Environ. Res. Lett. 11 (2016) 074003 doi:10.1088/1748-9326/11/7/074003

LETTER

Weekly cycles in peak time temperatures and urban heat islandintensity

Nick Earl1, Ian Simmonds1 andNigel Tapper2

1 School of Earth Sciences, TheUniversity ofMelbourne, Parkville,Melbourne, Victoria, 3010, Australia2 School of Earth, Atmosphere and Environment,MonashUniversity andCooperative ResearchCentre forWater Sensitive Cities, Clayton,

Melbourne, Victoria, 3800, Australia

E-mail: nearl@unimelb.edu.au

Keywords: anthropogenic activity, weekly cycles, diurnal cycle, urban heat island

Supplementarymaterial for this article is available online

AbstractRegular diurnal andweekly cycles (WCs) in temperature provide valuable insights into theconsequences of anthropogenic activity on the urban environment. Different locations experience arange of identifiedWCs and have very different structures. Two important sources of urban heat arethose associatedwith the effect of large urban structures on the radiation budget and energy storageand those from the heat generated as a consequence of anthropogenic activity. The former forcingwillremain relatively constant, but aWCwill appear in the latter.WCs for specific times of day and theurban heat island (UHI)have not been analysed heretofore.We use three-hourly surface (2m)temperature data to analyse theWCs of sevenmajor Australian cities at different times of day and todetermine towhat extent one of ourmajor city’s (Melbourne)UHI exhibits aWC.We show that theWCof temperature inmajor cities differs according to the time of day and that theUHI intensity ofMelbourne is affected on aWC. This provides crucial information that can contribute toward thepush for healthier urban environments in the face of amore extreme climate.

1. Introduction

Regular diurnal and weekly cycles (WCs) of meteor-ological parameters provide valuable insights into theconsequences of human activities (e.g. industrialactivity, generating electricity, biomass burning andpowering motor vehicles, which commonly change atweekends), particularly in urban areas. Such insightsare possible because many human activities followfairly regular diurnal and weekly schedules. SinceAshworth (1929) ‘accidentally’ discovered a weeklySunday minimum when investigating rainfall in theindustrial town of Rochdale, England, the concept ofhumans affecting their surroundings on a daily scaleinspiredmanyWC studies as summarised by Sanchez-Lorenzo et al (2012).

Different locations experience a wide range ofidentified WCs, and these are often contradictory innature. Many researchers have found significant WCsin various meteorological parameters in and aroundurban/industrial areas (Fujibe 1987, 2010, Simmonds

and Keay 1997, Cerveny and Balling 1998, 2005, For-ster and Solomon 2003, Shutters and Balling 2006,Gong et al 2006, 2007, Bell et al 2008, Laux and Kunst-mann 2008, Bell et al 2009, Sitnov 2010, Rosenfeld andBell 2011, Gruzdev 2013, Farias et al 2014, Georgouliaset al 2015)while others found that the observed signalsare no stronger than the background variability(DeLisi et al 2001, Barmet et al 2009, Hendricks Frans-sen et al 2009, Stjern 2011). Daniel et al (2012) suggestthat a wide variety of statistical tests are often usedinappropriately and this is at least partly responsiblefor the contradiction. Anthropogenic activity linked toWCs can be thought of as due to human heat genera-tion and the release of atmospheric pollution (Sim-monds and Keay 1997), with the accompanyingemission of moisture also affecting the urban energybudget (Sailor 2011). The three major sources ofanthropogenic heat, closely related to energy con-sumption, are transportation, buildings and industry,as explained by Sailor (2011). However, atmosphericdynamics, aerosol andmoisture release complicate the

OPEN ACCESS

RECEIVED

7March 2016

REVISED

23May 2016

ACCEPTED FOR PUBLICATION

2 June 2016

PUBLISHED

1 July 2016

Original content from thisworkmay be used underthe terms of the CreativeCommonsAttribution 3.0licence.

Any further distribution ofthis workmustmaintainattribution to theauthor(s) and the title ofthework, journal citationandDOI.

© 2016 IOPPublishing Ltd

http://dx.doi.org/10.1088/1748-9326/11/7/074003mailto:nearl@unimelb.edu.auhttp://dx.doi.org/10.1088/1748-9326/11/7/074003http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/11/7/074003&domain=pdf&date_stamp=2016-07-01http://crossmark.crossref.org/dialog/?doi=10.1088/1748-9326/11/7/074003&domain=pdf&date_stamp=2016-07-01http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0http://creativecommons.org/licenses/by/3.0

observed surface air temperature relationship withanthropogenic activity. Aerosol and moisture interac-tion with solar radiation (direct effect) and aerosolsaltering clouds characteristics and thermodynamics(indirect effect) vary greatly with time of year, surfaceland use and a wide range of other influences (e.g. citysize, location, types of local circulations; Bell andRosenfeld 2008). Therefore, the net effect of aerosol onsurface temperature is highly variable and underdebate.

The presence and/or strength of WCs in urbanareas are related to the driving forces for the anthro-pogenic use of energy. These include population, char-acter of industrial activity, patterns and intensity ofvehicular traffic, typical synoptic patterns (and impli-cations for, e.g., ventilation), physical geography of thecity, proximity of the city to large water bodies andtheir direction in relation to the prevailing winds, fre-quency of inversions (and their implications for con-vection and for trapping of heat) and typical rain-producing mechanisms of the area. The relativeimportance of these factors differs between cities,strongly influenced by background climate.

Diurnal rhythms also present perspectives onanthropogenic influences. The typical diurnal warm-ing and cooling regime of high-density urban areas isvery different from that of rural or low-density urbanareas. After nightfall, urban sites cool relatively slowlydue to a number of anthropogenic causal mechan-isms, discussed in detail by Oke (1982). This steadydecline continues throughout the night, whereas ruralareas cool fast soon after sunset and then more slowlylater as radiation emission decreases. After sunrise, therural sites warm faster, catching up with the urbanlocations and generally both reach a temperature peakin mid-afternoon. This rural warming can lead to anearly morning urban ‘cool island’ in some cities due tothe atmospheric boundary layer being deeper over theurban area, reducing the apparent earlymorning heat-ing compared to the rural environment (Oke 1982,Theeuwes et al 2015). However other factors includingshading in urban canyons, leading to a smaller surfacearea of accessible insolation, greater total urban sur-face area (compared toflat surface) and enhanced earlymorning absorption of radiation by urban materials(e.g. concrete and asphalt) may also be involved. Thismeans that the urban heat island (UHI) intensityundergoes a marked diurnal variation with the largestdifference typically 3–5 h after sunset and smallest,which may sometimes be negative, from early morn-ing tomid-afternoon. For reasons similar to those dis-cussed above, the diurnal evolution of the UHI candiffer significantly with location. The diurnal cycle,combined with theWC, provides us with the opportu-nity to examine when anthropogenic activity has thelargest influence on the environment. This is the firstmajor focus of the paper.

There is a range of postulated causes of urban heat(Oke 1982). Two of the more important sources of

urban heat are those associated with the effect of largeurban structures on the radiation budget and energystorage and from the heat generated from anthro-pogenic activity (Rizwan et al 2008). The former for-cing will remain relatively constant, but a WC willappear in the latter. The nature (including the diurnalprogression) and intensity of the UHI changes withcity characteristics (size, density, industrial activity,traffic intensity, amount of green space etc; Stewartand Oke 2012), latitude, time of day (e.g. Kolokotroniet al 2006), background climate (e.g. Zhao et al 2014)and weather types (e.g. Lowry 1977, Morris and Sim-monds 2000). This means that the UHI in cities in tro-pical areas may contrast with those in temperateregions.With Australia possessing cites with a range ofclimates, it is an ideal location for this study. Lowry(1977) provides a conceptual framework that high-lights the difficulties in defining and measuring theUHI, with factors including background climate, localtopography and effect of local urbanisation. Under-standing the characteristics of the UHI is of consider-able importance due to the rises in urban population(table S1). UHIs have a profound effect on the lives ofurban residents and the health impact of heatwaves isone factor that motivates the growing efforts to miti-gateUHI (Tapper et al 2014, Zhao et al 2014).

One aspect of UHIs yet to be comprehensively stu-died is how it varies on a WC. Furthermore, surfaceobservation based WC temperature studies are essen-tially based on daily mean temperature (Bäumer andVogel 2007, Laux and Kunstmann 2008, Fujibe 2010),maximum or minimum temperatures (Simmondsand Keay 1997, Laux and Kunstmann 2008, Kimet al 2009) or daily temperature range (Forster andSolomon 2003, Gong et al 2006, Bäumer andVogel 2007, Laux and Kunstmann 2008, Kimet al 2009, Kim et al 2010), not at specific times of day.Therefore, there are two main questions this paperaims to answer, once WCs have been identified asexpected for at least the largest cities: How does thesurface temperatureWCof Australianmajor cities dif-fer with time of day? Is the UHI intensity of a majorcity,Melbourne, affected on aWC?

2.Data andmethod

In order to investigate the temperature WCs fordifferent times of day, three-hourly surface (2 m)observation (dating as far back as 1943) data from theAustralian Bureau of Meteorology are utilised. Thesegive eight separate local times of day from which theWCs are analysed from 0000 to 2100 h. We use datafrom seven major Australian cities (see table S2)including Melbourne, which is also the focus of theUHI analysis, together with three additional sites torepresent the surrounding rural area. The studyutilises data on vehicular traffic that is a major sourceof anthropogenic heat release and generally acts as a

2

Environ. Res. Lett. 11 (2016) 074003

good proxy for indicating when peak times occur withregard to anthropogenic activity. To test whetherWCsare statistically significant, we use aMonte Carlo basedmethod (Earl et al 2015).

2.1. Station temperature dataThemeteorological stations are located at sites that arenear the centres of the urban areas, and hencesignificantly affected by anthropogenic activities (tableS2). Included in the analysis is every state capital, aswell as the reasonably densely populated regionalcentre of Cairns in Queensland, due to its long recordand tropical location. Territory capitals will not beanalysed due to Canberra’s station being moved manytimes and Darwin’s station being located too far fromthe urban area.

2.2.MelbourneUHI station dataThe three non-urban sites, to be compared with theMelbourne CBD site, include the stations located atMelbourne Airport, Moorabbin Airport and LavertonAirport (table S3). We use the same period of1973–2013 for the analysis, as this is an epoch when allthree non-urban sites have good data coverage. Thesesites were used by Morris and Simmonds (2000) intheir study of Melbourne’s UHI, and they providefurther site description details. These sites are locatedbetween 17 and 23 km away, in different directionsfrom Melbourne’s centre (see figure 3) and are atsimilar altitudes. We did not average the non-urbansites, because of the varying microclimatic effects ateach specific location; some important signals wouldlikely be smoothed out. Some sites are ‘more urban’than others, depending on wind direction and are notideal locations for rural representation, falling underthe category of ‘urban affected’ (see Lowry 1977), butare the best available.

2.3. Vehicular traffic peak timesTraffic data provided by Vic Roads (www.vicroads.gov.au) for a typical junction located approximately1 km from the CBD of the major Australian city ofMelbourne are utilised.We use the traffic volume data(every 15 min total) from October 2014, which can beconsidered as a representativemonth.

2.4. Statistical testingThe temperature data are run through the dailydivergence from a 31 d running mean (Bäumer andVogel 2007, Earl et al 2015) filter to remove the effectsof interannual and intraannual cycles, with each timeof day treated as a separate time series.We use aMonteCarlo test based on the range between the days of theweek with the minimum and maximum temperatureanomalies. We randomly remove 5% of the data (run10 000 times), retaining the order of the remainingcomponents of the time series and therefore effectivelyretaining the autocorrelation of the data, which is

important when analysing temperature data due to thememory, meaning that independence cannot beassumed. The WC temperature anomalies are calcu-lated and the maximum-to-minimum range takenfrom each simulated WC (WCR), creating a MonteCarlo based PDF (of range in temperature) (Earlet al 2015). To test whether the weekends differ fromthe weekdays, we take the average Monday–Fridaytemperature minus the average Saturday–Sundaytemperature from the original data and compare thatto each of the 10 000 Monte Carlo based WCs (WD-WE). This is also done with mean Monday–Saturdayminus Sunday alone (WD-S). Mondays at 0000, 0300and 0600 h are considered as representing Sundaynights for this statistical test due to being influenced bythe (lack of) Sunday night urban activity.

3. Results and discussion

InAustralian cities (and in developed nations through-out the world), there tends to be two peaks whenmostpeople travel to and fromwork, take children to schooletc. These occur during working weekdays (Mondayto Friday) in the morning (typically running from7–9 am) and there is a less pronounced afternoon/evening peak (4–7 pm). At the weekend these peaks donot occur, so there is a large difference in anthropo-genic activity between weekday and weekend activityat these times, especially in the morning. Traditionallythere is less total traffic seen in cities likeMelbourne onSaturdays and especially Sundays (Simmonds andKeay 1997). During these times on weekdays otherhuman activities also occur in city centres, such as theuse of air conditioning units or heating systems, andfurther afield with electricity generation and biomassburning (Earl et al 2015), producing additional wasteheat and aerosols. In western cultures, with Saturdaysand Sundays being days off work for a significantproportion of the population, Friday and Saturdayevenings have become popular with people going torestaurants, pubs, theatres etc. This means that thereare anthropogenic activities in the city centres on theseevenings through to the early hours of the followingday. This activity does not include typical industrialactivities seen during weekday daytimes, e.g. biomassburning, but increased usage of patio heaters or airconditioning units and perhaps even metabolic heatfrom the increased number of people in the area (QuahandRoth 2012).

3.1. Typical traffic volumesFigure 1 displays the traffic volume of a typical roadjunction near the Melbourne CBD. A similar patternwas found on a suburban arterial road 10 km southeastof the CBD (www.vicroads.gov.au), indicating thatthis activity is not confined to inner Melbourne. Ithighlights the difference between the weekdays andweekend (figure 1(a)) showing the general nocturnal

3

Environ. Res. Lett. 11 (2016) 074003

http://www.vicroads.gov.auhttp://www.vicroads.gov.auhttp://www.vicroads.gov.au

increase and morning reduction at weekends compared with the weekdays. There is a distinct morningand afternoon peak with the weekend showing a lessextreme single peak during the middle of the day.When comparing the averaged Monday–Thursdayand Friday, Saturday and Sunday volumes (1b), theFriday and Saturday night activity is highlighted. Theweekend nocturnal increase is comparable to theweekday evening peak hour volume. Friday followsthe weekday (Monday–Thursday) pattern until 2100 hwhere activity is increased into Saturday night, similarto Saturday evening and Sunday night. Figure 1(c)displays the weekend dip in traffic volume, especiallySunday, with Friday being a weekday and weekendnight, possessing the most traffic. Monday has slightlyless total traffic than Saturday, due to the high volumesin the early hours of Saturday and the fact thatMondayhas slightly lower traffic volumes throughout the daythan the other weekdays (not shown).

3.2.Morning peak timeIn the diurnal cycle of an average day, the morningpeak occurs the time at which the atmosphere is in thetransition between nocturnal and diurnal conditions,which means that the incoming solar radiation is onlybeginning to warm the surface and start the processesof mixing in the lower troposphere (Oke 1982). At thistime there will often be a more stable surface layer oreven a temperature inversion relatively close to thesurface, whichmay act to trap the anthropogenic waste

heat, which will build up more than later in the day,especially during winter in certain stable weather types(Morris and Simmonds 2000). Also, this is the timewhen the weekday to weekend anthropogenic activitydifferential is at its maximum. This means that wewould expect a strong WC of surface temperatureduring themorning peak, especially in the larger cities.This time provides the clearest insight into the short-term effects of anthropogenic activity.

Figure 2 shows that Australia’s largest cities havelower 0900 h temperatures on the weekend, with a sig-nal well above the statistical noise level. The strength ofthe Melbourne and Sydney signals is surprisingbecause they include all times of the year and allsynoptic weather types, not just those which arefavourable. Both cities have signals that are stronglystatistically significant both in the WCR and WD-WE(figure 2(a)). The Melbourne signal displays a WCsinusoidal pattern, peaking on Thursday with a Sun-day minimum. Sydney has a less coherent pattern,peaking on Friday, with a Sunday minimum. Bris-bane’s old monitoring site (1955–1972) also displays astatistically significantWD-WE, but not forWCR, andnot as strongly asMelbourne’s probably because of sitecharacteristics (it is located in the City Botanic Gar-dens, close to the river and further from the source ofheat, though still clearly affected by the morninganthropogenic activity). Adelaide does not display astrongly statistically significant signal, except for whenSunday alone is compared to weekdays (WD-S) as

Figure 1.Mean four-week, 1st October–28thOctober 2014, traffic volume at a typical junction in centralMelbourne (300 m frommeteorological station). (a)Average weekday, weekend and overall (per 15 min). (b)AverageMonday–Thursday volumeswith Friday,Saturday and Sunday (per 15 min). (c)Total vehicles (per day).

4

Environ. Res. Lett. 11 (2016) 074003

Figure 2. (a)Major Australian city 0900 h temperature weekly cycles.WCR test p-value (left) andWD-WE (right). (b)Melbournemaximumandminimum temperatureweekly cyclesWCR test p-value (left) andWD-WE (right),WE considered as Sunday andMonday forminimum temperature. (c)Aswith (a) but for 1800 h.

5

Environ. Res. Lett. 11 (2016) 074003

discussed below. The signals in the other cities did nothave statistically significant WCs at 0900 h, probablybecause these are not large or busy enough to producea signal above the statistical noise level. Fujibe (2010)also found this in daily mean temperature minimumfor major Japanese cities. However, these Australianpatterns are different from German cities, studied byBäumer and Vogel (2007) and Laux and Kunstmann(2008), most of which had aWednesday (or Thursday)peak and Saturday minimum. These studies are notsufficiently similar to ours for any robust comparisondue to them using mean temperature rather than0900 h temperature.

The Melbourne 0900 h signal is stronger statisti-cally than any of the Melbourne maximum and mini-mum seasonal temperature analyses carried out bySimmonds and Keay (1997). The 0900 h signal is alsostronger than the updated Melbourne maximum andminimum temperature WCs (figure 2(b)). This dis-plays the Monday–Friday working week, which has asimilar pattern to the 0900 h, though not as amplifiedor as strong statistically and with a Friday maximum.The minimum Melbourne temperature WC(figure 2(b)) is very different from the 0900 h, with aclear Sunday and especially Monday minimum, withagain smaller amplitude and is less strong statistically.The Sunday night (Monday morning) minimum isapparent (discussed below).

3.3. Evening peak timeIn summer, especially in temperate cites, the weekdayevening peak time occurs before or soon after sunset,so urban areas have a positive convective sensible heatflux with significant mixing occurring in the unstablelower atmosphere (Oke 1982). During winter how-ever, the Sun has already set at this time and nocturnalprocesses have begun. Like the morning, the increasedvolume of traffic (figure 1) and relatively stable atmo-spheric conditions, waste heat and aerosols can buildup. This means that we would expect to see a WCduring the evening peak, though not as pronounced asfor themorning.

Australia’s two largest cities display lower tem-peratures on the weekend than during the week, at1800 h (figure 2(c)), statistically significant for WD-WE, but not as strong as for 0900 h and the WCR wasnot significant. TheMelbourne signal is not sinusoidalas for the morning, though does display a Thursdaypeak and Sunday minimum. The 1800 h temperaturedoes not correspond particularly closely with max-imum temperature as seen forMelbourne (figures 2(b)and (c)), though is likely to be more similar duringsummer, with the maximum temperature oftenoccurring around this time. The Sydney 1800 h alsoshow a Fridaymaximum and Sundayminimum, simi-lar to the 0900 h results. The other cities did not dis-play statistically significantWCs at this time due to the

daytime processes causing a great deal ofmixing henceinhibiting the build-up of heat in these smaller cities.

3.4. Night life effectAnthropogenic activity during the night that createsany waste heat is more likely to produce a WC signaldue to the more stable lower atmosphere during thenight, especially on calm nights. This means that wewould expect a Saturday and Sunday 0000 h and0300 h maxima rather than the minima seen at 0900and 1800 h.

There is a statistically significant 0000 h WC inMelbourne and the new Brisbane site in the WCR test(table 1). TheWE-WD test is also significant but at theend of the other tail (compared to 0900 and 1800 hanalysis), weekends are warmer than weekdays asexpected. Sydney does not display a Saturday and Sun-day maxima, the reason for which is unclear, howeverwe can speculate that the location of the Sydney site ison relatively high ground, especially compared to theadjacent highway, so perhaps the site is relatively unaf-fected by the nocturnal anthropogenic heat island andis also well ventilated. Table 1 indicates that the 0300 htemperatures do not show strong statistically sig-nificant patterns seen at 0000 h for any location, otherthan Melbourne when Sunday is compared to the restof theweek.

All cities have a noticeableMondaymorning (Sun-day night) minimum, statistically significant in theWD-S test for Sydney, Melbourne and Brisbane’s newsite, and close to significance in Perth (old site),Hobart and Cairns as shown in table 1. This may bedue to aerosol loading during the week, with Sundaynight being the longest time (over two days) since thelast aerosol-producing weekday. Thismeans that thesecity sites may be relatively clean, cloud free and betterventilated on Monday morning, allowing for moreefficient cooling (e.g. Cerveny and Balling 2005).Simmonds and Keay (1997) found a sharp Sundaydrop in aerosols in Melbourne and other studies havefound a similar pattern (Cerveney and Balling 1998,Kim et al 2009), so this could be the causalmechanism.Kim et al (2009) found for Korean cities that there wasoften a Tuesday/Wednesday minimum in minimumtemperature and a Saturday peak, whichmay occur fora similar reason, with urban activities occurring onSaturday night, though the authors link the peak toincreased cloud cover from aerosol interaction. Itcould also be due to less Sunday urban activity asfigure 1 shows with fewer vehicles on the roadsthroughout the day, meaning less build-up of wasteheat. This pattern is not seen in Europe for minimumtemperature (Laux andKunstmann 2008), though thisdoes not always correspond to 0000 h temperature.

3.5.Other times of the dayThe table 1 column for Melbourne indicates that thesite displays a WC in temperature at all times of day

6

Environ. Res. Lett. 11 (2016) 074003

Table 1. Summary of statisticalWC testsa.

Time Sydney (1955–2013) Melbourne (1955–2013) Brisbane (1955–1972) Brisbane (2000–2013) Perth (1964–1991) Perth (1995–2014) Adelaide (1979–2013) Hobart (1995–2014) Cairns (1979–2013)

Weekly cycle range

0000 ** SaM ** SaM

0300

0600 *WSa

0900 *** F Su ***Th Su *Th Su *Th F

1200 **MoTh

1500 ***MTh **ThF

1800 *Th Su **MTh *Th F

2100 *Th Su

Weekdayminusweekend

0000 −−− −−− +0300 − −0600 ++ +0900 +++ +++ +++ + +1200 − +1500 + − −1800 +++ +++ +2100 ++ ++ +Monday–Saturdayminus Sunday

0000 ++ +++ + ++ + + +0300 + ++ +0600 ++ ++0900 +++ +++ ++ ++1200 + −− ++1500 −− ++ ++1800 ++ +++ + −− ++2100 + +++ ++ −− + +

a Tests include WCR, WD-WE and WD-S (see section 2.4) for each city (in order of population) station. * Indicates significance in the WCR test and the day of maximum (left) and minimum (right), + indicates a significant positivedifference and—a significant negative difference. The number of *,+ or− shows significance level, one is to the 10% level, two the 5% level and three to the 1% level.

7

Environ.R

es.Lett.11(2016)074003

except at mid-afternoon 1200 and 1500 h. This is to beexpected as this is the time when the atmosphere is atits most energetic and maximum mixing occurs(Oke 1982), not allowing near surface heat build-upfrom anthropogenic activities. The afternoon is thetime when the anthropogenic activity weekday toweekend difference is at a minimum (figure 1). Also,weekday build-up of aerosols during themorningmaybe reducing the amount of shortwave irradiancereaching the surface (e.g. Georgoulias et al 2015)reducing the effect of any increased waste heat. Thetimes of day that Sydney exhibits a WC are 0900, 1800and 0000 h as discussed above. With the site relativelyexposed, as mentioned, any signal from other times ofday may be overcome by advection of heat out of thearea, alongwith the anthropogenic activity beingmoresimilar at these times on weekends and weekdays. TheBrisbane sites have differing WCs, with the new sitehaving a strong WD-WE negative anomaly at 0000 hand the old site a strong positive anomaly at 0900 h asmentioned above. This is likely to be due to a change inlifestyle and Brisbane’s weekend nocturnal anthropo-genic activity, with Brisbane’s population more thandoubling during the two time periods (table S1) andmany parts of the inner city becoming gentrified,leading to more nocturnal weekend activity and lessweekday industrial activity. It could also be due to thechange in location to a site closer to anthropogenicactivity and further from the moderating effects of theBrisbane River. The original Perth station does notshow any significant WC signal though it has a weakpositive WD-WE and WD-S at 0900 h and 0000 hrespectively. The new site has a very different WCpattern (also from all other Australian urban stations)with Thursday minima and Monday maxima in theafternoon and evening, along with negative Sundayanomalies at these times. This is due to the sites havingvery disparate source areas, with the old site located tothe west of the city centre, upstream of the urbancentre from the prevailing south-west winds and thenew site located north-east of the urban centre, next toa golf course, which is often irrigated at mandatedtimes and days (www.watercorporation.com.au). Thishighlights the dangers of interpreting long-termmicro-scale data when site changes are involved(Stewart and Oke 2012), especially with variable sitecharacteristics. Adelaide and Hobart only have a WCsignal forWD-S, indicating that Sunday is the only daywhere anthropogenic activity in these urban centres islow enough to produce a signal. The only significantWC signal in Cairns is a Thursday maximum andFriday minimum for the WCR test. It is unclear whythis occurs.

These results show, for the first time, how there aretemperature WCs that occur at different times of dayand that vary in character. This is highly significantbecause understanding the temperature character-istics of the urban environment is becoming more

important due to the rise in global urban populationand heat-health implications.

3.6.MelbourneUHI WCMelbourne is a growing metropolis and has increasedin population by 65% from 1973 to 2014 (table S1). Ithas aWC for almost all times of the day (table 1), so wecan expect the UHI to also display a WC, assumingthat the surrounding rural areas do not display a WCas strongly.

The Melbourne UHI peaks between 2100 and0000 h and is at its minimum between 0900 and1500 h, depending on the day of the week and whichlocation is used to represent the rural environment(figure 3). This is generally similar to the typical UHIdiurnal cycle described by Oke (1982) with the Mel-bourne CBD—Laverton anomaly on weekdays themost similar to Oke’s cycle. The weekend 0900 h dip isthe main contrast with the classic diurnal UHI cycle,but is in agreement with the result of Theeuwes et al(2015), and is likely to be masked by averaging in non-WC UHI studies. The weekend CBD 0900 h anoma-lies compared to all rural sites contrast strongly withthe other days of the week, the extent of which is high-lighted by the divergence from the mean graphs infigure 3. The 0900 h WCs are all highly significant forboth the WCR and WD-WE tests. This is of no sur-prise when we consider the weekday to weekend dif-ference in the morning peak in vehicular activity(figure 1), which is likely to be less extreme in themorerural areas. UHI are often associated with the changesin energy budgets of the surface materials, howeverthis confirms that the anthropogenic activity withinthese areas also makes an important contribution,with the reduction of the UHI especially at 0900 h onSunday. The Sunday 0900 h minimum in the CBD isso extreme that there is a slight urban ‘cool island’when compared to Moorabbin airport. This may bepartly due to the fact that theMoorabbin site has someurban characteristics compared with Laverton andespecially Tullamarine. The 1800 h UHI is not asextreme as the 0900 h one but is also statistically sig-nificant at each site for theWD-WE test but not for theWCR. This is to be expected since the anthropogenicactivity weekday toweekend evening peak difference isnot as extreme as during the morning. The Saturdayand Sunday 0000 h anomalies have the opposite signfor the WD-WE test, with the weekend morning (Fri-day and Saturday nights) UHI anomaly significantlyhigher than during the rest of the week as expected.The line graphs display how the Saturday and SundayUHI anomalies transition from being the highest at0000 h, to the lowest by 0900 h when compared toeach location. This is consistent with the change intraffic volumes (figure 1) representing anthropogenicactivity, also displaying the extreme weekday versusweekend anomalies being at 0900 h.

8

Environ. Res. Lett. 11 (2016) 074003

http://www.watercorporation.com.au

These results show that theUHI of amajor city canhave a WC, especially at 0900 h, but also at 1800 and0000 h, while it is not significant at other times of day(not shown). This result gives us an indication of theinfluence of anthropogenic activity in enhancing theUHI significantly, confounding the effect of the urbansurface characteristics that produce the UHI by chan-ging the local energy budgets.

4. Conclusions

These results show, for the first time, that the WC oftemperature in major cities differs according to thetime of day and that theUHI intensity of amajor city isaffected on a WC. This has profound implicationsduring extreme heatwaves, showing that humans havethe capability to manage anthropogenic heat flux bychanging the activities to the characteristics of aweekend day, and vice versa for the night. With thenumber and severity of heatwaves set to increase(Cowan et al 2014), this could have significant implica-tions for the health of urban populations. Recent workfor Melbourne (Nicholls et al 2008, Tapper et al 2014)has shown distinct threshold temperatures abovewhich human mortality greatly increases. Advancewarning of extreme heat events provides opportunities

to better manage (e.g., incentives for drivers to avoidusing their cars at the warmest part of the day) humanactivity that will add to the heat. Australian govern-ment concern with the need to create cooler, greenerandmore liveable cities in the face of climate change isevidenced in the recent (January 2016) ministerialstatement on cities. Clearly the results reported hereprovide further important information that can con-tribute to this push for healthier urban environmentsin the face of amore extreme climate.

Introducing the concepts ofWCs at different timesof day and identifying a WC in the UHI were the keyaims of the study. Planned future workwill address theimportant questions of seasonality, weather typing(e.g. Morris and Simmonds 2000), normalising theSunrise and sunset times (e.g. Fortuniak et al 2006)and the impact of windspeed (e.g. Morris et al 2001) totheWC signals.

Acknowledgments

Parts of this research were made possible by fundingfrom the Australian Research Council (Project Num-ber DP130103562). Temperature data can be down-loaded from the Bureau ofMeteorology (http://www.bom.gov.au/climate/data-services/station-data.

Figure 3.Temperature anomalies from theMelbourne site (blue dot in the centre of themap)minus the three airport site (red dots)temperature weekly cycles with the test p-value for (left) theWCR and (right)WD-WE. The top line graphs represent the diurnaltemperature anomalies for each day. The lower graphs plot the divergence from themean for each observation time.

9

Environ. Res. Lett. 11 (2016) 074003

http://bom.gov.au/climate/data-services/station-data.shtmlhttp://bom.gov.au/climate/data-services/station-data.shtml

shtml) and traffic data fromVic Roads (www.vicroads.gov.au).

References

Ashworth J R 1929 Influence of smoke and hot gases from factorychimneys on rainfallQuart. J. R.Meteorol. Soc. 55 341–50

Barmet P, Kuster T,Muhlbauer A and LohmannU2009Weeklycycle in particulatematter versus weekly cycle in precipitationover Switzerland J. Geophys. Res. 114D05206

BäumerD andVogel B 2007Anunexpected pattern of distinctweekly periodicities in climatological variables inGermanyGeophys. Res. Lett. 34 L03819

Bell T L andRosenfeldD 2008Comment on ‘weekly precipitationcycles? Lack of evidence fromUnited States surface stations’byDMSchultz et al Geophys. Res. Lett. 35 L09803

Bell T L, RosenfeldD andKimK-M2009Weekly cycle of lightning:evidence of storm invigoration by pollutionGeophys. Res.Lett. 36 L23805

Bell T L, RosenfeldD, KimK-M, Yoo J-M, LeeM-I andHahnenbergerM2008Midweek increase inUS summer rainand stormheights suggests air pollution invigoratesrainstorms J. Geophys. Res. 113D02209

Cerveny R S andBalling RC Jr 1998Weekly cycles of air pollutantsprecipitation and tropical cyclones in the coastal NWAtlanticregionNature 394 561–3

Cerveny R S andBalling RC 2005Variations in the diurnalcharacter of tropical cyclonewind speedsGeophys. Res. Lett.32 L06706

CowanT, PurichA, Perkins S, PezzaA, Boschat G and Sadler K 2014More frequent longer and hotter heat waves for Australia inthe twenty-first century J. Clim. 27 5851–71

Daniel J S, PortmannRW, Solomon S andMurphyDM2012Identifyingweekly cycles inmeteorological variables: theimportance of an appropriate statistical analysis J. Geophys.Res. 117D13203

DeLisiMP, CopeAMand Franklin J K 2001Weekly precipitationcycles along the northeast corridor?Weather Forecast. 16343–53

EarlN, Simmonds I andTapperN2015Weekly cycles of globalfires—associations with religionwealth and culture and insightsinto anthropogenic influences on global climateGeophys. Res.Lett. 42 9579–89

FariasWR, PintoO, Pinto I R andNaccaratoKP 2014Theinfluence of urban effect on lightning activity: evidence ofweekly cycleAtmos. Res. 135 370–3

Forster PMF and Solomon S 2003Observations of a ‘weekendeffect’ in diurnal temperature range Proc. Natl Acad. Sci. USA100 11225–30

FortuniakK, Kłysik K andWibig J 2006Urban–rural contrasts ofmeteorological parameters in ŁódźTheor. App. Climatol. 8491–101

Fujibe F 1987Weekday–weekend differences of urban climates: I.Temporal variation of air temperature and othermeteorological parameters in the central part of TokyoJ.Meteorol. Soc. Japan 65 923–9

Fujibe F 2010Day-of-the-week variations of urban temperature andtheir long-term trends in JapanTheor. App. Clim. 102393–401

Georgoulias AK,Kourtidis KA, Alexandri G, Rapsomanikis S andSanchez-LorenzoA 2015Common summertime total cloudcover and aerosol optical depthweekly variabilities overEurope: sign of the aerosol indirect effects?Atmos. Res. 15359–73

GongDY,GuoD andHoCH2006Weekend effect in diurnaltemperature range inChina: opposite signals betweenwinterand summer J. Geophys. Res. 111D18113

GongDY,HoCH,ChenD,QianY, Choi Y S andKim J 2007Weekly cycle of aerosol-meteorology interaction over ChinaJ. Geophys. Res. 112D22202

GruzdevAN2013Analysis of the weekly cycle in the atmospherenearMoscow Izv. Atmos. Ocean Phys. 49 137–47

Hendricks FranssenH J, Kuster T, Barmet P and LohmannU2009Comment on ‘winter ‘weekend effect’ in Southern Europeand its connectionwith periodicities in atmosphericdynamics’ by a Sanchez‐Lorenzo et al Geophys. Res. Lett. 36L13706

KimBG, ChoiMHandHoCH2009Weekly periodicities ofmeteorological variables and their possible associationwithaerosols in KoreaAtmos. Environ. 43 6058–65

KimKY, ParkR J, KimKR andNaH2010Weekend effect:anthropogenic or natural?Geophys. Res. Lett. 37 1–6

KolokotroniM,Giannitsaris I andWatkins R 2006The effect of theLondonurban heat island on building summer coolingdemand and night ventilation strategies Sol. Energy 80 383–92

Laux P andKunstmannH2008Detection of regional weeklyweather cycles across Europe Environ. Res. Lett. 3 044005

LowryWP 1977Empirical estimation of urban effects on climate: aproblem analysis J. Appl.Meteorol. 16 129–35

Morris C JG and Simmonds I 2000Associations between varyingmagnitudes of the urban heat island and the synopticclimatology inMelbourne, Australia Int. J. Climatol. 201931–54

Morris C JG, Simmonds I and PlummerN2001Quantification ofthe influences of wind and cloud on the nocturnal urban heatisland of a large city J. Appl.Meteorol. 40 169–82

Nicholls N, Skinner C, LoughnanMandTapperN 2008A simpleheat alert system forMelbourne Australia Int. J. Biomet. 52375–84

OkeTR 1982The energetic basis of the urban heat islandQ. J. R.Meteorol. Soc. 108 455

QuahAKL andRothM2012Diurnal andweekly variation ofanthropogenic heat emissions in a tropical city, SingaporeAtmos. Environ. 46 92–103

RizwanAM,Dennis L Y L and LiuC 2008A review on thegeneration determination andmitigation ofUrbanHeatIsland J. Environ. Sci. 20 120–8

RosenfeldD andBell T L 2011Why do tornados and hailstorms restonweekends? J. Geophys. Res. 116D20211

SailorD J 2011A review ofmethods for estimating anthropogenicheat andmoisture emissions in the urban environment Int. J.Climatol. 31 189–99

Sanchez-LorenzoA, Laux P,Hendricks FranssenH J, Calbó J,Vogl S, Georgoulias AK andQuaas J 2012Assessing large-scaleweekly cycles inmeteorological variables: a reviewAtmos. Chem. Phys. 12 5755–71

Shutters S T andBalling RC 2006Weekly periodicity ofenvironmental variables in Pheonix, ArizonaAtmos. Environ.40 304–10

Simmonds I andKeayK 1997Weekly cycle ofmeteorologicalvariations inMelbourne and the role of pollution andanthropogenic heat releaseAtmos. Environ. 31 1589–603

Sitnov SA 2010Weekly cycle ofmeteorological parameters overmoscow regionDokl. Earth. Sci. 431 507–14

Stewart ID andOkeTR 2012 Local climate zones for urbantemperature studiesBull. Am.Met. Soc. 93 1879–900

StjernCW2011Weekly cycles in precipitation and othermeteorological variables in a polluted region of EuropeAtmos. Chem. Phys. 11 4095–104

TapperN, Coutts A, LoughnanMandPankhaniaD 2014Urbanpopulation vulnerability to climate extremes:mitigatingurban heat through technology andwater-sensitive urbandesign LowCarbonCities: TransformingUrban Systems(Earthscan Book Series on Sustainable Design vol 3) edS Lehmann (London: Routledge) pp 361–74, ch 20

TheeuwesNE, SteeneveldG J, RondaR J, RotachMWandHoltslag AA 2015Cool citymornings by urban heatEnviron.Res. Lett. 10 114022

Zhao L, LeeX, Smith RB andOlesonK 2014 Strong contributions oflocal background climate to urban heat islandsNature 511216–9

10

Environ. Res. Lett. 11 (2016) 074003

http://bom.gov.au/climate/data-services/station-data.shtmlhttp://www.vicroads.gov.auhttp://www.vicroads.gov.auhttp://dx.doi.org/10.1002/qj.49705523205http://dx.doi.org/10.1002/qj.49705523205http://dx.doi.org/10.1002/qj.49705523205http://dx.doi.org/10.1029/2008JD011192http://dx.doi.org/10.1029/2006GL028559http://dx.doi.org/10.1029/2007GL033046http://dx.doi.org/10.1029/2009GL040915http://dx.doi.org/10.1029/2007JD008623http://dx.doi.org/10.1038/29043http://dx.doi.org/10.1038/29043http://dx.doi.org/10.1038/29043http://dx.doi.org/10.1029/2004GL021177http://dx.doi.org/10.1175/JCLI-D-14-00092.1http://dx.doi.org/10.1175/JCLI-D-14-00092.1http://dx.doi.org/10.1175/JCLI-D-14-00092.1http://dx.doi.org/10.1029/2012JD017574http://dx.doi.org/10.1175/1520-0434(2001)0162.0.CO;2http://dx.doi.org/10.1175/1520-0434(2001)0162.0.CO;2http://dx.doi.org/10.1175/1520-0434(2001)0162.0.CO;2http://dx.doi.org/10.1175/1520-0434(2001)0162.0.CO;2http://dx.doi.org/10.1002/2015GL066383http://dx.doi.org/10.1002/2015GL066383http://dx.doi.org/10.1002/2015GL066383http://dx.doi.org/10.1016/j.atmosres.2012.09.007http://dx.doi.org/10.1016/j.atmosres.2012.09.007http://dx.doi.org/10.1016/j.atmosres.2012.09.007http://dx.doi.org/10.1073/pnas.2034034100http://dx.doi.org/10.1073/pnas.2034034100http://dx.doi.org/10.1073/pnas.2034034100http://dx.doi.org/10.1007/s00704-005-0147-yhttp://dx.doi.org/10.1007/s00704-005-0147-yhttp://dx.doi.org/10.1007/s00704-005-0147-yhttp://dx.doi.org/10.1007/s00704-005-0147-yhttp://dx.doi.org/10.1007/s00704-010-0266-yhttp://dx.doi.org/10.1007/s00704-010-0266-yhttp://dx.doi.org/10.1007/s00704-010-0266-yhttp://dx.doi.org/10.1007/s00704-010-0266-yhttp://dx.doi.org/10.1016/j.atmosres.2014.07.031http://dx.doi.org/10.1016/j.atmosres.2014.07.031http://dx.doi.org/10.1016/j.atmosres.2014.07.031http://dx.doi.org/10.1016/j.atmosres.2014.07.031http://dx.doi.org/10.1029/2006JD007068http://dx.doi.org/10.1029/2007JD008888http://dx.doi.org/10.1134/s0001433813020096http://dx.doi.org/10.1134/s0001433813020096http://dx.doi.org/10.1134/s0001433813020096http://dx.doi.org/10.1029/2008GL036774http://dx.doi.org/10.1029/2008GL036774http://dx.doi.org/10.1016/j.atmosenv.2009.08.023http://dx.doi.org/10.1016/j.atmosenv.2009.08.023http://dx.doi.org/10.1016/j.atmosenv.2009.08.023http://dx.doi.org/10.1029/2010GL043233http://dx.doi.org/10.1029/2010GL043233http://dx.doi.org/10.1029/2010GL043233http://dx.doi.org/10.1016/j.solener.2005.03.010http://dx.doi.org/10.1016/j.solener.2005.03.010http://dx.doi.org/10.1016/j.solener.2005.03.010http://dx.doi.org/10.1088/1748-9326/3/4/044005http://dx.doi.org/10.1175/1520-0450(1977)0162.0.CO;2http://dx.doi.org/10.1175/1520-0450(1977)0162.0.CO;2http://dx.doi.org/10.1175/1520-0450(1977)0162.0.CO;2http://dx.doi.org/10.1002/1097-0088(200012)20:153.0.CO;2-Dhttp://dx.doi.org/10.1002/1097-0088(200012)20:153.0.CO;2-Dhttp://dx.doi.org/10.1002/1097-0088(200012)20:153.0.CO;2-Dhttp://dx.doi.org/10.1002/1097-0088(200012)20:153.0.CO;2-Dhttp://dx.doi.org/10.1175/1520-0450(2001)0402.0.CO;2http://dx.doi.org/10.1175/1520-0450(2001)0402.0.CO;2http://dx.doi.org/10.1175/1520-0450(2001)0402.0.CO;2http://dx.doi.org/10.1007/s00484-007-0132-5http://dx.doi.org/10.1007/s00484-007-0132-5http://dx.doi.org/10.1007/s00484-007-0132-5http://dx.doi.org/10.1007/s00484-007-0132-5http://dx.doi.org/10.1002/qj.49710845502http://dx.doi.org/10.1016/j.atmosenv.2011.10.015http://dx.doi.org/10.1016/j.atmosenv.2011.10.015http://dx.doi.org/10.1016/j.atmosenv.2011.10.015http://dx.doi.org/10.1016/S1001-0742(08)60019-4http://dx.doi.org/10.1016/S1001-0742(08)60019-4http://dx.doi.org/10.1016/S1001-0742(08)60019-4http://dx.doi.org/10.1029/2011JD016214http://dx.doi.org/10.1002/joc.2106http://dx.doi.org/10.1002/joc.2106http://dx.doi.org/10.1002/joc.2106http://dx.doi.org/10.5194/acp-12-5755-2012http://dx.doi.org/10.5194/acp-12-5755-2012http://dx.doi.org/10.5194/acp-12-5755-2012http://dx.doi.org/10.1016/j.atmosenv.2005.09.037http://dx.doi.org/10.1016/j.atmosenv.2005.09.037http://dx.doi.org/10.1016/j.atmosenv.2005.09.037http://dx.doi.org/10.1016/S1352-2310(96)00344-5http://dx.doi.org/10.1016/S1352-2310(96)00344-5http://dx.doi.org/10.1016/S1352-2310(96)00344-5http://dx.doi.org/10.1134/S1028334X10040203http://dx.doi.org/10.1134/S1028334X10040203http://dx.doi.org/10.1134/S1028334X10040203http://dx.doi.org/10.1175/BAMS-D-11-00019.1http://dx.doi.org/10.1175/BAMS-D-11-00019.1http://dx.doi.org/10.1175/BAMS-D-11-00019.1http://dx.doi.org/10.5194/acp-11-4095-2011http://dx.doi.org/10.5194/acp-11-4095-2011http://dx.doi.org/10.5194/acp-11-4095-2011http://dx.doi.org/10.1088/1748-9326/10/11/114022http://dx.doi.org/10.1038/nature13462http://dx.doi.org/10.1038/nature13462http://dx.doi.org/10.1038/nature13462http://dx.doi.org/10.1038/nature13462

1. Introduction2. Data and method2.1. Station temperature data2.2. Melbourne UHI station data2.3. Vehicular traffic peak times2.4. Statistical testing

3. Results and discussion3.1. Typical traffic volumes3.2. Morning peak time3.3. Evening peak time3.4. Night life effect3.5. Other times of the day3.6. Melbourne UHI WC

4. ConclusionsAcknowledgmentsReferences