Six years of dew observations in the Negev Desert, Israel

Journal of Arid Environments (1996) 32: 361–371

Six years of dew observations in the Negev Desert, Israel

Abraham Zangvil The Jacob Blaustein Institute for Desert Research, Ben-Gurion University of the Negev, Sede Boker Campus, 84990, Israel (Received 16 August 1994, accepted 9 December 1994) Dew measurements taken at the desert site of Sede Boker for nearly 6 years are analysed. The instrument used is a Hiltner-type dew balance. Several parameters describing various aspects of dew formation are discussed. The total monthly amount of dew and the distribution of the number of dew nights per month shows two maxima (in September and in December– January) and two minima (in April and November). The average dew deposit per dew night behaves differently: the most striking feature of this quantity is the appearance of distinct summer and winter regimes, with the winter having more dew per dew night. With respect to the total monthly dew hours, the year appears to be divided in half: first, the 6 months from August to January, with an average of 145 h per month, and second, the 6 months from February to June, with 80 h per month. The average duration of dew per dew night appears to follow very closely the length of the night: there is a clear maximum of dew duration in December (9·7 h) and a clear minimum in July (5·5 h). Finally, the rate of dew accumulation is found to have a distinct dry season regime and a winter, rainy season regime. ©1996 Academic Press Limited Keywords: dew; Negev Desert; Israel; climatology

Introduction One of the least explored aspects of desert meteorology is dew formation. Dew may play an important role in the water balance and growth regime of certain desert plants in coastal deserts. This in turn may have an effect on the whole desert ecosystem. Also, in newly developed desert agriculture, dew may influence the occurrence of plant disease (Wallin, 1967). Dew measurements in agricultural areas have been reported by Lloyd (1961), Newton & Riley (1964) and Getz (1978). The Negev heights are a barren desert area about 400–600 m above Mediterranean sea level (MSL). The region is located on the north to north-western slopes of the Negev mountains (Har Ha’Negev). The average rainfall in the region is about 90 mm and most of it falls during December to March. In certain dry years, the total annual rainfall can be as low as 25 mm. Because of the close proximity of the Mediterranean Sea (Sede Boker is 75 km from the sea) there is moisture available in the lower atmosphere and consequently there is a considerable number of dew nights. The agent 0140–1963/96/040361 + 11 $18.00/0

© 1996 Academic Press Limited

362

A. ZANGVIL

supplying the moisture is the sea breeze circulation which is active throughout most of the year. Records of frequency of occurrence, duration and amounts of dew in Israel are relatively scarce. Duvdevani (1947) initiated a network of observations for agricultural applications in Israel. Results of these observations were summarised by Gilead & Rosenan (1954). However, almost no observations exist for the Negev heights. One such study is reported by Evenari et al. (1971) summarising the results of 4 years of dew observations at Avdat about 10 km south of Sede Boker. Using the Duvdevani dew blocks, they found about 180 dew nights and a water equivalent of 30 mm annually. However, with this device (wooden blocks) it is not possible to measure the dew duration or the rate of accumulation. In order to study the atmospheric conditions involved in dew formation in the Negev, dew monitoring was initiated in January 1977, at the meteorological experimental site of the Institute for Desert Research at Sede Boker (see Fig. 1 for location). Zangvil & Druian (1980) analysed the results of the first 2 years of dew measurements. However, since only 2 or 3 months of observations were available for each calendar month the results in the above paper may not be representative. The purpose of the present paper is to extend the analysis to approximately 6 years. It will be shown that this period may be sufficient to establish a stable dew ‘climatology’. The general atmospheric and synoptic conditions prone to dew formation in Israel are described in the next section. In the following section, the instrumentation, site of measurements and method are described. The results of 6 years of dew monitoring are presented next.

Israel

Haila

Tel Aviv TEL AVIV

Jerusalem

Beer Sheva

Sede Boker

300 km 0

Figure 1. Study area and surroundings.

Km 25

50

Ellat

DEW OBSERVATIONS IN THE NEGEV

363

Micrometeorological and synoptic aspects The optimal atmospheric conditions for dew formation are discussed by Neumann (1956), Monteith (1956, 1963) and others. The two basic ingredients for dew formation are availability of moisture and efficient nocturnal radiational cooling. Up to a distance of about 100 km from the Mediterranean Sea a supply of moisture is available with on-shore winds and the sea breeze circulation. In order for dew to form, the relative humidity near the ground must be high. This may occur when a stable layer of air overlies a shallow moist layer near the ground. To obtain maximum radiational cooling, the following conditions must be met: (1) clear skies; (2) light winds; and (3) cold, dry air overlying a shallow moist layer near the ground. Good radiation cooling conditions exist throughout the year in the Negev (Fig. 2(a)) while the supply of moisture is a crucial factor as we go further away from the Mediterranean. It should be noted that dew condensation on the gauge depends on the net radiation,

Outward net radiation (w.m–2)

80.00 75.00 70.00 65.00 60.00 55.00 50.00

2

1

3

4

5

6 7 Month

8

9

10 11 12

Figure 2(a). Monthly average maximum nocturnal outward net radiation at Shivta site, 17 km north-west of Sede Boker.

18

Vapour pressure (hPa)

16 14 12 10 8 6 4 2 0

1

2

3

4

5

6 7 Month

8

9

10 11 12

Figure 2(b). Monthly average vapour pressure at 1400h local time.

364

A. ZANGVIL

with more outgoing radiation facilitating more cooling. However, high amounts of water vapour may hamper radiational cooling. Thus, a very delicate interplay between net radiation and surface and vertical moisture distribution is needed for optimal dew formation conditions to occur. Optimal synoptic conditions for dew formation occur throughout the year when a ridge of high pressure prevails in the mid-troposphere and causes a subsidence inversion to form at the lower layers. Moisture brought in by on-shore wind may be trapped under this stable inversion layer. Under such conditions, the sky is usually clear and the wind light, giving rise to good radiation cooling conditions. From the above, it is clear that the type of low level flow pattern associated with the upper ridge is crucial for dew formation. As far as the low level flow is concerned, the year may be divided into four seasons (e.g. Zangvil & Shemer, 1986): in the summer (mid-June to September), the Monsoon low pressure area in southern Asia extends to the north-eastern Mediterranean in the form of a low pressure trough while a weak ridge extends from the Azores High to North Africa and to the south-eastern Mediterranean. This pressure system causes a north-westerly flow over Israel, modified by the daily sea and land breeze circulation. The summer large scale synoptic flow pattern usually enhances the sea breeze. In the transition period from summer to winter (October to mid-November) high pressure ridges begin to build up over Turkey and a low pressure trough extends from the Sudan to the south-eastern Mediterranean. This flow configuration causes the winds to blow from a northerly to north-easterly direction. In addition, the land–sea temperature contrast diminishes and thus on average we can expect infrequent sea breezes and less humidity to be brought inland (Fig. 2(b)). During the winter season (mid-November to March) migrating cyclonic storms affect the region, causing variable atmospheric conditions such as strong winds and cloudiness which decrease the number of dew events. On the other hand, rainfall increases the soil moisture and also the humidity at the lower atmospheric layers. In the transition period from winter to summer (April to mid-June) the surface flow pattern is characterised by frequent occurrence of Khamsin depressions moving along the North African Coast eastward across Israel. Some cold lows may still affect the region at the beginning of the period and, in addition, the Sudan trough extends northward toward Israel from time to time. These highly variable conditions affect different regions of Israel in different ways as far as dew is concerned (e.g. Levi, 1967).

Instrumentation, data and procedure The measurement of dew has inherent problems not encountered in other meteorological observations. The difficulty is both in the definition of dew and in its actual measurement (Nagel, 1962; Noffsinger, 1965). In the literature (e.g. Noffsinger, 1965) dew is classified into several categories based mostly on the source of moisture (soil, atmosphere, etc.). However, since the distinction is not easily made in practice, it was decided in our case to define dew as any moisture, different from falling raindrops, which accumulates during the night on non-hygroscopic objects as a result of radiative cooling. In most cases of dew in the desert, the source of moisture is atmospheric. Thus, the desert offers an excellent environment for the study of pure atmospheric dew. On some occasions during the rainy season an additional source of dew water is, probably, moist soil (Lloyd, 1961). Another form of dew is that resulting from the interception of fog droplets by the receiving surface of the dew gauge. There is no method of distinguishing this type of dew deposit from the pure type of condensation due to radiational cooling. As mentioned at the outset, there is no standard, internationally accepted method or instrument for measuring dew.

DEW OBSERVATIONS IN THE NEGEV

365

The instrument used in this study is a Hiltner dew balance, manufactured by Lambrecht. In the original instrument, the receiving surface was made of a thin nylon mesh. The thinness of this mesh prohibits it from cooling radiationally below the ambient air temperature because of sensible heat flux from the air to the mesh. Thus, on nights with light dew, the original instrument did not record any dew at all. This fact was verified during the first 6 months of dew monitoring from January–June 1977. Because of this problem, it was decided to modify the receiving surface of the dew balance by adding a plastic disc, 0·2 mm thick, on top of the original nylon mesh. After this modification was made dew was recorded by the instrument whenever dew was observed on nearby objects. Nevertheless, it is realised that dew recorded by this instrument is still dependent on the instrument. No simple solution to this problem can be offered at this time. The dew gauge is located on bare loess soil at the meteorological observation site of the Institute for Desert Research. This plot is located on the north-easterly corner of the Sede Boker campus in Sede Zin, 45 km south of Beer Sheva, 480 m.a.s.l. The original data consisted of weekly charts on which the weight of dew deposit is continuously recorded as a function of time, from July 1977 to December 1982. The initial processing of the charts consisted of: (1) counting the number of dew events, nj per month, j = 1,…6; (2) reading off the peak weights of dew water, wi, and converting them into a depth equivalent, di, for each dew event. If several peaks occurred during the event, the highest one was selected. The total dewfall per month was calculated by n

Dj = ∑ di

j = 1,...,6

(Eqn 1)

i=1

and (3) measuring on the charts the total number of hours, hi, for which dew was increasing or remained unchanged, and summing up for each month. n

Hj = ∑ hi

j = 1,...,6

(Eqn 2)

i=1

The quantity H is the total monthly dew duration. (The actual time of dew wetness may be longer than hi by 1–3 h, depending on meteorological conditions.) The time lapse from the beginning to the end of dew evaporation after sunrise or during the night was not measured. ¯ and H ¯ — Using the above three basic parameters averaged over the 6 years — n¯ , D for each month we derived three more quantities: (1) the average dewfall per dew ¯ n (mm.day–1); (2) the average duration of dew event: H/¯ ¯ n (h.day–1); and (3) event: D/¯ –1 ¯ ¯ the average rate of dew accumulation: D/H (mm.h ). For the months January through June, 5 months were available, and for July through December, 6 months were available. The overbar (–) denotes averaging over 5 or 6 months.

Results and discussion All the six statistics calculated from the dew charts are presented in this section. In the following, the individual statistics and their physical and synoptic implications are discussed.

Total monthly dew amounts and number of dew events ¯ is shown in Fig. 3. Two main The total monthly amount of dew accumulation, D, features stand out. The existence of two main maxima and two minima. The summer maximum occurs in September and the winter maximum in December and January,

366

A. ZANGVIL

2.5

D (mm)

2.0 1.5 1.0 0.5 0.0

1

2

3

4

5

6 7 Month

8

9

10 11 12

¯ Figure 3. Total monthly dew amounts, D.

the latter being the absolute maximum. The minima are found in April and November. The summer maximum can be easily explained by the existence of the prevailing moist north-westerly flow over the eastern Mediterranean. The fact that the maximum occurs in late summer and not in mid-summer can be explained by the increased length of nights and lower nocturnal temperatures in late summer. In conjunction with the high vapour pressure in September (Fig. 2(b)) that means higher relative humidity, which is conducive to dew formation. The pronounced minimum in spring can be explained by the frequent occurrence of Khamsin depressions. Similarly, the less pronounced minimum in the fall is related to the development of the Sudan trough which causes frequent dry easterly flows in the region. The Sede Boker area is well exposed from the east to these winds. A possible explanation for the winter maximum is the increased relative humidity in the atmosphere which is partly a result of evaporation from the soil after rain. Another is the fact that Sede Boker is protected by hills from the winter westerlies, allowing more radiational cooling. In general, this result is consistent with Evenari et al. (1971), who found one main maximum in November and a minimum in April, at Avdat, 10 km south of Sede Boker. This slight inconsistency may be attributed to the fact that Avdat is more exposed to the winter westerlies than Sede Boker, and also to the different period of analysis. The Sede Boker total annual amount of ~ 17·0 mm of dew deposition is smaller than Avdat’s reported amount of ~ 30·0 mm. This discrepancy can be explained by the different device used by Evenari et al. Their thick wooden blocks allow for efficient surface radiational cooling. On the other hand, the thin receiving surface of our dew gauge permits more sensible heat exchange with the environment and less cooling, and thus, less dew. The distribution of the number of dew nights per month, n¯ , is shown in Fig. 4. The main features of this diagram are quite similar to those of Fig. 3: the location of the maxima and the minima is almost the same. However, the summer maximum is more pronounced than that of the winter maximum. Also, the winter maximum shifted from December–January in Fig. 3 to January. The fact that the summer maximum is more pronounced than the winter maximum can be explained by observing that in the summer, regular north-westerly flow associated with the sea breeze exists almost daily, bringing moisture from the Mediterranean, while in the winter there is a variable circulation.

DEW OBSERVATIONS IN THE NEGEV

367

30 25

n (days)

20 15 10 5 0

1

2

3

4

5

6 7 Month

8

9

10

11

12

Figure 4. Total monthly number of dew events, n¯ .

Dew duration ¯ was determined from the original charts as The average dew duration per month, H, described above. It is interesting to compare this quantity (Fig. 5) with the total dew ¯ and number of dew nights, n¯ , shown in Figs 3 and 4, respectively. The amounts, D, behaviour of this quantity retains the main features of Fig. 3. From the point of view of total monthly dew duration the dry season from May to October is clearly divided into two parts: an early part, consisting of the months of May through July, totalling about 75 dew h.month–1 (about 2·5 h night–1); and a later part, the dry months of August through October, having an average of 144 dew h month–1 (almost 5 h night–1). January and December also have over 5 h of dew per night. (The actual length of wetness per night may be longer by 1–3 h.) This finding may have some important implications on the growth regime of some desert plants and on the insects or other animals that feed on these plants. During the dry part of the year (April through November) a very high correlation exists between the vapour pressure, measured in a standard meteorological screen (see Fig. 2(b)), and the total monthly dew amount, dew hours and to a lesser extent, the monthly number of dew nights.

200 180 160

H (h)

140 120 100 80 60 40 20 0

1

2

3

¯ Figure 5. Total month dew hours, H.

4

5

7 6 Month

8

9

10

11

12

368

A. ZANGVIL

11

H/n (h.day–1)

10 9 8 7 6 5

1

2

3

4

5

6 7 Month

8

9

10

11

12

¯ n. Figure 6. Monthly average number of dew h.dew event–1, H/¯

On the other hand, it is noted that nocturnal net radiation in the Sede Boker area is maximum (outgoing) in April and May and minimum (outgoing) in October and September (Fig. 2(a)). Apparently, dew formation is influenced primarily by the humidity maximum in September rather than by the maximum net radiation (and low humidity in April–May). From the above data we have calculated the average duration of a dew night for each ¯ n is affected by two principal factors: firstly, the month (Fig. 6). This quantity, H/¯ length of the night and secondly, atmospheric conditions. In the following discussion, it is assumed that if dew duration per night is positively correlated with the length of the night in the month, astronomical factors dominate atmospheric factors. If such a proportion does not exist, atmospheric factors dominate. In Fig. 6 there is a maximum of dew duration in December (9·7 h.dew night–1) and a minimum in July (5·4 h.dew night–1). Recalling that the length of the night in December and July is 14 and 10·2 h, respectively, and observing that the dew starts to evaporate shortly after sunrise, we note that in December dew forms about 4·3 h after sunset while in July dew forms about 4·8 h after sunset. This finding, together with the fact that the night is longer in December than in July, indicates that dew duration in December is 9·7/14 or 69% of night length and in July only 5·4/10·2 or 53% of night length. This statistic can be interpreted as meaning that in early winter, atmospheric conditions on dew nights are more favourable for dew formation than in early summer. These favourable conditions may be increased atmospheric moisture, on the one hand (probably due to soil moisture evaporation), and frequent calm nights, on the other. In early summer there appears to be ample supply of moisture by the sea breeze reaching Sede Boker in the early afternoon (around 1500h local time). The sea breeze at this time of the year is usually combined with a moderate north-westerly geostrophic component which enhances its force and causes it to blow through the afternoon, and well into the evening. Usually, the wind starts to subside after 2100h or even later. Thus, dew does not start to form until midnight. In winter on calm nights dew may form very soon after sunset. Rate of dew accumulation and extreme events Another quantity which can be derived from the data is the average dew deposit per ¯ n. This is shown in Fig. 7(a). This quantity behaves rather differently dew night, D/¯ ¯ and n¯ (Figs 3 and 4). The most striking feature of this diagram is the from D

DEW OBSERVATIONS IN THE NEGEV

369

0.16

D/n (mm.day–1)

0.14 0.12 0.10 0.08 0.06 0.04 0.02 0.00

1

2

3

4

5

6 7 Month

8

9

10

11

12

¯ n. Figure 7(a). Monthly average dew per dew event, D/¯

appearance of a ‘dry’ and ‘wet’ regime, with 0·075 mm day–1 and 0·125 mm day–1, respectively. The amount of dew per dew night is larger in the winter season probably because of the longer nights but possibly also because of increased moisture supply near the ground due to rain. This is closely tied with the hourly rate of dew accumulation discussed below. The winter regime (December through March) with high dew deposits per dew night, coincides with the 4 rainiest months. During the dry months, ¯ n increases gradually due to the increase in the length of the nights. D/¯ Extreme values of dew accumulation per dew event are shown in Fig. 7(b). The absolute maximum measured during the period of observations is 0·32 mm day–1 in September. In general the month by month variation of the maximum dew amount does not appear to be well correlated with the monthly mean amount of dew per dew event. Rather, it appears to be best correlated with the total monthly dew amounts (Fig. 3). The most interesting quantity from the point of view of the microphysics of dew is ¯ H, ¯ shown in Fig. 8. This the monthly average hourly rate of dew accumulation, D/ quantity represents the quality of atmospheric and soil condition pertinent to dew ¯ H, ¯ contrary to the quantities discussed above, does not formation. The behaviour of D/

0.40

Dew (mm.day–1)

0.35 0.30 0.25 0.20 0.15 0.10 0.05 0.00

1

2

3

4

5

6 7 Month

Figure 7(b). Extreme amounts during dew events.

8

9

10

11

12

370

A. ZANGVIL

0.015 0.014

D/H (mm.h–1)

0.013 0.012 0.011 0.010 0.009 0.008 0.007 0.006 0.005

1

2

3

4

5

7 6 Month

8

9

10

11

12

¯ H. ¯ Figure 8. Monthly average dewfall rate, D/

show marked peaks or lows. The general feature of this figure is a rather uniform average rate of dew accumulation throughout the year ( ~ 0·012 mm h–1). Upon closer examination, two main peaks and two main lows can be identified and there is some resemblance to Fig. 7(a). A summer peak occurs in July, when the soil is almost dry (as far as the effect of its moisture content on dew formation is concerned). This peak can be interpreted as a manifestation of optimal atmospheric condition for dew formation in the month of July. The main peak which occurred in December while the soil was considerably moist represents an optimal combination of soil and atmospheric conditions for dew formation. The additional effect of soil moisture availability causes the dew rate in December to exceed that of July. The relative minimum in January can be explained by the fact that the mixing ratio of water vapour in the atmosphere in January, the coldest month, is lower than in December (see Fig. 2(b)), hence, to generate the amount of downward moisture flux as in December, an unrealistically large temperature gradient from the dew gauge plate to the surrounding air would be needed.

Conclusion Nearly 6 years of dew measurements at Sede Boker yielded a unique climatology of dew at this coastal desert location. The month with the largest percent of variation, 100% S.D. mean–1, was March, with 43%, but the remaining months had values of less than 30% and the months of August, September and October had values of less than 10%. This means that 6 years of data analysed yield mean values within 5% of the expected population mean. It is interesting to note that the monthly rainfall data from December through March have a percent of variation of nearly 70%. The annual total dew amount had a percent of variation of 12·3%, compared to 53% for annual rainfall. These results mean that despite its smaller amounts, dew may be a much more reliable source of moisture for micro-organisms or vegetation that are able to utilize it (e.g. lichen). The author wishes to thank Mrs Perla Druian for the data processing and Dr Ilia Apterman for graphics.

DEW OBSERVATIONS IN THE NEGEV

371

References Duvdevani, S. (1947). An optical method of dew estimation. Quarterly Journal of the Royal Meteorological Society, 73: 282–296. Evenari, M., Shanan, L. & Tadmor, A. (1971). The Negev — The Challenge of a Desert. Cambridge, Mass: Harvard University Press. 345 pp. Getz, R.R. (1978). Dew monitoring network in the Southeast. Bulletin of the American Meteorological Society, 59: 1150–1154. Gilead, M. & Rosenan, N. (1954). Ten years of dew observation in Israel. Israel Exploration Journal, 4 (2). Levi, M. (1967). Fog in Israel. Israel Journal of Earth Sciences, 16: 7–21. Lloyd, M.G. (1961). The contribution of dew to the summer water budget of northern Idaho. Bulletin of the American Meteorological Society, 42: 572–580. Monteith, J.L. (1956). Dew. Quarterly Journal of the Royal Meteorological Society, 42: 572–580. Monteith, J.L. (1963). Dew: Facts and Fallacies. In: Rulter, A.J. & Wholehead, F.W. (Eds), The Water Relations of Plants. London: Blackwell. Nagel, J.F. (1962). On the measurement of dew. Archives of Meteorology, Geophysics and Bioclimotology Series B, 11: 403–423. Neumann, J. (1956). Estimating the amount of dewfall. Archives of Meteorology, Geophysics and Bioclimotology Series A, 9: 197–203. Newton, O.H. & Riley, J.A. (1964). Dew in the Mississippi delta in the fall. Monthly Weather Review, 92: 369–373. Noffsinger, T.L. (1965). Survey of techniques of measuring dew. In: Wexler (Ed.), Humidity and Moisture, Vol. II. New York: Reinhold. Wallin, J.R. (1967). Agrometeorological aspects of dew. Agricultural Meteorology, 4: 85–102. Zangvil, A. & Druian, P. (1980). Measurement of dew at a desert site in Southern Israel. Geographical Research Forum, 2: 26–34. Zangvil, A. & Shemer, D. (1986). Climate. In: Stern, E., Gradus, Y., Meir, A., Krakover, S. & Tsoar, H. (Eds), Atlas of the Negev, pp. 59–66. Jerusalem: Keter. 95 pp.