1. Introduction
It has been well demonstrated that urbanization leaves a distinctive thermal signature, the urban heat island (UHI). The UHI occurs when an urbanized area, resulting from a modified energy balance, has a warmer temperature than its surroundings. This effect may be up to 10°C warmer, although there is considerable variation due to different local landscapes and meteorological conditions (Oke 1973, 1982; Eliasson 1994; Stewart 2000; Kalnay and Cai 2003; Sakakibara and Owa 2005; Chen et al. 2006; Stewart and Oke 2009; Stewart 2010; Mohsin and Gough 2012; Tam et al. 2015). Historically, the UHI has been detected by examining rural–urban pairs of climate stations (Oke 1973; Mohsin and Gough 2012), but more recently, other methods have emerged including the use of thermal day-to-day variation metrics (Gough 2008; Tam et al. 2015; Gough and Hu 2016; Wu et al. 2017; Anderson et al. 2018; Gough and Shi 2020). In this work, we explore a new thermal metric and its behavior as a function of the nature of urbanization that yields a distinctive peri-urban thermal signature.
Daily temperature is a commonly used climate measure. The precise determination of this measure has been surprisingly challenging. Methods for doing so range from the mean of the minimum and maximum temperature of the day, sampling temperature at different, and what are believed to be key, times of the day, and the averaging of hourly observations (Gough et al. 2020). These considerations have produced a large body of interesting scientific literature (Hartzell 1919; Brooks 1921, Collison and Tabony 1984; Reicosky et al. 1989; Weber 1993; Harris and Pedersen 1995; Gough and Leung 2002; Zeng and Wang 2012; Trewin 2004; Weiss and Hays 2005; Bonacci et al. 2013; Gough and He 2015; Bonacci and Zeljkovic 2018; Bernhardt et al. 2018; Bernhardt and Carleton 2019; Bernhardt 2020; Gough et al. 2020). While the bulk of this literature is focused on the most accurate algorithm for capturing the daily mean temperature including the climate observing window and the potential for bias in the assessment of the climate record with particular application to climate data homogenization (Vincent et al. 2009, 2012; Ma and Guttorp 2013; Wang 2015; Li et al. 2016; Zhou and Wang 2016; Zaknic-Catovic and Gough 2018; Villarini et al. 2017; Bernhardt et al. 2018; Bonacci and Zeljkovic 2018), the differences in methods can also provide insight into the local climate, such as the presence of fog (Gough and He 2015; Bonacci and Zeljkovic 2018; Bernhardt and Carleton 2019).
These subtle variations of temperature on a daily time scale have been used, in addition to fog, to identify freeze–thaw cycles (Ho and Gough 2006), cloud cover (Laidler et al. 2009), sea ice (McGovern and Gough 2015), urbanization (Tam and Gough 2012; Tam et al. 2015; Gough and Hu 2016; Anderson et al. 2018), and coastalization (Gough and Shi 2020). These analyses include diurnal variation (Ho and Gough 2006), spatial variation (McGovern and Gough 2015), the difference between daily temperature calculation methods (Gough and Leung 2002; Gough and He 2015; Bonacci and Zeljkovic 2018; Bernhardt and Carleton 2019), and day-to-day temperature variability (Gough 2008; Tam et al. 2015).
In this work, we introduce a new metric, based on the nature of day-to-day warm and cold transitions, and apply it to urbanization. This new metric is a variant on the day-to-day variability framework introduced by Karl et al. (1995) and further developed by Gough (2008) and Tam et al. (2015). In this framework, the absolute difference from the previous day is used as a more realistic estimate of variability as it is not dependent on the assumption of randomness as is standard deviation. However, in this work, the sign and magnitude of the change from day to day are considered. For a given time period (e.g., month or year), consider the number of times the day-to-day temperature change is positive (warming) as well as the number of times the change is negative (cooling). If the two counts are not the same for a given period and there is no net change in temperature during the time period, this implies the magnitude of the two differs (warming and cooling) and is not the same on average. For example, if during a year the number of warm transitions exceed the number of cold transitions (with no net change of temperature at the end of the year), the magnitude of the average cold change must exceed the magnitude of the average warm change, and the ratio of these changes is the inverse of the ratio of the number of occurrences.
Although a priori we do not make any predictions how this ratio would be affected by urbanization, we examine climate stations from urban and rural sites anticipating that urbanization and other land-use change have distinctive thermal signatures (e.g., Kalnay and Cai 2003; Cao et al. 2018). As a result of our analysis we reintroduce a third landscape category, “peri-urban” (Jones et al. 1990), that has a thermal signature that is distinct from that of rural and urban locations. The term peri-urban is widely used in the social sciences (e.g., Pannell 2006; Sorensen 2016). We define peri-urban as landscapes that form at a transition zone between urban and rural environments. This has also been labeled in the literature as the “wildland–urban interface (WUI)” (e.g., Platt 2010) or the “rural–urban interface (RUI)” (e.g., Hoffmann et al. 2017). The new warm/cold transition ratio for the minimum temperature of the day is distinctively sensitive to peri-urban landscapes, and a mechanism based on a country breeze at the urban–rural interface is proposed.
2. Data and methods
a. Data
Climate data for 28 urban and rural sites from five Canadian provinces whose landscape characteristics have been identified in previous work (Gough and Rosanov 2001; Mohsin and Gough 2012; Allen et al. 2015; Tam et al. 2015; Gough and Sokappadu 2016; Anderson and Gough 2017; Zaknic-Catovic and Gough 2018; Anderson et al. 2018) were analyzed. All data were obtained from the climate data archive housed at Environment and Climate Change Canada. Stations used are listed in Table 1. Figure 1 shows the main urban locations of these stations: Vancouver, British Columbia; Edmonton, Alberta; Toronto, Ontario; Ottawa, Ontario; Montreal, Quebec; and Halifax, Nova Scotia. Because of the reduction of weather/climate stations after 2000, a common period of 1991–2000 was used. The urban Montreal stations (Montreal McGill and Montreal Lafontaine) were an exception, and for them the time period of 1982–91 was used because of the closure of the stations in 1993. A longer time series was used for Toronto (1851–2000) and Pearson Airport (1941–2000) to explore temporal development of the warm/cold transition metric. Toronto population data from 1834 to 2001 are used as a proxy for the progressive spatial urbanization of Toronto. The data were obtained from the Canadian census and the City of Toronto Archives.
List of stations, organized by longitude, from west to east.
Main locations for climate stations: Vancouver, Edmonton, Toronto, Ottawa, Montreal, and Halifax.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
b. Methods
For each year, the difference in temperature between successive days was calculated for each of minimum, maximum, and mean temperature. The average of the magnitude of the positive and negative transitions (ΔT+, ΔT−) are calculated for all three temperature (maximum, minimum, and mean) elements [Eqs. (1) and (2)] where N+ is the number of warm transitions and N− is the number of cold transitions:
The three elements are compared for the three landscape types urban, peri-urban, and rural. The Wilcoxon signed-rank test (McDonald 2014) is used to assess the significance of differences found. Data normality is determined using the Shapiro–Wilk test (Shapiro and Wilk 1965). A longitudinal analysis is done for two Toronto stations, generating a time series spanning from the 1850s to the present for Toronto and 1941 to the present for Toronto Pearson. The 1890s and 1980s for Toronto are examined in detail to explore the seasonality of the new metric.
3. Results
a. Decadal analysis
For most locations and all three temperature variables there is a skewed distribution, indicating that the magnitude of the transition is different among the stations and across the three thermal variables (Table 2). Among the “urban” stations there appears to be bimodal behavior, especially seen in the RΔT for Tmin. A group of stations have values of RΔT well above 1.0 and a second group with values of RΔT at or below 1.0. These are illustrated in Fig. 2. Using a RΔT for Tmin of 1.05 as the threshold, those above are deemed peri-urban (orange diamonds) and those below as urban (blue diamonds). The peri-urban bin includes Toronto Pearson, Buttonville Airport, Trenton, Pierre E. Trudeau Airport, Mirabel Airport, Toronto Island Airport, Ottawa Airport, Vancouver International Airport, Edmonton Airport, Edmonton City Centre, Saint Hubert Airport, Halifax Stanfield Airport, and Shearwater Airport. Those remaining as urban are Toronto, Montreal McGill, Montreal Lafontaine, Ottawa Canadian Department of Agriculture (CDA), Surrey–Newton, Edmonton Stony Plain, and Halifax Citadel. Those grouped as peri-urban are located at the edge of urban areas, often airports, or at a land–water interface (Toronto Island Airport, Vancouver International Airport). The latter two are in close proximity to major bodies of water and likely experiencing some coastalization effect (Gough and Shi 2020) that may be distinct from the peri-urban effect. Those labeled urban are not airports and are located near the center of urban areas. Although the urban and rural bins are normally distributed, the peri-urban bin is not (using the Shapiro–Wilk test). We thus use a Wilcoxon signed-rank test rather than a t test to compare landscape types. This is done by using paired data for each of the major urban areas studied: Vancouver, Edmonton, Toronto, Ottawa, Montreal, and Halifax (Fig. 1). The paired locations are listed in Table 3. These are graphically illustrated in Fig. 3 for Tmin RΔT. The Wilcoxon signed-rank test (Table 4) indicates that peri-urban is statistically different from both urban and rural locations (p < 0.05) as a visual inspection of Fig. 3 clearly illustrates with the peri-urban sites having the largest values at all six city locations. This is not true for the comparison between rural and urban using this test.
The magnitude of warm (ΔT+) and cold (ΔT−) transitions for minimum, maximum, and mean temperature of the day for the urban, peri-urban, and rural stations; RΔT is the ratio of the warm transitions to cold transitions.
Bifurcating urban stations using the annual Tmin RΔT threshold of 1.05. Those above 1.05 are designated as peri-urban and those below as urban. The following color scheme is used: peri-urban (orange), urban (blue), and rural (red).
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
List of paired locations for the six city locations used in the Wilcoxon signed-rank test.
Paired urban–peri-urban–rural Tmin RΔT for Vancouver, Edmonton, Ottawa, Halifax, Toronto, and Montreal (labels 1–6, respectively, on the x axis). Peri-urban locations are indicated by orange squares, urban locations are indicated by blue diamonds, and rural locations are indicated by red triangles.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
Average values of ΔT+, ΔT−, and RΔT for Tmax, Tmin, and Tmean for urban, peri-urban, and rural bins. Superscripts indicate statistical significance (p < 0.05) using a two sided Wilcoxon signed-rank test.
b. Longitudinal analysis
The Toronto climate record, observed at the University of Toronto, is the longest in the country of Canada. The record begins in 1841. We use data for the 150-yr period spanning from 1851 to 2000. As the ratio (RΔT) of positive to negative temperature transitions for minimum temperature proved to be the most sensitive to variation in landscape (Table 4, Fig. 3), it is plotted as decadal averages for the 150-yr period in Fig. 4. The plot illustrates that Toronto in the 1850s and 1860s had a ratio a slightly lower than rural sites found in the previous section. However, it is the time of the largest magnitude of warm and cold transitions, consistent with a rural landscape (Table 4). At the time of its creation (1821) the University of Toronto was provided land at the northern fringe of the developing settlement. Figure 5 shows the population of Toronto as a function of time (the 1850s span from 1851 to 1860, etc.) that we use as a proxy for the expanding urban sprawl of the city. In the 1860s the population was estimated to be around 60 000. The population grew steadily reaching over 100 000 by 1900 and over 800 000 by 1930. Figure 4 shows that Toronto displayed peri-urban characteristics with a ratio near 1.2 in the 1870s, 1880s and 1890s. After this time, the ratio gradually subsided to values typical of urban landscapes found in the previous section (Table 4).
The ΔT+/ΔT− ratio (RΔT) for decadal averages for minimum temperature for the Toronto climate station from 1851 to 2000.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
Population of Toronto from 1834 to 2001. Data from 1931 to 2001 were taken from the Canadian census. Earlier data were derived from City of Toronto Archives.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
Longitudinal data of a shorter period are available for Toronto Pearson (Fig. 6). The airport was created in a rural area west of Toronto in 1939, and climate data are available from 1940. A similar decadal analysis was done for these data (1941 to 2000). The ratio begins around one for the earliest two decades with consistently large warm and cold transitions typical of a rural landscape. The ratio jumps to over 1.2 in the 1960s with a coincident drop in warm and cold transition magnitudes and remains there for subsequent decades. This transition to peri-urban from rural is consistent with the expanding urban sprawl reaching the Toronto Pearson location (Fig. 5).
The ΔT+/ΔT− ratio (RΔT) for decadal averages for minimum temperature for Pearson Airport from 1940 to 2000.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
c. Physical mechanism
The analysis of the previous two sections introduces a thermal metric that clearly identifies peri-urban landscapes, in both spatial (six urban centers) and temporal analyses (two long-term time series from Toronto stations). The question that remains is why do peri-urban landscapes produce such a striking behavior? Temperature transitions arise from both energy balance and dynamic changes in the atmosphere. On the first point the diurnal cycle of temperature largely arises from the daily energy balance cycle with insolation as a net provider of energy to a location that is modified by surface conditions (albedo, heat storage) and atmosphere conditions (radiatively active gases, cloud, and fog coverage). The atmosphere is a fluid and thermal variations can occur due to shifting air masses and the intermittent presence of midlatitude cyclones, convection storms, and sea/lake breezes that characterize many areas of southern Canada. Given the distinctive changes that appear in peri-urban environments it is most likely that the explanation lies in localized radiative and dynamic features that occur at the boundary of the urban–rural interface. To assist with this, two decades from Toronto are examined in detail, examining seasonal variation in the 1890s (a peri-urban period) and the 1980s (an urban period). Figures 7 and 8 show the difference in the warm and cold transitions (absolute values) for the two different periods for minimum temperature. Generally, the warm transitions are larger in magnitude and cold transitions are smaller in magnitude for the peri-urban period compared to the urban period. The differences were larger for the warm transitions, particularly in winter.
The monthly magnitude of the warm transitions (ΔT+) in the 1890s (blue) and the 1980s (red) for minimum temperature for the Toronto climate station.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
The monthly magnitude of the cold transitions (|ΔT−|) in the 1890s (blue) and the 1980s (red) for minimum temperature for the Toronto climate station.
Citation: Journal of Applied Meteorology and Climatology 59, 9; 10.1175/JAMC-D-19-0292.1
From this analysis, we are able to speculate on what is taking place in the peri-urban environments that can explain the changes particularly with the behavior of minimum temperature. In Laidler et al. (2009), day-to-day temperature considerations led them to conclude that the fall season in Foxe Basin, Nunavut, had experienced more cloud cover in recent years, likely the result of concurrent changing sea ice conditions. Cloud/fog cover could explain the radiative changes experienced in peri-urban landscapes. Cloud cover and the increased absolute humidity associated with it serves to trap outgoing radiation, thus enhancing warm transitions and mitigating cold transitions. The urban heat island is known to increase cloud cover and influence precipitation patterns through the development of a country breeze, also known as an urban breeze (Chou and Zhang 1982; Barlag and Kuttler 1990; Eliasson and Holmer 1990; Hidalgo et al. 2008; Wang et al. 2009; Bourscheidt et al. 2016). The country breeze arises from the horizontal temperature gradient that forms between rural and urbanized landscapes. The less dense air over the urbanized landscape rises and thus draws in air from rural areas that tend to be moister. This breeze has been observed to be stronger at night (Barlag and Kuttler 1990) and this is related to well documented urban ventilation (e.g., Fan et al. 2017; Ren et al. 2018). This type of circulation is ripe for the formation of low-lying stratus clouds. For particularly large urban sprawls this country breeze effect becomes limited in extent and the country breeze does not penetrate to the core of the urbanized landscape, and thus there is a different thermal regime unaffected by country breeze induced cloud cover (Mohsin and Gough 2010). However, if this breeze and the production of cloud and fog were to occur every night the impact on day-to-day variation would be negligible and the Tmin transition from day to day would be minimized, not enhanced. Not every night, though, develops a country breeze, especially when regional winds due to air masses and storm activity dominate the weather. In addition, not every country breeze will induce the instability or lower the dewpoint sufficiently for the production of fog or cloud. It is the accumulated variability of these interplaying mechanisms that enables the Tmin transitions to be larger than rural or urban settings. This is consistent with the behavior at the urban sites of Vancouver, Edmonton, Toronto, Ottawa, Montreal, and Halifax, and it is particularly notable for Toronto and Montreal (Fig. 3), the largest two urban areas among the six. It also explains the peri-urban behavior of Toronto in the 1870s to 1890s when Toronto was developing as a city.
There is one caveat: The stations at Vancouver International Airport and Toronto Island Airport are located in very close proximity to large bodies of water. Although these were categorized as peri-urban, it is possible that an additional marine category needs to be developed consistent with the recent work of Gough and Shi (2020) that examined the impacts of coastalization on day-to-day metrics. These two stations are more likely to be influenced by a sea/lake breeze than a country breeze.
4. Summary
In this work a new thermal metric was introduced: the ratio of the absolute temperature change for daily warm and cold day transitions (RΔT). This was examined for the maximum, minimum, and mean temperatures of the day. The ratio for the minimum temperature proved to be particularly adept at detecting the thermal signature of peri-urban landscapes (landscapes located near the urban–rural interface), which is particularly notable at airports. For these landscapes, the ratio was consistently above 1, indicating that warm transitions were of greater magnitude than cold transitions, which was not evident for urban and rural landscapes. It was speculated that in peri-urban environments there is intermittent cloud cover at night. As peri-urban environments are warmer than surrounding rural areas, a thermal circulation develops as a country or urban breeze. This breeze brings cooler, moister rural air into the peri-urban areas. This moister air rises above the peri-urban area and, on occasion, forms fog or stratus clouds. These clouds trap outgoing longwave radiation, tending to enhance the magnitude of the transition to warmer days and mitigating the magnitude to colder days. An urban sprawl has this effect in the peripheral areas (i.e., peri-urban) and the effect is diminished and disappears at the core of the urban region due to the limited range of the country breeze. This was spatially observed in climate stations in the large urban areas of Toronto and Montreal. It was also illustrated in the time series of Toronto and Pearson data. The 1850s and 1860s for Toronto displayed rural characteristics. This transformed to a ratio above 1 typical of the peri-urban landscapes examined in this work from the 1870s to the 1890s. After this time period, the ratio gradually declined to values typical of urban values (below one). For Pearson, the 1940s and 1950s displayed rural characteristics with a ratio of 1 that increased to peri-urban values for the decades after this.
This new metric is a promising tool in the subtleties of detecting the thermal signature of peri-urban landscapes. This may have particular application to the homogenization of climate datasets. The goal of homogenization is to reduce the climate record to the variations of weather and climate and screen other compounding influences such as changes in instrumentation and calibration, station relocation, and urbanization and other land-cover changes (e.g., Vincent et al. 2012; Lakhraj-Govender et al. 2017). This metric could be useful in identifying the changing nature of a particular station such as spreading urbanization, as clearly seen with the Toronto and Toronto Pearson climate time series.
The characterization of surface climate stations is important for the assessment of the climate record and the detection of climate change. This work suggests that climate stations are fundamentally influenced by their surroundings, particularly the degree of urbanization and proximity to rural environments. Whether a station is rural, urban, or peri-urban and whether this status has changed with time is critically important to the use of such stations to assess climate change and to assess the strength of the urban heat island. It also has application in assessing “urban ventilation,” an important topic in urban planning (Ren et al. 2018). This ventilation allows the penetration of country breezes that provide thermal relief to the center of an urban area. In addition, the urban–rural boundary is inherently of interest, and this metric enables the clear identification of an area that is experiencing a peri-urban climate.
Acknowledgments
The author is supported by NSERC Grant RGPIN-2018-06801.
REFERENCES
Allen, S. M. J., W. A. Gough, and T. Mohsin, 2015: Changes in the frequency of extreme temperatures records in Toronto, Ontario, Canada. Theor. Appl. Climatol., 119, 481–491, https://doi.org/10.1007/s00704-014-1131-1.
Anderson, C. I., and W. A. Gough, 2017: Evolution of winter temperature in Toronto, Ontario, Canada: A case study of the winters 2013/14 and 2014-/5. J. Climate, 30, 5361–5376, https://doi.org/10.1175/JCLI-D-16-0562.1.
Anderson, C. I., W. A. Gough, and T. Mohsin, 2018: Characterization of the urban heat island at Toronto: Revisiting the choice of rural sites using a measure of day-to-day variation. Urban Climate, 25, 187–195, https://doi.org/10.1016/j.uclim.2018.07.002.
Barlag, A.-B., and W. Kuttler, 1990: The significance of country breezes in urban planning. Energy Build., 15, 291–297, https://doi.org/10.1016/0378-7788(90)90001-Y.
Bernhardt, J., 2020: Comparing daily temperature averaging methods: The role of synoptic climatology in determining spatial and seasonal variability. Phys. Geogr., 41, 272–288, https://doi.org/10.1080/02723646.2019.1657332.
Bernhardt, J., and A. M. Carleton, 2019: Comparing daily temperature averaging methods: The role of surface and atmosphere variables in determining spatial and seasonal variability. Theor. Appl. Climatol., 136, 499–512, https://doi.org/10.1007/s00704-018-2504-7.
Bernhardt, J., A. M. Carleton, and C. LaMagna, 2018: A comparison of daily temperature-averaging methods: Spatial variability and recent change for the CONUS. J. Climate, 31, 979–996, https://doi.org/10.1175/JCLI-D-17-0089.1.
Bonacci, O., and I. Zeljkovic, 2018: Differences between true mean temperatures and means calculated with four different approaches: A case study from three Croatian stations. Theor. Appl. Climatol., 131, 733–743, https://doi.org/10.1007/s00704-016-1993-5.
Bonacci, O., I. Zeljkovic, R. Sakic Trogrlic, and J. Milkovic, 2013: Differences between true mean daily, monthly and annual air temperatures and air temperatures calculated with three equations: A case study from three Croatian stations. Theor. Appl. Climatol., 114, 271–279, https://doi.org/10.1007/s00704-012-0830-8.
Bourscheidt, V., O. Pinto Jr., and K. P. Naccarato, 2016: The effects of Sao Paulo urban heat island on lightning activity: Decadal analysis (1999–2009). J. Geophys. Res. Atmos., 121, 4429–4442, https://doi.org/10.1002/2016JD024782.
Brooks, C. E. P., 1921: True mean temperature. Mon. Wea. Rev., 49, 226–229, https://doi.org/10.1175/1520-0493(1921)49<226:TMT>2.0.CO;2.
Cao, Q., D. Yu, M. Georgescu, and J. Wu, 2018: Substantial impacts of landscape changes on summer climate with major regional differences: The case of China. Sci. Total Environ., 625, 416–427, https://doi.org/10.1016/j.scitotenv.2017.12.290.
Chen, X.-L., H.-M. Zhao, P.-X. Li, and Z.-Y. Yin, 2006: Remote sensing image-based analysis of the relationship between urban heat island and land use/cover changes. Remote Sens. Environ., 104, 133–146, https://doi.org/10.1016/j.rse.2005.11.016.
Chou, S. Z., and C. Zhang, 1982: On the Shanghai heat island effect. Acta Geogr. Sin., 37, 372–382.
Collison, P., and R. C. Tabony, 1984: The estimation of mean temperature from daily maxima and minima. Meteor. Mag., 113, 329–337.
Eliasson, I., 1994: Urban-suburban-rural air temperature differences related to street geometry. Phys. Geogr., 15 (1), 1–22, https://doi.org/10.1080/02723646.1994.10642501.
Eliasson, I., and B. Holmer, 1990: Urban heat island in Goteborg, Sweden. Theor. Appl. Climatol., 42, 187–196, https://doi.org/10.1007/BF00866874.
Fan, Y., J. C. R. Hunt, and Y. Li, 2017: Buoyancy and turbulence-driven atmospheric circulation over urban areas. J. Environ. Sci., 59, 63–71, https://doi.org/10.1016/j.jes.2017.01.009.
Gough, W. A., 2008: Theoretical considerations of day-to-day temperature variability applied to Toronto and Calgary, Canada data. Theor. Appl. Climatol., 94, 97–105, https://doi.org/10.1007/s00704-007-0346-9.
Gough, W. A., and Y. Rosanov, 2001: Aspects of Toronto’s climate: Heat island and lake breeze. Can. Meteor. Oceanogr. Soc. Bull., 29, 67–71.
Gough, W. A., and A. Leung, 2002: Nature and fate of Hudson Bay permafrost. Reg. Environ. Change, 2, 177–184, https://doi.org/10.1007/s10113-002-0048-x.
Gough, W. A., and D. He, 2015: Diurnal temperature asymmetries and fog at Churchill, Manitoba. Theor. Appl. Climatol., 121, 113–119, https://doi.org/10.1007/s00704-014-1227-7.
Gough, W. A., and Y. Hu, 2016: Day to day temperature variability for four urban areas in China. Urban Climate, 17, 80–88, https://doi.org/10.1016/j.uclim.2016.06.002.
Gough, W. A., and S. Sokappadu, 2016: Climate context of the cold summer of 2014 in Toronto, ON, Canada. Theor. Appl. Climatol., 126, 183–189, https://doi.org/10.1007/s00704-015-1571-2.
Gough, W. A., and B. Shi, 2020: Impact of coastalization on day-to-day temperature variability along China’s east coast. J. Coast. Res., 36, 451–456, https://doi.org/10.2112/JCOASTRES-D-19-00167.1.
Gough, W. A., A. Zaknic-Catovic, and A. Zajch, 2020: Sampling frequency of climate data for the determination of daily temperature and daily temperature extrema. Int. J. Climatol., https://doi.org/10.1002/joc.6528, in press.
Harris, S. A., and J. H. Pedersen, 1995: Comparison of three methods of calculating air temperature from electronic measurements. Z. Geomorphol., 39, 203–210, https://doi.org/10.1016/S0883-9417(95)80025-5.
Hartzell, F. Z., 1919: Comparison of methods for computing daily mean temperatures: Effect of discrepancies upon investigations of climatologists and biologists. Mon. Wea. Rev., 47, 799–801, https://doi.org/10.1175/1520-0493(1919)47<799:COMFCD>2.0.CO;2.
Hidalgo, J., G. Pigeon, and V. Masson, 2008: Urban-breeze circulation during the CAPITOUL experiment: Observational data analysis approach. Meteor. Atmos. Phys., 102, 223–241, https://doi.org/10.1007/s00703-008-0329-0.
Ho, E., and W. A. Gough, 2006: Freeze thaw cycles in Toronto, Canada, in a changing climate. Theor. Appl. Climatol., 83, 203–210, https://doi.org/10.1007/s00704-005-0167-7.
Hoffmann, E. M., M. Jose, N. Nolke, and T. Mockel, 2017: Construction and use of a simple index of urbanisation in the rural–urban interface of Bangalore, India. Sustainability, 9, 2146, https://doi.org/10.3390/su9112146.
Jones, P. D., P. Ya. Groisman, M. Coughlan, N. Plummer, W. C. Wang, and T. R. Karl, 1990: Assessment of urbanization effects in time series of surface air temperature over land. Nature, 347, 169–172, https://doi.org/10.1038/347169a0.
Kalnay, E., and M. Cai, 2003: Impact of urbanization and land-use change on climate. Nature, 423, 528–531, https://doi.org/10.1038/nature01675.
Karl, T. R., R. W. Knight, and N. Plummer, 1995: Trends in high-frequency climate variability in the twentieth century. Nature, 377, 217–220, https://doi.org/10.1038/377217a0.
Laidler, G. J., J. Ford, W. A. Gough, T. Ikummaq, A. S. Gagnon, and S. Kowal, 2009: Travelling and hunting in a changing Arctic: Assessing Inuit vulnerability to sea ice change in Igloolik, Nunavut. Climatic Change, 94, 363–397, https://doi.org/10.1007/s10584-008-9512-z.
Lakhraj-Govender, R., S. Grab, and N. E. Ndebele, 2017: A homogenized long term temperature record for the Western Cape Province in South Africa: 1916-2013. Int. J. Climatol., 37, 2337–2353, https://doi.org/10.1002/joc.4849.
Li, Z., K. Wang, C. Zhou, and L. Wang, 2016: Modelling the true monthly mean temperature from continuous measurements over global land. Int. J. Climatol., 36, 2103–2110, https://doi.org/10.1002/joc.4445.
Ma, Y., and P. Guttorp, 2013: Estimating daily mean temperature from synoptic climate observations. Int. J. Climatol., 33, 1264–1269, https://doi.org/10.1002/joc.3510.
McDonald, J. H., 2014: Handbook of Biological Statistics. 3rd ed. Sparky House Publishing, 313 pp.
McGovern, P. G., and W. A. Gough, 2015: East-west asymmetry in coastal temperatures of Hudson Bay as a proxy for sea ice. Arctic, 68, 445–452, https://doi.org/10.14430/arctic4522.
Mohsin, T., and W. A. Gough, 2010: Trend analysis of long-term temperature time series in the Greater Toronto Area (GTA). Theor. Appl. Climatol., 101, 311–327, https://doi.org/10.1007/s00704-009-0214-x.
Mohsin, T., and W. A. Gough, 2012: Characterization and estimation of urban heat island at Toronto: Impact of the choice of rural sites. Theor. Appl. Climatol., 108, 105–117, https://doi.org/10.1007/s00704-011-0516-7.
Oke, T. R., 1973: City size and the urban heat island. Atmos. Environ., 7, 769–779, https://doi.org/10.1016/0004-6981(73)90140-6.
Oke, T. R., 1982: The energetic basis of the urban heat island. Quart. J. Roy. Meteor. Soc., 108, 1–24, https://doi.org/10.1002/qj.49710845502.
Pannell, C. W., 2006: Peri-urbanism in globalizing China: A critique. Eurasian Geogr. Econ., 47, 54–57, https://doi.org/10.2747/1538-7216.47.1.54.
Platt, R. V., 2010: The wildland-urban interface: Evaluating the definition effect. J. For., 108, 9–15, https://doi.org/10.1093/jof/108.1.9.
Reicosky, D. C., L. J. Winkelman, J. M. Baker, and D. G. Baker, 1989: Accuracy of hourly air temperatures calculated from daily minima and maxima. Agric. For. Meteor., 46, 193–209, https://doi.org/10.1016/0168-1923(89)90064-6.
Ren, C., and Coauthors, 2018: Creating breathing cities by adopting urban ventilation assessment and wind corridor plan—The implementation in Chinese cities. J. Wind Eng. Ind. Aerodyn., 182, 170–188, https://doi.org/10.1016/j.jweia.2018.09.023.
Sakakibara, Y., and K. Owa, 2005: Urban–rural temperature differences in coastal cities: Influence of rural sites. Int. J. Climatol., 25, 811–820, https://doi.org/10.1002/joc.1180.
Shapiro, S. S., and M. B. Wilk, 1965: An analysis of variance test for normality (complete samples). Biometrika, 52, 591–611, https://doi.org/10.1093/biomet/52.3-4.591.
Sorensen, A., 2016: Periurbanization as the institutionalization of place: The case of Japan. Cities, 53, 134–140, https://doi.org/10.1016/j.cities.2016.03.009.
Stewart, I. D., 2000: Influence of meteorological conditions on the intensity and form of the urban heat island effect in Regina. Can. Geogr., 44, 271–285, https://doi.org/10.1111/j.1541-0064.2000.tb00709.x.
Stewart, I. D., 2010: A systematic review and scientific critique of methodology in modern urban heat island literature. Int. J. Climatol., 31, 200–217, https://doi.org/10.1002/joc.2141.
Stewart, I. D., and T. R. Oke, 2009: Newly developed “Thermal Climate Zones” for defining and measuring urban heat island magnitude in the canopy layer. Eighth Symp. on the Urban Environment, Phoenix, AZ, Amer. Meteor. Soc., J8.2A, https://ams.confex.com/ams/pdfpapers/150476.pdf.
Tam, B. Y., and W. A. Gough, 2012: Examining past temperature variability in Moosonee, Thunder Bay and Toronto, Ontario, Canada through a day-to-day variability framework. Theor. Appl. Climatol., 110, 103–113, https://doi.org/10.1007/s00704-012-0622-1.
Tam, B. Y., W. A. Gough, and T. Mohsin, 2015: The impact of urbanization and the urban heat island effect on day to day temperature variation. Urban Climate, 12, 1–10, https://doi.org/10.1016/j.uclim.2014.12.004.
Trewin, B., 2004: Effects of changes in algorithms used for the calculation of Australian mean temperature. Aust. Meteor. Mag., 53, 1–11.
Villarini, G., A. Khouahki, and E. Cunningham, 2017: On the impacts of computing daily temperatures as the average of the daily minimum and maximum temperatures. Atmos. Res., 198, 145–150, https://doi.org/10.1016/j.atmosres.2017.08.020.
Vincent, L. A., L. W. Xiolan, E. J. Milewska, R. Hopkinson, and R. Malone, 2009: Bias in minimum temperature introduced by a redefinition of the climatological day at the Canadian Stations. J. Appl. Meteor. Climatol., 48, 2160–2168, https://doi.org/10.1175/2009JAMC2191.1.
Vincent, L. A., L. W. Xiolan, E. J. Milewska, H. Wan, F. Yang, and V. Swail, 2012: A second generation of homogenized Canadian monthly surface air temperature for climate trend analysis. J. Geol. Res., 117, D18110, https://doi.org/10.1029/2012JD017859.
Wang, K., 2015: Sampling biases in datasets of historical mean air temperature over land. Sci. Rep., 4, 4637, https://doi.org/10.1038/srep04637.
Wang, X. Q., Z. F. Wang, Y. B. Qi, and H. Guo, 2009: Effect of urbanization on the winter precipitation distribution in Beijing area. Sci. China., 52D, 250–256, https://doi.org/10.1007/s11430-009-0019-x.
Weber, R. O., 1993: Influence of different daily mean formulas on monthly and annual averages on temperature. Theor. Appl. Climatol., 47, 205–213, https://doi.org/10.1007/BF00866241.
Weiss, A., and C. J. Hays, 2005: Calculating daily mean temperatures by different methods: Implications from a non-linear algorithm. Agric. For. Meteor., 128, 57–65, https://doi.org/10.1016/j.agrformet.2004.08.008.
Wu, F.-T., C. Fu, Y. Qian, Y. Gao, and S.-Y. Wang, 2017: High-frequency daily temperature variability in China and its relationship to large-scale circulation. Int. J. Climatol., 37, 570–582, https://doi.org/10.1002/joc.4722.
Žaknić-Ćatović, A., and W. A. Gough, 2018: A comparison of climatological observing windows and their impact on detecting daily temperature extrema. Theor. Appl. Climatol., 32, 41–54, https://doi.org/10.1007/s00704-017-2068-y.
Zeng, X., and A. Wang, 2012: What is monthly mean land surface air temperature? Eos, Trans. Amer. Geophys. Union, 93, 156, https://doi.org/10.1029/2012EO150006.
Zhou, C., and K. Wang, 2016: Spatiotemporal divergence of the warming hiatus over land based on different definitions of mean temperature. Sci. Rep., 6, 31789, https://doi.org/10.1038/srep31789.
