A primary Study on the Dynamic Variation Characteristics of Urban Dew in Guangzhou, China

YE Youhua , ZHOU Kai, PENG Shaolin*

Abstract： Urban dew is one of the important importing water resources, however, information on the dynamic variation characteristics of urban dew is rare by far. The purpose of this paper is to study that through the field observation. Cloth-plate method was used and four typical urban landscapes were employed in Guangzhou city, China. It was found that the frequency of urban dew in April, July, October and January was 43%, 46%, 73% and 62%, respectively. Dew frequency and dew amounts changed from landscape to landscape, and nearly no similar dynamic rule was found. Dew amounts obtained in April (0.026 mm) and July (0.027 mm) were higher than those in October (0.014 mm) and January (0.017 mm). According to the max dew amounts, the highest value was found in January (0.073 mm), followed by July (0.066 mm), April (0.059 mm) and October (0.049 mm). Results indicated that the dynamic variation characteristics of urban dew were impacted by the urban landscapes, different urban landscapes and different collecting time exhibited different dynamic variation characteristics on dew frequency and dew amounts. The average dew amounts showed the order as following： Baiyun Mountain Forest Park>Fangcun Industrial Area>Tianhe Commercial Area>Zhongda Community. The differences of mean dew amounts between Baiyun Mountain Forest Park and other sites arrived at significant level,as well as the differences between Fangcun Industrial Area and Zhongda Community also reached significant levels. Autumn and winter were the possible fitting seasons for dew events to happen, and spring and summer were the possible fitting time for dew amounts to condense. The dew in urban areas is affected by both seasonal variation and functional area.

Key words： urban dew; urban landscapes; dew frequency; dew amounts; dynamic variation

Dew is one of the common climatic phenomena in our surrounding, and the investigation and utilization have been undertaken over 200 years(Stone, 1957a). Though much work has been done before, it is still not familiar to us when comparing to other climatic factors such as temperature,precipitation, wind, etc., for the small quantity and indirect effect on human being. As one of the importing water resource, it was important for plants and microorganisms, and its ecological effects could not be ignored (Stone, 1957a, 1957b, 1963; Kume et al., 2001; Jackson et al., 1999; Csintalan et al. 2000;Kidron et al., 2002; Prado et al., 2007).

Nowadays more and more work on it has being done in semiarid and arid area, mountain and urban area, but these work behaved clear regional characteristics (Zangvil, 1996; Malek et al., 1999;Kidron, 1999; Kidron et al., 2000; Kidron, 2005;Richards, 2005; Ye et al., 2007). Though there is an increasing interest in the study of dew (Richards,2004), reports on the dynamic variation characteristics of dew were quite few (Ashbel, 1949; Yan et al., 2004;Liu et al., 1998; Zhou et al., 2002; Wang et al., 2006),especially in the non-arid city.

Profound changes in urban environment have occurred as a result of rapid urbanization, which has attracted worldwide attention. The condensation of urban dew would be impacted by the urbanization which would lead to a changing dew value. What kind of dynamic variation characteristics of dew will appear under the influence of urban climate? Which element is the main factor responsible for the variation of dew? Is the variation of dew similar among the urban landscapes? Knowledge about the dynamic characteristics is scarce. In order to answer these questions, more work should be done. In present study, the authors tried to analyze the dynamic variation characteristics of dew through field observation in Guangzhou City, China.

1 Studied area and methods

1.1 Studied area

Guangzhou city (112°57′～114°3′E, 22°26′～23°56′N) is a political, economic and cultural center of Guangdong Province, China (Fig.1), with a permanent population of 10.04 millions in 2008 and a total area of 3718.5 km2in urban area (Guangdong Provincial Bureau of Statistics, 2008). It has a subtropical oceanic monsoon climate with an average annual temperature of 21.7～22.9 ℃, and the average annual precipitation ranges from 1385.5 to 1619.5 mm. The city center is about 30 km far from the sea,and Pear River, one of the biggest rivers in China,crosses the city.

1.2 Experimental landscapes and time

Four typical landscapes were employed to study the dynamic variation characteristics of dew in urban area including residential landscapes, commercial landscapes, industrial landscapes and forest landscapes. In the field observation, Zhongda residential landscape (ZRL), Tianhe commercial landscape (TCL), Fangcun industrial landscape (FIL)and Baiyunshan forest landscape (BFL) were chosen as the representative of the above landscapes,respectively. Dew collections were performed on lawn in all landscapes except BFL where the collecting site was on the canopy of the forest. In each landscape,three collectors were set and performed nearly at the same time.

Fig. 1 Location of the studied sitesZRL-Zhongda residential landscape, TCL-Tianhe commercial landscape, FIL-Fangcun industrial landscape, BFL-Baiyunshan forest landscape

In order to study the dynamic variation characteristics of urban dew, four months were chosen including January, April, July and October. Field observations were carried out in July and October of 2005, January and April of 2006. Dew was measured through the whole fine night when no rains or fogs appeared. Therefore, characteristics of temperature and urban heat island in urban landscapes were collected (Table 1).

1.3 Methods

In present study, dew is the condensation of water vapor on a surface, whose temperature is reduced by radiate cooling to below the dew point of clear air in contact with it (WMO, 1966; HMSO,1991), including dewfall and distillation (Monteith,1963).

Dew was measured using cloth-plate (Kidron,1998) in the experiment, and the detailed measuring method of dew and related meteorological factors was described by Ye et al. (2007). The materials used in the experiment included plywood (100 cm×100 cm×0.50 cm, the size of length, width and height, the same meaning as follows), synthetic velvet cloth (100 cm×90 cm×0.15 cm), polyethylene plate (100 cm×100 cm×0.05 cm), polyethylene bottles and stakes with 15 cm height. The collector was made according to the condition and the structure of the collector would not impact on the formation of dew. Though the synthetic velvet cloth could not capture absolute dew amounts at night for some dew escapes, it was an excellence material for its fine moisture absorption characteristic.

The plywood affixed with polyethylene plates(polyethylene plate facing to the sky) were fixed on the stakes before dew collection. Cloths were attached to polyethylene plate at half an hour after sunset(06：00 pm) in the evening and collected at half an hour before sunrise early in the next morning (06：00 am) according to Yan et al. (2004). It’s needed to know that urban dew amount is much small and the time of maximum value appears is unclear, thus dew must be collected before sunrise for reducing dew escape. Dew was collected in different landscapes synchronously. Once collected, the cloths were transferred to separate pre-weighed polyethylene bottles and sealed immediately, then weighed as soon as possible using electronic balance (accuracy is ±0.1 g, Guangzhou Pubo Instrument Co., Ltd., China) in a nearby laboratory. Mass datas were converted to units of mm depth later (accuracy is ±0.001 mm). Because the dew quantity is small, only a thin drip can be seen during the field observation, i.e., dew sample collection is difficult. In present study, the cloths were dried, weighed and sealed in the pre-weighed polyethylene bottles before dew collection, dew quantity could be expressed as：

Where mdis the dew amounts, mtis the total mass of the collection samples including dew amounts, the mass of the cloth and bottle, mcis the mass of the cloth, mbis the mass of the bottle.

Surface air temperature, relative humidity, wind speed and evaporation were investigated and urban heat island was evaluated, and these methods were also described by Ye et al. (2007). Air temperature (t)and relative humidity (RH) were measured using a ZJ1-2A auto-recording hygrothermograph (accuracy is±0.1 ℃ and ±5% RH) (Shanghai Meteorological Instrument Factory Co., Ltd., China), and wind speed(WS) was measured using a portable auto-recording wind detecting instrument (accuracy=0.1 m·s-1)(Shanghai Fengyun Meteorological Instrument Management Factory, China). RH, t and WS were measured at a height of 1.5 m on lawn from 06：00 pm to 06：00 am of the next day. Evaporation (E) was measured using a laminable steel pan in 20 cmdiameter (Shanghai Meteorological Instrument Factory Co., Ltd., China) at 0.7 m height during the whole day.

Table 1 Characteristics of temperature and urban heat island in urban landscapes

Urban heat island (UHI), a very popular phenomenon in urban area, was chosen as a parameter to assess its effect on dew. It was evaluated according to the definition of Kim et al. (2004), and expressed as：

Where UHI was the mean urban heat island, tuwas the mean temperature of each measured urban landscapes, and tcwas the mean temperature of Conghua countryside (source from Meteorological Bureau of Conghua).

The differences in dew frequency, dew amounts among urban landscapes were evaluated by One-way Analysis of Variance (ANOVA) at 0.01 and 0.05 level,respectively. All statistical analyses were performed using SPSS 11.5 software package (SPSS Inc, USA).

2 Results

2.1 Dew frequency

The dynamic variation characteristics of urban dew frequency in ZRL and FIL were the same, but both of them were different from those in TCL and BFL (Fig.2). Though some variances appeared among the landscapes, no significant difference was found among them (P>0.05) based on the mean dew frequency (60%). It was found that the mean frequency of urban dew in January, April, July and October was 62%, 43%, 46% and 73%, respectively.Results indicated that different urban landscapes might lead to different dew events, and autumn and winter were the possible fitting seasons for dew events to happen.

Fig. 2 Frequency of urban dew in different urban landscapesZRL： Zhongda residential landscape; TCL： Tianhe commercial landscape; FIL： Fangcun industrial landscape; BFL： Baiyunshan forest landscape. The same below

2.2 Dew amounts

The mean dew amounts at fine night were measured in each urban landscapes with results presented in Fig. 3. In ZRL and BFL, similar property appeared with the highest mean value found in July,followed by April, January and October, respectively,which was different from those in TCL and FIL. The differences between July and January, July and October, and April and October in ZRL and BFL were significant (P＜0.05), respectively. In commercial landscape (TCL), the dew amounts in these measured times were January＜July＜October＜April, but significant difference was only found between April and January(P＜0.05). In industrial landscape (FIL), the dew quantity presented as the following rule： October＜April＜ January＜July, and significant difference was only found between July and October (P＜0.05).Results showed that urban landscapes had an impact on the dynamic variation of dew quantity, and different landscape might lead to a different changing rule of dew amounts.

Fig. 3 Condensation characteristics of mean dew amounts in different urban landscapesBars represent one standard error. Same letters indicate non-significant differences while different letters indicate significant differences among landscapes. The same below

Dynamic characteristic of mean dew amounts in urban landscapes were shown in Fig. 4. From which,April (0.026 mm) and July (0.027 mm) obtained a significantly higher value than those in October (0.014 mm) and January (0.017 mm) respectively (P＜0.05).Results indicated that spring and summer was the possible fitting time for dew amounts to condense.

Fig. 4 Dynamic characteristic of mean dew amounts in seasonsThe number of collecting samples in January, April, July, October is 79, 53, 57 and 90, respectively.

The max dew amounts of each collecting time were analyzed, and results showed that the max dew amounts changed from landscape to landscape and from month to month. The dynamic variation characteristics of max dew amounts in urban area was：October (0.049 mm)＜April (0.059 mm)＜July (0.066 mm)＜January (0.073 mm).

Mean dew amounts of the sample collecting time in urban landscapes were showed in Fig. 5. Among the four landscapes, an increasing rule was observed from ZAL to BFL. It was found that BFL obtained the highest value among the four landscapes, which was significantly higher than those in ZRL, TCL and FIL(P＜0.05). It could also be seen that there was a significant difference in mean value between ZRL and FIL (P＜0.05).

Fig. 5 Mean dew amounts of the sample collecting time in urban landscapes

3 Discussions

During the study, the values of dew frequency and dew amounts showed relatively big changes among the measured urban landscapes and the collecting time, even significant differences occurred among them. The changes, presumably, reflected the dynamics of dew over the study period with a result of the influence of urban landscapes and the urban micro-climate.

In most of the previous studies, the investigation time of dew was short such as one month or even several days (Chiwa et al., 2003; Camuffo et al.,2003), thus reports on dew frequency was few(Ashbel, 1949; Yan et al., 2004). There was no related information on the dynamic variation of dew frequency in urban area so far. Some experts have done the work on the seasonal variation of dew frequency outside the city and their work suggested that the seasonal variation of dew would be different in different area (Ashbel, 1949). The dynamic variation of dew frequency in present study differed from the previous results, the difference of dew frequency in each landscape might be related to the region's general pattern of weather conditions (Liu et al., 1998), which would result in the difference of precipitation, wind, fog, frost, ice, etc., and consequently lead to a different rule of dew frequency.

Due to the fact that the four urban landscapes in present study belonged to the same weather pattern,the difference in dew frequency among them might be ascribed to their micro-environment, as well as the local shower. The impact of micro-environment on dew formation was referred in some reports, but less report confirmed which factor would make the dew formation absence except the high temperature or urban heat island (Mattsson, 1962; Luo et al., 2000; Li 2002; Richards, 2004; Ye et al., 2007).

The changes of dew were referred to three aspects because of the differences of environment among landscapes, or the heterogeneities of the landscapes (Ye et al., 2007). The landscapes in present study were divided into different function by administrator, and the operations, land coverage, traffic were different from landscape to landscape, which would have an effect on dew. The heterogeneities of landscapes resulted in an alteration of the near surface atmospheric parameter,which would affect the formation process of dew directly. Therefore, the meteorological factors near the surface air play a key role on dew frequency and dew amounts.

Considering the physical process of dew formation, several parameters could be related to dew,including temperature, humidity, wind and surface property (Beysens, 1995; Li, 2002). Urban heat island is also a factor impacting the dew (Ackerman, 1987;Peng et al., 2005). In addition, evaporation could not be neglected when analyzing the correlation factors with dew (Ye et al., 2007).

Based on the previous researches, we investigated air temperature, urban heat island,relative humidity, evaporation and wind speed in the landscapes during the study period, results showed that there was no significant correlation between dew frequency and the mean value of the measured factors in each month, which was not consisted with previous researches referred above (Data and Figures were omitted). However, when comparing air temperature(Table 1) with the dew frequency (Fig. 2), it could be found that when the air temperature or urban heat island was the highest value among the landscapes in all collecting time (except July), the relevant dew frequency was the lowest, meaning that the higher temperature or urban heat island restrained the formation of dew. In July, on contrary, the highest dew frequency was observed under the condition of highest temperature or urban heat island. This result was also contrast with other researches (Hage, 1975; Ye et al.,2007). Temperature or urban heat island could make the dew event absence, but only if it arrived at an enough high level. Possibly, the dynamic variation of dew frequency was impacted by an integrated impact of several factors.

The dynamic characteristics of dew amounts were found in some reports, for example, the dew amounts in summer (0.033 mm) was lower than that in autumn(0.063 mm) in the sand dune of Minqin County (Zhen,1963), while contrary result was found on the fringes of denser oasis of Linze County where the dew amounts in summer (0.118 mm) was higher than that in autumn (0.087 mm) (Fang et al., 2005) in Gansu Province, China. Similar phenomena appeared in other investigations (Zhou et al., 2002; Wang et al., 2006;Yan et al., 2004). Actually, the information about the dynamic variation of dew amounts was limited and no document about that in urban area was found by far except some sporadic information. The dew amounts in April was lower than that in summer in present study was consistent with previous research (Wang et al., 2006), while the dew amounts in autumn was lower than that in summer was contrary to some researches(Zhen, 1963; Yan et al., 2004; Fang et al., 2005).

Dew frequency was considered as one of the important factor for the dynamic variation of dew amounts (Yan et al., 2004), the more dew frequency,the more dew amounts. However, contrary results appeared in present study, thus dew frequency was not the only impacting factor for the dynamic variation of dew amounts. Because the formation of dew was an exposed process, the difference of dynamic variation of dew amounts in different landscapes was also referred to several climatic factors, such as urban heat island, relative humidity, and evaporation (Monteith,1963; Angus, 1958; Duvdevani, 1964; Zangvil, 1996;Li, 2002).

In order to assess the effect of temperature, urban heat island, relative humidity, wind speed and evaporation on the dynamic variation of dew amounts,the correlations were analyzed and showed in Table 2～Table 6. It was found that the dynamic variation of relative humidity was positively related to the dynamic variation of dew amounts in most of the collecting time and measured sites, and the correlation in January and October was significant at the 0.05 level (Table 2), meaning that the relative humidity had an impact on the dynamic variation of dew amounts,and the higher RH would result in increasing dew amounts. The correlation between dew amounts and wind speed was complex (Table 3), changing fromlandscape to landscape and from time to time, of which, significant correlation was only found in two landscapes in January and April, respectively. In present field observation, dew amounts were positively correlated to temperature in most of the samples, and significant relationships were found in all landscapes in January and in FIL in October (Table 4). Results indicated that the increasing temperature was beneficial to the dew formation which was contrary to previous researches and the definition of dew (Hage,1975; Ackerman, 1987; Richards, 2004). The results between dew amounts and UHI were also puzzled though the negative correlation rate was over 50%(Table 5). Most of dew amounts were negatively related to evaporation (Table 6), meaning that the increasing evaporation was the possible reason for the decreased dew amounts in urban landscapes. Though the correlation results were out of expectation, dew condensation might be mainly related to relative humidity, urban heat island and evaporation (Ye et al.,2007).

Table 2 Relationships between urban dew amounts and nocturnal mean relative humidity

Table 3 Relationships between urban dew amounts and nocturnal mean wind speed

Table 4 Relationships between urban dew amounts and nocturnal mean temperature

Table 5 Relationships between urban dew amounts and nocturnal mean urban heat island

Table 6 Relationships between urban dew amounts and daily evaporation

4 Conclusions and application

The dynamic variation characteristics of urban dew were impacted by the urban landscapes, different urban landscapes and different collecting time exhibited different dynamic variation characteristics on dew frequency and dew amounts.

In current study, the measured sites belongs to the Pear River Basin, results were important information for the condensation resources in humid zone and sub-tropic region, and were the necessary supplement for dew data in urban area. Results presented some important dynamic information on dew amounts and dew frequency which were useful to promote the understanding of the dynamic variation of dew in urban area and non-arid region.The observation carried out would be benefic to the dew researches and any other related researches.The results in present study indicated that urban landscapes had an impact on the dew frequency and dew amounts, meaning that the urban management departments and the urban planning institutes should learn some knowledge such as dew condensation, and present results would be the basis for them. Though the dynamic variation characteristics of urban dew were analyzed in present study, results could not be applied directly in other cities for the different property in geography and urbanization.

In present study, the field observation was carried out in four months and four landscapes based on the limited fund, and some valid data were obtained. In order to gain more information on the dynamic variation of urban dew, future researches should consider the time series and more landscapes.In fact, urbanization results in lots of landscapes which would lead to different surface micro-climate,thus more researches in other types of landscapes would be meaningful. During the study process,relative humidity, temperature, wind speed and evaporation as well as the urban heat island were measured, but precipitation was not referred.Precipitation will influence the situation of other climate factors such as temperature, moisture, etc.,and consequently have an effect on dew condensation, so precipitation is the necessary factor in further study on dew. In our study, the height of the dew sample collecting set was confirmed according to the real condition of the landscapes in urban area, and no more attention was paid to the influence of height on dew amounts. Therefore, probably, it is a research direction in future.