The altered land surface and anthropogenic heat release increase the sensible heat flux from the land surface to the atmosphere in city areas, leading to higher temperatures than in surrounding rural areas— a phenomenon known as the urban heat island (UHI) (Howard, 1820; Balchin and Pye, 1947; Oke, 1982; Gordon, 1994; Kalnay and Cai, 2003; Li et al., 2004; Zhang et al., 2005; Liu et al., 2007; Ren et al., 2007; Miao et al., 2009; Oleson et al., 2010; Yang et al., 2013; Georgescu et al., 2014). The UHI can not only produce detrimental impacts, such as degradation in air quality and water quality, amplification in heat waves and storms, and increases in energy demands for air conditioning in summer (e.g., Weaver et al., 2009; Milojevic et al., 2011; Myhre et al., 2013), but also beneficial effects, such as less demands for heating in winter (e.g., Taha, 1997; Stewart and Oke, 2012). Therefore, it is important to understand the nature and causes of the UHI, to develop mitigation or adaptation strategies in the future.

The Chinese New Year (CNY), the beginning of the lunar New Year, is traditionally China's most important public holiday, officially lasting for one week. Since Reform and Opning-up, many people from rural or underdeveloped areas have migrated to big cities. During the CNY holiday, many people in big cities return to their native places for traditional family gatherings.

The reduction of the urban population during the CNY holiday decreases human activities in urban areas, thus leading to reductions in anthropogenic heat emissions from such as vehicles, heating, industrial energy consumption, and human metabolism. The effects of anthropogenic heat on the UHI are relatively large in mid- and high-latitude cities in winter due to weaker solar radiation input, shallower boundary layer, and greater energy use for heating (KŁ ysik, 1996; Ichinose et al., 1999; Fan and Sailor, 2005; Bohnenstengel et al., 2014). Less human activities can also directly and indirectly affect other processes and thus change the UHI (Bonan, 2008). However, how the mass human migration affects the UHI during the CNY holiday is still unknown.

Harbin City is the capital of Heilongjiang Province, Northeast China, and covers the area from 44° 04'N to 46° 40'N and from 125° 42' E to 130° 10'E. The city is known for its coldest climate and longest winter among the major cities of China. Harbin City is a key cultural, political, economic, and scientific center in Northeast China. It has experienced a rapid growth in population from 4.22 million in 1990 to 9.80 million in 2006. Migrant workers and college students usually leave the city for family reunions before the CNY holiday. Also, some local residents leave the city for visiting their relatives and friends, or other purposes. This study investigates the role of modern human migration for the UHI during the CNY holiday in Harbin City for the period of 1992-2006.

The homogenized daily mean (T_mean), maximum (T_max) and minimum (T_min) temperatures used in this study are obtained from the China Meteorological Administration. The UHI effect is defined as the temperature differences between urban (Harbin) and rural (Fangzheng) stations (DT= T_urban-T_rural). The description of the two stations is provided in Table 1.

The CNY day is determined according to the lunar calendar, therefore, the date of the CNY day changes with year. Table 2 shows that the date of the CNY day varies between 22 January and 19 February during the period of 1992-2006. In this study, the CNY day is denoted as day +1, and the day before as day -1. The CNY week is from day +1 to day +7. Our study period includes nine weeks from 28 days before to 34 days after the CNY day. The dates of the CNY day, the CNY week, and the start (day -28) and end (day +35) days of the study period are listed in Table 2.

Figure 1 shows daily mean (DT_mean), maximum (DT_max), and minimum (DT_min) temperature differences between urban and rural stations from 28 days before the CNY day (day -28) to 34 days after the CNY day (day +35) averaged over the period of 1992-2006 in Harbin City. DT_mean, DT_max, and DT_min from day -28 to day +35 show a similar feature. All three temperature variables have relatively lower values during the CNY week. Meanwhile, some differences exist. DT_mean ranges between 0.37° C and 3.11° C, and the mean value is 1.43° C for day -28 to day +35. During the CNY week (day +1 to day +7), DT_mean is lower than the mean value. DT_max values during day -28 to day +35 exhibit smaller magnitudes than DT_mean with a mean value of 0.62° C, and during the CNY week, they are lower than the mean value except on day +1. Comparatively, DT_min has the larger magnitude than DT_mean, varying from -0.26° C to 3.36° C during day -28 to day +35. DT_min during the CNY week is much lower than the mean value of 1.49° C.

Table 1

Table 1

Table 1 The description of urban and rural stations. Harbin Fangzheng Latitude (° N) 45° 45' 45° 51' Longitude (° E) 126° 46' 128° 51' Altitude (m) 143 119 Location Urban region Rural region

Table 1 The description of urban and rural stations.

Table 2

Table 2

Table 2 Dates of the Chinese New Year (CNY), CNY week, 28 days before the CNY day (day -28), and 34 days after the CNY day (day +35) from 1992 to 2006. Year CNY day CNY week Day -28 Day +35 1992 4 February 4-10 February 7 January 9 March 1993 23 January 23-29 January 26 December 26 February 1994 10 February 10-16 February 13 January 16 March 1995 31 January 31 January - 6 February 3 January 6 March 1996 19 February 19-25 February 22 January 24 March 1997 7 February 7-13 February 10 January 13 March 1998 28 January 28 January - 3 February 31 December 3 March 1999 16 February 16-22 February 19 January 22 March 2000 5 February 5-11 February 8 January 10 March 2001 24 January 24-30 January 27 December 27 February 2002 12 February 12-18 February 15 January 18 March 2003 1 February 1-7 February 4 January 7 March 2004 22 January 22-28 January 25 December 25 February 2005 9 February 9-15 February 12 January 15 March 2006 29 January 29 January - 4 February 1 January 4 March

Table 2 Dates of the Chinese New Year (CNY), CNY week, 28 days before the CNY day (day -28), and 34 days after the CNY day (day +35) from 1992 to 2006.

Figure 1

Figure Option ViewDownloadNew Window

Figure 1 The urban heat island (UHI) effects from 28 days before the CNY day (day -28) to 34 days after the CNY day (day +35), averaged over the period of 1992-2006 in Harbin City: (a) DTmean, (b) DTmax, and (c) DTmin. DTmean, DTmax, and DTmin represent the daily mean, maximum, and minimum temperature differences between urban and rural stations, respectively. The UHI effect during the CNY week is highlighted by the gray shaded area, and the blue dashed line represents the mean UHI value for day -28 to day +35.

Since the CNY holiday officially lasts for one week, we further examine the weekly means of DT_mean, DT_max, and DT_min during day -28 to day +35 averaged over the period of 1992-2006 (Fig. 2). Here, we define the CNY week as week +1, one week before as week -1, one week after as week +2, and so on. DT_mean has the lowest value of 0.86° C in the CNY week, and the weekly mean DT_mean values fluctuate between 1.25° C and 1.71° C in non-holiday weeks. DT_min is much stronger than DT_max in non-holiday weeks. However, during the CNY week, DT_max and DT_min only have small differences with magnitudes of 0.48° C and 0.34° C, respectively. Therefore, the reduction of DT_min is much larger than that of DT_max during the CNY week, compared to non-holiday weeks.

Figure 2

	Figure Option ViewDownloadNew Window
	Figure 2 Weekly means of the UHI effects during day -28 to day +35 averaged over the period of 1992-2006 in Harbin City: (a) DTmean, (b) DTmax, and (c) DTmin. The CNY week is denoted as week +1, one week before as week -1, one week after as week +2, and so on.

To further compare the UHI differences between the CNY holiday and non-holiday times, we define four weeks before and four weeks after the CNY week as the background period. Table 3 lists the values of the UHI during the CNY week and the background period, the differences between them, and the relative changes of the UHI during the CNY week to the background period averaged over the period of 1992-2006. DT_mean, DT_max, and DT_min during the CNY week are consistently much lower than during the background period. The differences between them are 0.65° C, 0.31° C, and 1.14° C, respectively. The relative changes of DT_mean and DT_max during the CNY to the background period have the similar value of 43% and 48%. Comparatively, the change of DT_min can reach 71%. The differences in DT_mean and DT_min are significant at the 99% confidence level by Student’ s t-test, and the DT_max difference is significant at the 98% confidence level.

Harbin City, located at the highest latitude, has the coldest climate among major cities in China. The city has undergone rapid urbanization in the last decades. During the CNY holiday, many people in the city return to their native places or leave for other purposes for celebrating the most important Chinese holiday. This provides us an unique opportunity to explore how mass human migration affects the UHI. In this study, we investigate the role of mass human migration for the UHI during the CNY holiday in Harbin City for the period of 1992-2006 using observational data.

Table 3

Table 3

Table 3 Statistics of the UHI effects expressed as daily mean (DTmean), maximum (DTmax), and minimum (DTmin) temperature differences between urban and rural stations averaged over the period of 1992-2006. The background period is defined as week -4 to week -1 and week +2 to week +5. UHI effect during the CNY week (° C) UHI effect during the background period (° C) Difference of UHI effect between the CNY week and the background period (° C) Relative change of the UHI effect during the CNY week to the background period (%) DTmean 0.86 1.50 -0.65* * 43 DTmax 0.34 0.65 -0.31* 48 DTmin 0.48 1.62 -1.14* * 71 * Significant at the 98% confidence level by Student’ s t-test. * * Significant at the 99% confidence level by Student’ s t-test.

Table 3 Statistics of the UHI effects expressed as daily mean (DTmean), maximum (DTmax), and minimum (DTmin) temperature differences between urban and rural stations averaged over the period of 1992-2006. The background period is defined as week -4 to week -1 and week +2 to week +5.

Our results indicate that the UHI effects during the CNY week are much lower than during the background period (four weeks before and four weeks after the CNY week). The reduction of DT_mean during the CNY week is 0.65° C, with respect to the background period. The reductions of DT_max, and DT_min are asymmetric, with magnitudes of 0.31° C and 1.14° C, respectively. The relative changes of DT_mean, DT_max, and DT_min to those during the background period are 43%, 48%, and 71%, respectively. These changes in DT_mean and DT_min are significant at the 99% confidence level by Student’ s t-test, and the DT_max change is significant at the 98% confidence level. The reduced population during the CNY holiday in the urban area of Harbin City leads to less human activities, which results in less anthropogenic heat emissions and also affects other processes, thus significantly reducing the UHI effects. Our findings provide observational evidence that mass human migration can significantly affect the UHI.