July and August 2013, a record-breaking heat event occurred in central–east China including the YRD. This was a large spatial event of intense heat, which persisted for a long period with extensive socio–economic impacts (Zhang et al., 2014). Previous studies of this event focus on the attribution of the HW in terms of large-scale atmosphere circulation (Peng, 2014; Sun, 2014) and climate change (Xia et al., 2016). The influences of human activities and urban areas in the 2013 HW are also detectable. Changes in anthropogenic forcing at both global and urban scales are found to increase the probability of HW events (Ma et al., 2017; Wang et al., 2017). However, the urban thermal effects in both the canopy and boundary layers across large city clusters under extreme heat wave conditions are still not clear.

      The objective of this study is to examine the three-dimensional (3D) UTE of the extensively urbanized YRD based on data from a dense observational network including automatic weather stations (Fig. 1), radiosondes (Table 1), and surface flux sensors. Meanwhile, influences of an extreme heat wave on the YRD UTE in both the canopy and boundary layers are also explored.

      Height (m AGL)Data sourceMean (Tmean)Minimum (Tmin)Maximum (Tmax)
      ~2AWS–MSAImean,2AImin,2AImax,2A
      ~2NWSImean,2N
      ~50–100, 750, 1500, 3000RadiosondeImean,R
      ERA5Imean,E
      Note: MSA denotes moving spatial anomaly.

      Table 1.  Notation used to distinguish the different aspects of the urban thermal effects (UTE)

    2.   Data and methodology
    • The YRD region in this study includes Shanghai Municipality, and Jiangsu and Zhejiang provinces (Fig. 1). It is home to about 150 million registered residents (National Bureau of Statistics of China, 2018), of whom approximately 70% are urban dwelling (Liao et al., 2015). Cities in this region include Shanghai, Nanjing, Hangzhou, Suzhou, Ningbo, Wuxi, and Changzhou. Population data for Shanghai Municipality, Jiangsu, and Zhejiang provinces (National Bureau of Statistics of China, 2014) combined with impervious and pervious land-co-ver fraction data (Gong et al., 2019a, b) indicate the extent urbanized (Fig. 2). There have been large population increases since the 1980s, with a “Z-shape” city cluster becoming apparent between 1993 and 2013 (Zhang N. et al., 2010; Zhong et al., 2017). Both the “Yangtze River Economic Belt” and the “Belt and Road Initiative” (Lu et al., 2018) cross the YRD.

      Figure 2.  Variations of urban characteristics through time for YRD: (a) population density (data from China Statistical Bureau 1983–2013) and (b) impervious fraction > 0.5 (impervious cover includes building roofs, driveways, and parking lots) in different time periods of first occurrence (color shaded; data from Gong et al., 2019a, b).

      We use the impervious land cover (Gong et al., 2019a, b) to identify urban stations. These are defined as having an impervious fraction ≥ 0.5 in 2013 in their surroundings (circle area: 10 or 100 km2; i.e., radius of 1.785 or 5.642 km). To obtain values for other years, we assume a linear change of ≥ 0.01 yr−1 (Jiang et al., 2020) derived from data for the 1981–2013 period. This follows the method used by others in the Chinese mainland (Ren and Ren, 2011) and United States of America (Hausfather et al., 2013).

    • In 2013, air temperature (T2m) was measured hourly at 2 m by automatic weather stations (AWS, Vaisala MILOS500, Finland) with a radiation shield (Tan et al., 2015) at 2269 sites (circles, Fig. 1). To reduce topographic site heterogeneity, those located at higher elevations (> 500 m above sea level; Dairaku et al., 2000) are excluded from our analysis.

      The AWS sites are flat open areas maintained by the China Meteorological Administration (CMA) regional offices (CMA, 2018). The hourly data are quality-controlled (QC) following World Meteorological Organization guidelines (Estévez, 2011) by the local meteorological bureaus of Shanghai, Jiangsu, and Zhejiang. The QC includes checks of consistency (e.g., internal, temporal, and spatial) and climatic range. The QC data are used to determine the daily mean, minimum, and maxi-mum 2-m air temperature (Tmean,2m, Tmin,2m, and Tmax,2m; Table 1).

      The urban or non-urban AWS class assignment (Fig. 1) using impervious fraction (Section 2.1) differs little between the two sizes of area (10 and 100 km2) analyzed. Overall, 584 (26%) of the sites are classified as urban. Most of these sites occur near the major city cluster and some smaller cities scattered across northern Jiangsu and southern Zhejiang.

      Quality-controlled daily 2-m air temperature data of CMA National Meteorological Information Center , observed four times per day [0200, 0800, 1400, and 2000 Beijing Time (BT)], are available from 141 YRD national weather stations (NWS; Zhai and Pan, 2003; Ren et al., 2005). Using the same approach for classification as for the AWS sites, 101 are urban. In this study, we analyze these data for the period 1980–2017 (Table 1).

      Over the region, there are seven CMA radiosonde stations (squares; Fig. 1, Table 1). These collect air temperature observations twice a day (0800 and 2000 BT) at 11 pressure levels (Guo et al., 2008; Guo and Ding, 2009). QC of these data includes assessment of allowable value range and gross errors (http://data.cma.cn/data/cdcdetail/dataCode/B.0011.0001C.html).

      As high-resolution L-band radiosonde data (Jiang et al., 2017; Han et al., 2019) are unavailable in the region, we obtain additional vertical information from ERA5 reanalysis data (Copernicus Climate Change Service, 2018). This data are hourly with 30-km horizontal resolution. Although there is not an urban scheme in the Integrated Forecast System (IFS) reanalysis model, the radiosonde and aircraft data from the region are assimilated.

      For both the radiosonde and ERA5 data, four levels (1000, 925, 850, and 700 hPa) are analyzed (Table 1), corresponding to about 50–100, 750, 1500, and 3000 m AGL (above ground level). To classify the radiosonde data, the impervious fraction within the 100-km2 circular area around the ground station is used as a proxy for the source area for the sensor as it rises above the surface. Three sites are classified as urban (four non-urban). The urban (non-urban) classification for the 2269 AWS stations is assigned to the ERA5 grid (0.25° × 0.25°).

      Shortwave and longwave radiation, turbulent sensible, and latent heat fluxes are measured at two sites in Shanghai: (1) commercial–residential area Xujiahui (XJH; 31.19°N, 121.43°E), and (2) grass area within the Shanghai World Expo Park (SWEP; 31.19°N, 121.47°E). The measurements at XJH are undertaken at 80 m AGL (25 m instrument mast on top of a 55-m building) and at SWEP 1.5 m AG. Sensor details and data processing techniques are the same for both sites (Ao et al., 2016, 2018). The net all-wave radiation (Q*), turbulent sensible (QH), and latent (QE) heat fluxes are directly measured by using 10-Hz data to determine 30-min fluxes. However, neither the anthropogenic heat flux (QF) nor net storage heat flux (ΔQS) are observed. Here, QF is not estimated. Although QF may increase under HW conditions (Stone, 2012), the QF heat emission can be expected to be relatively small (Ichinose et al., 1999) compared to large summertime shortwave radiation. The HW and non-HW day fluxes are compared at both sites using the median fluxes.

    • There are 584 (1685) urban (non-urban) AWS sites available. To reduce the uncertainty from the AWS instrument sites, a regional UTE is calculated by using a filtering window that is varied to detect spatial heterogeneities at different spatial scales (Wu and Yang, 2013). To reduce large-scale influences (e.g., of atmospheric circulation, global warming) and urban observation environment impacts, we use very large (radii of 200, 250, and 300 km) moving spatial anomaly (MSA) windows around each AWS site relative to canopy layer air temperature source area (order 1 km).

      The MSA provides mean temperature classified as urban $ \left({T}_{\rm{u}}' \right) $ and non-urban ($ {T}_{\rm{r}}' $). A regional urban–non-urban difference Iav(t) is determined for each day (t):

      $$ {I}_{{\rm{av}}}\left(t\right)=\sum _{i=1}^{{n}_{\rm{u}}}\frac{{T}_{{\rm{u}},i(t)}'}{{n}_{\rm{u}}}-\sum _{j=1}^{{n}_{\rm{r}}}\frac{{T}_{{\rm{r}},j(t)}'}{{n}_{\rm r}},$$ (1)

      using the number of urban (nu) and non-urban (nr) sites within the area. Similarly, for each urban site relative to all the non-urban AWS within the area:

      $$ {I}_{i}\left(t\right)={T}_{{\rm{u}},{i}}' -\sum _{j=1}^{{n}_{{\rm{r}}}}\frac{{T}_{{\rm{r}},{j\left(t\right)}}'}{{n}_{\rm{r}}}.$$ (1)

      Positive $ {I}_{i}\left(t\right) $ values indicate a strong UTE. The same techniques are used for both types of canopy layer data (AWS, NWS), but for the radiosonde (R) and ERA5 data (E), MSA are not used (Table 1). As the soundings are sparse and the source area for the sensor increases as it rises above the surface, direct urban–rural differences are used to measure the boundary layer UTE for the R and E data without MSA.

      Statistical properties of the daily T2m are assessed by using the skewness, kurtosis, interquartile range (IQR), median, and frequency distribution (Mardia, 1970). Skewness measures the asymmetry of the distribution, with positive (negative) skew indicating the presence of a long tail on the high (low) end of the distribution. Kurtosis measures the width of the distribution, with negative (positive) excess kurtosis describing a distribution that is wider (narrower) than the normal distribution. IQR gives the spread between 25th and 75th percentiles values (median is the 50th percentile). Linear regression of Tmean,2m is used to assess the relation between the UTE intensity and HW.

    • Following Lau and Nath (2012) and Li et al. (2015), HW days are identified by using $T $max,2m and $T $min,2m thresholds. The CMA operational regulation (Huang et al., 2011) defines high temperatures as daily $T $max,2m ≥ 35°C and people’s discomfort coming from nocturnal $T $min,2m ≥ 25°C (Zittis et al., 2015). HW days need to meet both criteria (i.e., $T $max,2m ≥ 35°C and $T $min,2m ≥ 25°C). Using the areal average of the 2269 YRD AWS, 34 days satisfy $ {T}_{{\rm{min}},2{\rm{m}}\; \geqslant \;25}^{{\rm{max}},2{\rm{m}}\; \geqslant \; 35} $ during July–August of 2013.

    3.   Regional urban thermal effects (UTE) during the heat wave
    • In summer 2013, 97.5% of the YRD AWS recorded at least one Tmax,2m greater than 35°C. At many YRD sites, the summer 2013 temperature observations broke local records (Ma et al., 2017; Wang et al., 2017). Both warm days (i.e., Tmax,2m ≥ 35°C) and nights (i.e., Tmin,2m ≥ 25°C) were extremely persistent, occurring on more than 40 days over the YRD Z-shaped city cluster in July–August of 2013 (Fig. 3).

      Figure 3.  Number of days in July–August 2013 in the Yangtze River Delta (YRD) for (a) daily Tmax,2m ≥ 35°C and (b) both daily Tmax,2m ≥ 35°C and Tmin,2m ≥ 25°C, based on the data from the automatic weather stations (AWS).

    • Analysis of all the AWS air temperature data in the YRD region (Section 2.3) shows that the city cluster has a negative skewness (−0.63°C; Fig. 4, Table 2). The urban (red; Fig. 4) temperature distribution is shifted to the right of the non-urban (blue, Fig. 4), especially for Tmin,2m (Fig. 4b), indicating the urban sites are obviously warmer at night. The urban Tmin,2m and Tmean,2m distributions are narrower (i.e., higher kurtosis or closeness to the mean) than the non-urban observations, but differences between urban and non-urban sites for Tmax,2m are undetectable (Fig. 4c). Thus, the frequency distribution (0.5°C bins) of Tmin,2m has the expected feature of a regional UTE.

      Figure 4.  Frequency distributions (0.5°C bins; %) during July–August 2013 of (a, d) Tmean,2m (°C), (b, e) Tmin,2m (°C), and (c, f) Tmax,2m (°C) at all urban (U, red) and non-urban (nU, blue) automatic weather station (AWS) sites over the (a–c) entire period and the period of (d–f) heat wave (HW, 34 days) and non-HW (nHW, 28 days) conditions. Values given within each plot are the medians.

      Tmean,2m (°C)Tmin,2m (°C)Tmax,2m (°C)
      UnUUnUUnU
      July–AugustSkewness−0.59−0.36−0.91−0.37−0.43−0.29
      Kurtosis0.490.102.470.38−0.10−0.11
      IQR3.13.42.63.34.24.8
      Median30.929.227.225.236.035.4
      HWSkewness−0.63−0.35−0.66−0.19−0.43−0.37
      Kurtosis0.34−0.010.39−0.170.490.18
      IQR2.83.22.63.43.64.2
      Median31.629.827.525.436.836.3
      non-HWSkewness−0.78−0.50−1.37−0.78−0.23−0.10
      Kurtosis0.670.232.941.08−0.59−0.29
      IQR2.93.02.72.94.24.7
      Median29.227.526.024.533.332.5

      Table 2.  Analysis of the frequency curves (Fig. 6) of daily Tmean,2m (°C), Tmin,2m (°C), and Tmax,2m (°C) from AWS under HW and non-HW conditions for urban (U) and non-urban (nU) sites in the YRD (Fig. 1) during July–August 2013. For statistical methods see Section 2.3

      In this region, during July–August 2013, 34 days meet the HW definition (Section 2.4) and 28 not. The expected skew towards higher temperature occurs in both urban and non-urban temperature distribution on HW days (Fig. 4, Table 2), most evident in Tmin,2m. The Tmin,2m IQR is larger under HW conditions for non-urban than urban sites. There is a pronounced tail toward higher Tmin,2m. The canopy layer UTE is largest for Tmin,2m [Eq. (1), Iav,min]. It varies between 0.7 and 1.6°C on average for the three radii extents (Fig. 5b); whereas Iav,mean,2A is slightly less (0.6–1.2°C, Fig. 5a) and Iav,max,2A is the smallest (0.2–0.5°C, Fig. 5c). The large Iav,min,2A is consistent with previous studies over the YRD (e.g., Nanjing, Shanghai; Zeng et al., 2009; Cui and Shi, 2012). The UTE for Tmax,2m is much smaller than for Tmin,2m and Tmean,2m. Negative UTE Tmax,2m occurs on some days, as observed in traditional studies of urban–rural afternoon temperature differences (Basara et al., 2008). This is a consequence of lower urban sky view factors. Here, negative UTE Tmax,2m occurs on a few days during the weaker part of the HW (i.e., Tmax,2m just greater than 35°C) with southeasterly flows from the sea causing the urban Tmax,2m to be less than that downwind of non-urban areas.

      Figure 5.  Temopral evolution of daily regional averaged canopy layer urban thermal effects (UTE) [Iav,2A; °C] for (a) Tmean,2m (°C), (b) Tmin,2m (°C), and (c) Tmax,2m (°C) temperature derived from the automatic weather station (AWS) data, using three areal extents [MSA radii (km): 200 (blue), 250 (yellow), and 300 (red)] during July–August 2013 in the YRD. For methods see Section 2.3.

      Both Iav,mean,2A (0.2–0.3°C) and Iav,min,2A (0.5–0.6°C) are significantly enhanced under HW conditions (p < 0.01), similar to the enhanced individual city’s CL-UHI intensity under the same conditions (Ao et al., 2019; Jiang et al., 2019). However, Iav,max,2A does not have a significant change under HW conditions. There are somewhat lower Iav,max,2A values during HW than non-HW conditions. This feature is also found over other city clusters in China, such as the Beijing–Tianjin–Hebei metropolitan region, but the urban effects in Beijing may be greater in winter than summer (Yang et al., 2013).

      Analysis of the canopy layer UTE using NWS data Tmean,2m (Iav,mean,2N) for July–August of 1980–2017 shows that the UTE in 2013 is the largest (Fig. 6), probably caused by the extensive urbanization over this period and the extreme HW in 2013. The UTE may be enhanced by extreme HWs during midsummer as the correlation between UTE and Tmean,2m increases to 0.85 from annually averaged 0.76. Linear regression between Iav,mean,2N and Tmean,2m suggests that Iav,mean,2N is strengthened by 0.2°C when Tmean,2m increases by 1°C.

      Figure 6.  Interannual variation of the average urban thermal effect (UTE) [Iav,mean,2N; °C, black] and mean temperature (Tmean,2m; °C, orange) derived from the national weather station (NWS; Fig. 1) data during July−August of 1980–2017 in the YRD.

    • To examine the canopy layer UTE across the region, MSA is computed for all YRD AWS Ii [Eq. (2), Section 2.3]. For a 200-km radius, the results show the positive (warmer) Ii,mean,2A and Ii,min,2A (Figs. 7a, b) have a Z-shape pattern similar to the YRD urbanization spatial pattern (Fig. 1). Areas with the largest canopy layer UTE (i.e., both Ii,mean,2A > 0.5°C and Ii,min,2A > 0.5°C) cover the Z-shape city cluster, and smaller economically developed city groups scattered across Jiangsu and Zhejiang provinces. However, positive Ii,max occurs south of the Z-shape city cluster (i.e., northern Zhejiang, Fig. 7c) with fewer areas above 0.5°C. This lowers Iav,max,2A under HW conditions. The canopy layer UTE extends into nearby non-urban areas under HW conditions for both Ii,mean,2A (Fig. 7d) and Ii,min,2A (Fig. 7e) > 0.5°C. The latter area increases the most. During the HW, Ii,max,2A intensity is greater than in non-HW conditions (Fig. 7f). Areas where Ii,mean,2A and Ii,min,2A are > 0.5°C (Figs. 7d, e) indicate that the canopy layer UTE extends to nearby non-urban area under HW conditions, with Ii,mean,2A having the largest extension. The canopy layer UTE extension to the nearby non-urban area under HW conditions may result from the combination of individual city’s CL-UHI being amplified by the HW (Ao et al., 2019) and the predominant southerly winds (Jiang et al., 2019) being advected northwards. Similar canopy layer UTE distributions are found (figure omitted) for MSA areas with larger radii (e.g., 250 and 300 km).

      Figure 7.  Spatial distributions of the average urban thermal effect (UTE) [Ii,mean,2A; °C] calculated from the automatic weather station (AWS) data using a 200-km radius during (a–c) non-HW (nHW; 28 days) and (d–f) heat wave (HW; 34 days) conditions for (a, d) mean temperature (Ti,mean,2A; °C), (b, e) minimum temperature (Ti,min,2A; °C), and (c, f) maximum temperature (Ti,max,2A; °C) in the YRD.

      Dominated by the East Asian summer monsoon, southerlies prevail in the YRD lower troposphere during July–August (Tao and Chen, 1987). Wind may advect urban heat from neighboring cities to non-urban areas. For example, high temperature anomalies expanded northwestward in Shanghai and its neighboring cities from 8 to 12 July 2013 (Jiang et al., 2019). Similarly, some non-urban areas in the YRD are influenced by urban heat during both non-HW and HW periods in the summer of 2013 (Fig. 7). By taking the non-urban areas with UTE index larger than 0.5°C as those influenced by urban heat, further calculations show that about 28% (non-HW) and 35% (HW) of the non-urban areas are affected during two periods. This suggests (1) some rural sites are contaminated by urban influences in the YRD, (2) the UTE is a more appropriate term than the “UHI” for discussing urban effects of a large city cluster (such as the YRD), and (3) the quantitative urban influences on non-urban areas differ with synoptic conditions (a topic to be explored in detail in future work).

      Consistent with the AWS results, the NWS data show that Ii,2N is enhanced during the HW period (figure omitted). Given that most NWS are urban (Section 2.2), the canopy layer UTE amplification at NWS is greater (0.2–0.9°C) than at the AWS during the HW period. The spatial distribution of Ii,2N is slightly different from Ii,2A. A higher canopy layer UTE in the city cluster and a high canopy layer UTE in south of the YRD are evident. However, more (> 72%) of the urban NWS are in the south of the YRD. Therefore, the AWS data show better both the strength and the spatial distribution of canopy layer UTE in the YRD. However, importantly the characteristics of the canopy layer UTE are similar in both the AWS and NWS data.

    • The surface energy balance (SEB) plays a key role in the spatial difference of temperature within and around cities (Oke, 1982; Grimmond et al., 1991), which varies under HW conditions (Li et al., 2015; Sun et al., 2017). The marked differences in energy partitioning between urban and non-urban surfaces are well-documented (Oke, 1982; Piringer et al., 2002; Oke et al., 2017), with sensible heat flux distinctively larger than the latent heat flux over areas with little vegetation, and increased storage heat fluxes in urban areas (Grimmond and Oke, 1999; Loridan and Grimmond, 2012). As Shanghai is the most urbanized city in the YRD, the surface energy flux differences are analyzed in HW and non-HW conditions (Fig. 8).

      Figure 8.  Diurnal differences (Δ = HW − nHW) in median surface energy balance fluxes: (a) net all-wave radiation (Q*; W m−2), (b) turbulent sensible heat (QH; W m−2), (c) turbulent latent heat (QE; W m−2), and (d) Bowen ratio (QH/QE) at an urban (XJH, U; red) and a grass (SWEP, nU; blue) site. For methods see Section 2.2. Note that the number of days for heat wave (HW) is 34 while that for non-HW (nHW) is 28.

      Daytime net all-wave radiation (Q*) is expected to be larger during the HW because of increased solar radiation from clearer skies (Schatz and Kucharik, 2015). The observed total cloud cover is much less during the HW (3.9/10.0) compared to the non-HW (6.6/10.0) at XJH and similarly (HW 4.0/10.0; non-HW 7.0/10.0) at SWEP. At both sites, Q* increases during the HW (Fig. 8a). However, Q* is similar between the two nearby sites (11 km separation, Fig. 1) and consistent with urban–non-urban Q* comparisons (Oke et al., 2017).

      The long-persisting HW in the YRD is associated with a strong and westward western Pacific subtropical high. Days tend to be cloudless and clear sky enhances the incoming shortwave radiation, therefore increasing Q*, which in turn provides more energy for sensible and latent heat fluxes. As expected, the urban sensible heat flux (QH) is positive throughout the day whereas the grass site QH is negative at night (2000–0600 BT, figure omitted). Under HW conditions, the QH increases in the afternoon (1300–1600 BT) at both flux sites with a larger increase at XJH (35.0 W m−2) than at the grass site (1.1 W m−2, Fig. 8b).

      The latent heat flux (QE) also increases in afternoon–early evening (1300–2100 BT) at both sites during the HW (Fig. 8c). The median increase at XJH is 1.3 W m−2; whereas at the grass site it is 32.1 W m−2. The median Bowen ratio (QH/QE, Fig. 8d) at XJH increased from 6.5 (non-HW) to 8.4 on HW days, but decreased at the grass site from 0.11 (non-HW) to 0.07 (HW). The relatively lower daytime urban QE increase allows the enhanced QH to warm the urban atmosphere under HW conditions, consistent with HW conditions in London (Ward and Grimmond, 2017). Larger latent heat fluxes (or mass equivalent evaporation) will dry the soil. Unfortunately, we lack soil moisture data to assess this. Reduction in soil moisture is regarded as a principal driver to UTE amplification, but there is a tradeoff between sensible heat and reduced non-urban latent cooling (Fischer et al., 2007), such as that seen in the record-breaking 2003 European summer HW.

    • The differences in SEB partitioning and canopy layer air temperatures are expected to cause differences in heating of the boundary layer (Zhang et al., 2020). The larger sensible heat flux from the urban surface enhances growth of the urban boundary layer, causing it to develop faster and to a greater height than the non-urban counterpart. Release of large urban storage heat fluxes helps the mixing layer to persist for longer. To explore the vertical characteristics of the UTE, both radiosonde and ERA5 data (denoted by the subscript R and E, respectively) are analyzed.

      Radiosonde air temperatures between 1000 and 700 hPa above both urban and non-urban areas are warmer on HW than non-HW days (Figs. 9a, b, d, e; black), with the differences (HW minus non-HW) decreasing with height. The Iav,mean,R profiles (Figs. 9c, f; black) show that the vertical UTE extends to at least 925 hPa (~750 m) but is nearly undetectable at 850 hPa (~1500 m) at both 0800 and 2000 BT. The HW-day Iav,mean,R is significantly (p < 0.05) warmer than that on non-HW days from 1000 to 925 hPa, but above 850 hPa the inverse occurs (i.e., cooler on HW than non-HW days). Analysis of the ERA5 air temperature reveals similar vertical profiles to those of radiosonde air temperature (Figs. 9a–e; gray). Although Iav,mean,E covers the whole YRD, it is much smaller than Iav,mean,R on both HW and non-HW days; the HW-day Iav,mean,E is significantly (p < 0.05) strengthened on HW days from 1000 to 925 hPa at both 0800 and 2000 BT (Figs. 9c, f), while it is also undetectable at 850 hPa at both 0800 and 2000 BT as that in Iav,mean,R.

      Figure 9.  Vertical profiles of air temperature (°C) and urban thermal effect (UTE) in the Yangtze River Delta (YRD) during July–August 2013 under heat wave (HW, 34 days, solid) and non-HW (nHW, 28 days, dashed) periods from radiosonde (R, black) and ERA5 (E, gray) data for (a, d) urban (U), (b, e) non-urban (nU), and (c, f) U−nU difference [Iav,mean, °C] at (a–c) 0800 and (d–f) 2000 BT. For methods see Section 2.3.

      The early evening (2000 BT) and morning (0800 BT) vertical UTEs decrease with height below 850 hPa under HW conditions (i.e., decreasing urban minus non-urban difference with height) as the heat transported from the underlying surface is mixed into an increasing larger volume of air. The subsidence inversion associated with the synoptic-scale high pressure inhibits the vertical UTE expanding above 850 hPa during the HW.

    4.   Final comments
    • During the 2013 extreme HW, the canopy layer air temperature distribution in the distinctive Z-shape YRD city-cluster (cf. non-city-cluster) is narrower and skewed towards warmer temperatures, especially at night. The mean canopy layer UTE in the region is up to 0.6–1.2°C, with the nocturnal UTE (0.7–1.6°C on average) larger than daytime (0.2–0.5°C on average). During the HW, the nocturnal canopy layer UTE is 0.5–0.6°C larger than under non-HW conditions. The stronger canopy layer UTE (> 0.5°C) is mainly located in the Z-shape city cluster but extends beyond under HW conditions. Thus, the HW strengthens the canopy layer UTE and extends horizontally.

      During the HW, the difference in Bowen ratios between urban and non-urban areas is enlarged, with both a larger Bowen ratio observed at the urban site and a smaller value over the grass site. The UTE is found to extend vertically to at least 925 hPa (~750 m) or 850 hPa (~1500 m). Unlike the canopy layer UHI (CL-UHI) in individual cities dominated by immediate surroundings, the 3D UTE in large city clusters is affected by presence of the urban area at its lower boundary. In this region with a large city cluster, the effect of the urban area extends in three dimensions (horizontally and vertically) under extreme HW conditions. The regional UTE contaminates the surrounding non-urban areas, thus making the signature of individual cities difficult to measure using the classical urban heat island assumptions (Stewart and Oke, 2012). The methodology proposed in the study allows the 3D regional UTE to be assessed.

      In summary, this study provides insight into the 3D UTE for a large city cluster. Insights gained from the record-setting July–August 2013 HW in YRD and the resulting UTE are important given such conditions are projected to increase through the mid to late 21st century. Unlike previous studies, we consider the 3D UTE in both the canopy and boundary layers and enhance our understanding of the effects of urban clusters, especially in the vertical dimension. Utilizing the densest automatic wea-ther station network, allows us to report the UTE variation at a high spatial resolution. In terms of future work, hourly and finer (minute) observations (i.e., higher temporal resolution) would be extremely beneficial to improve understanding of the rates of change and the variability (temporal and spatial). Similarly, a wider range of sensors (e.g., L-band radiosondes, Doppler radar, and thermal profilers) would provide more vertical information. These combined would enable a better understanding of the 3D impacts of large city clusters in a region.

      Acknowledgments. We gratefully acknowledge the reviewers and editors for their helpful suggestions to improve this manuscript.

Reference (72)

Catalog

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return