• Aboal Viñas, J., 1998: Los flujos netos hidrológicos y químicos asociados de un bosque de laurisilva en Tenerife (The hydrological and chemical net fluxes associated with a laurel forest in Tenerife). Ph.D. thesis, University of La Laguna, 184 pp. [Available online at http://dialnet.unirioja.es/servlet/tesis?codigo=637.].

  • Allen, R. G., Pereira L. S. , Raes D. , and Smith M. , 1998: Crop evapotranspiration: Guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56, Food and Agriculture Organization of the United Nations, 300 pp.

  • Aschan, G., 1998: Microclimate, Energy Budget and Water Balance of Tropical and Extratropical Forests. (in German). Biology Series, Vol. 159, Tectum-Verlag, 256 pp.

    • Search Google Scholar
    • Export Citation
  • Axelrod, D. I., 1975: Evolution and biogeography of Madrean-Tethyan sclerophyll vegetation. Ann. Mo. Bot. Gard., 62 , 280334.

  • Azevedo, J., and Morgan D. L. , 1974: Fog precipitation in coastal California forests. Ecology, 55 , 11351141.

  • Brewer, C. A., Smith W. K. , and Vogelmann T. C. , 1991: Functional interaction between leaf trichomes, leaf wettability and the optical properties of water droplets. Plant Cell Environ., 14 , 955962.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bruijnzeel, L. A., 2001: Hydrology of tropical montane cloud forests: A reassessment. Land Use Water Resour. Res., 1 , 1.11.18.

  • Bruijnzeel, L. A., and Proctor J. , 1995: Hydrology and biochemistry of tropical montane cloud forests: What do we really know? Tropical Montane Cloud Forests, L. S. Hamilton, J. O. Juvik, and F. N. Scatena, Eds., Ecological Studies, Vol. 110, Springer-Verlag, 38–78.

    • Search Google Scholar
    • Export Citation
  • Bruijnzeel, L. A., Waterloo M. J. , Proctor J. , Kuiters A. T. , and Kotterink B. , 1993: Hydrological observations in montane rain forests on Gunung Silam, Sabah, Malaysia, with special reference to the “Massenerhebung” effect. J. Ecol., 81 , 145167.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Brunn, R. J., Lewis W. , Perkins P. J. , and Serafini J. S. , 1955: Impingement of cloud droplets on a cylinder and procedure for measuring liquid-water content and droplet sizes in supercooled clouds by rotating multicylinder method. NASA Rep. 1215, 183 pp.

  • Brutsaert, W., 1975: Comments on surface roughness parameters and the height of dense vegetation. J. Meteor. Soc. Japan, 53 , 9697.

  • Cáceres, G., 1981: Hydrological importance of horizontal interception in a humid premontane forest in Balalaica, Turrialba, Costa Rica (in Spanish). M.S. thesis, Centro Agronómico Tropical de Investigación y Enseñanza, University of Costa Rica, 98 pp.

  • Cavelier, J., and Goldstein G. , 1989: Mist and fog interception in elfin cloud forests in Colombia and Venezuela. J. Trop. Ecol., 5 , 309322.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cavelier, J., Solis D. , and Jaramillo M. A. , 1996: Fog interception in montane forests across the Central Cordillera of Panamá. J. Trop. Ecol., 12 , 357369.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chang, S-C., Lai I-L. , and Wu J-T. , 2002: Estimation of fog deposition on epiphytic bryophytes in a subtropical montane forest ecosystem in northeastern Taiwan. Atmos. Res., 64 , 159167.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dohrenwend, R. E., 1979: Hydrologic behavior at the top of a tropical mountain. Ford Forestry Center Research Note 29, Michigan Technological University, 14 pp.

  • Eugster, W., Burkard R. , Holwerda F. , Scatena F. N. , and Bruijnzeel L. A. , 2006: Characteristics of fog and fogwater fluxes in a Puerto Rican elfin cloud forest. Agric. For. Meteor., 139 , 288306.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fischer, D. T., and Still C. J. , 2007: Evaluating patterns of fog water deposition and isotopic composition on the California Channel Islands. Water Resour. Res., 43 .W04420, doi:10.1029/2006WR005124.

    • Search Google Scholar
    • Export Citation
  • Førland, E. J., and Coauthors, 1996: Manual for operational correction of Nordic precipitation data. Norwegian Meteorological Institute Rep. 24/96, KLIMA, 66 pp.

  • Friedlander, S. K., 2000: Smoke, Dust, and Haze: Fundamentals of Aerosol Dynamics. Oxford University Press, 407 pp.

  • Giambelluca, T. W., and Nullet D. , 1991: Influence of the trade-wind inversion on the climate of a leeward mountain slope in Hawaii. Climate Res., 1 , 207216.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gioda, A., Maley J. , Guasp R. E. , and Baladón A. A. , 1995: Some low elevation fog forests of dry environments: Applications to African paleoenvironments. Tropical Montane Cloud Forests, L. S. Hamilton, J. O. Juvik, and F. N. Scatena, Eds., Ecological Studies, Vol. 110, Springer-Verlag, 156–164.

    • Search Google Scholar
    • Export Citation
  • Golubic, I., 2001: Vegetation analysis of laurel forest and similar stands within the research framework of the water balance in the Garajonay National Park, La Comera (Canary Islands) (in German). Diploma thesis, Department of Plant Sociology and Ecology, University of Duisburg-Essen, 107 pp.

  • González-Rodríguez, A. M., Morales D. , and Jiménez M. S. , 2002: Leaf gas exchange characteristics of a Canarian laurel forest tree species [Persea indica (L.) K. Spreng.] under natural conditions. J. Plant Physiol., 159 , 695704.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Goodman, J., 1982: Water potential from advection fog. Department of Meteorology Progress Rep. 1, San Jose State University, 45 pp.

  • Gunn, R., and Kinzer G. D. , 1949: The terminal velocity of fall for water droplets in stagnant air. J. Meteor., 6 , 243248.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gutiérrez, A. G., and Coauthors, 2008: Regeneration patterns and persistence of the fog-dependent Fray Jorge forest in semiarid Chile during the past two centuries. Global Change Biol., 14 , 161176.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hafkenscheid, R. L. L. J., 2000: Hydrology and biogeochemistry of montane rain forests of contrasting stature in the Blue Mountains of Jamaica. Ph.D. thesis, Vrije Universiteit, 302 pp.

  • Herckes, P., Mirabel P. , and Wortham H. , 2002: Cloud water deposition at a high-elevation site in the Vosges Mountains (France). Sci. Total Environ., 296 , 5975.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holder, C. D., 2007: Leaf water repellency of species in Guatemala and Colorado (USA) and its significance to forest hydrology studies. J. Hydrol., 336 , 147154.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Höllermann, P., 1981: Microenvironmental studies in the laurel forest of the Canary Islands. Mt. Res. Dev., 1 , 193207.

  • Holwerda, F., Scatena F. N. , and Bruijnzeel L. A. , 2006: Throughfall in a Puerto Rican lower montane rain forest: A comparison of sampling strategies. J. Hydrol., 327 , 592602.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hutley, L. B., Doley D. , Yates D. J. , and Boonsaner A. , 1997: Water balance of an Australian subtropical rainforest at altitude: The ecological and physiological significance of intercepted cloud and fog. Aust. J. Bot., 45 , 311329.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ingraham, N. L., and Matthews R. A. , 1988: Fog drip as a source of groundwater recharge in northern Kenya. Water Resour. Res., 24 , 14061410.

  • Joslin, J. D., Mueller S. F. , and Wolfe M. H. , 1990: Tests of models of cloudwater deposition to forest canopies using artificial and living collectors. Atmos. Environ., 24A , 30073019.

    • Search Google Scholar
    • Export Citation
  • Juvik, J. O., and Ekern P. C. , 1978: A climatology of mountain fog on Mauna Loa, Hawaii Island. Water Resources Research Center Tech. Rep. 118, University of Hawaii, 63 pp.

  • Juvik, J. O., and Nullet D. , 1995: Comments on “A proposed standard fog collector for use in high-elevation regions”. J. Appl. Meteor., 34 , 21082110.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kämmer, F., 1974: Klima und vegetation auf Tenerife, besonders in Hinblick auf den nebelniederschlag. Scr. Geobot., 7 , 178.

  • Klemm, O., Wrzesinsky T. , and Scheer C. , 2005: Fog water flux at a canopy top: Direct measurement versus one-dimensional model. Atmos. Environ., 39 , 53755386.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Langmuir, I., and Blodgett K. B. , 1946: A mathematical investigation of water droplet trajectories. The Collected Works of Irving Langmuir, V. J. Schaefer, Ed., Pergamon Press, 348–393.

    • Search Google Scholar
    • Export Citation
  • Laws, J. O., and Parsons D. A. , 1943: The relation of raindrop-size to intensity. Eos, Trans. Amer. Geophys. Union, 24 , 452460.

  • Lloyd, C. R., Gash J. H. C. , Shuttleworth W. J. U. , Marques F. , and De O F. , 1988: The measurement and modelling of rainfall interception by Amazonian rain forest. Agric. For. Meteor., 43 , 277294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lovett, G. M., 1984: Rates and mechanisms of cloud water deposition to a subalpine balsam fir forest. Atmos. Environ., 18 , 361371.

  • McComber, P., and Touzot G. , 1981: Calculation of the impingement of cloud droplets in a cylinder by the finite-element method. J. Atmos. Sci., 38 , 10271036.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Monteith, J., and Unsworth M. , 1990: Principles of Environmental Physics. 2nd ed. Edward Arnold, 291 pp.

  • Müller, S. F., 1991: Estimating cloud water deposition to subalpine spruce-fir forests. I: Modifications to an existing model. Atmos. Environ., 25A , 10931104.

    • Search Google Scholar
    • Export Citation
  • Müller, S. F., Joslin J. D. , and Wolfe M. H. , 1991: Estimating cloud water deposition to subalpine spruce-fir forests. II: Model testing. Atmos. Environ., 25A , 11051122.

    • Search Google Scholar
    • Export Citation
  • Padilla, H., Belmont R. , Torres M. C. , Garcia R. , and Baez A. , 1996: A field comparison of cloud water collectors in a mountainous region under low wind speed conditions in eastern Mexico. Atmósfera, 9 , 189199.

    • Search Google Scholar
    • Export Citation
  • Pérez de Paz, P. L., 1990: Garajonay National Park: World Heritage. (in Spanish). National Institute for the Conservation of Nature, Excmo. Cabildo Insular de La Gomera, 367 pp.

    • Search Google Scholar
    • Export Citation
  • Riehl, H., 1979: Climate and Weather in the Tropics. Academic Press, 611 pp.

  • Ritter, A., Regalado C. M. , and Aschan G. , 2007: An impaction model for estimating fog water collection in a subtropical laurel cloud forest of the Garajonay National Park. Proc. Fourth Int. Conf. on Fog, Fog Collection and Dew, La Serena, Chile, Pontificia Universidad Católica de Chile, 355–358.

  • Santana, L., 1986: Study of Precipitation Fog in Tenerife. (in Spanish). National Institute for the Conservation of Nature, 97 pp.

  • Schemenauer, R. S., and Cereceda P. , 1994: The role of wind in rainwater catchment and fog collection. Water Int., 19 , 7076.

  • Sevruk, B., 2005: Rainfall measurement: Gauges. Encyclopedia of Hydrological Sciences, M. G. Andersen, Ed., Vol. 1, Hydrometeorology, John Wiley, 529–536.

    • Search Google Scholar
    • Export Citation
  • Sharon, D., 1980: Distribution of hydrologically effective rainfall incident on sloping ground. J. Hydrol., 46 , 165188.

  • Shuttleworth, W. J., 1977: The exchange of wind-driven fog and mist between vegetation and the atmosphere. Bound.-Layer Meteor., 12 , 463489.

  • Shuttleworth, W. J., 1993: Evaporation. Handbook of Hydrology, D. R. Maidment, Ed., McGraw-Hill, 4.1–4.53.

  • Sperling, F. G., Washington R. , and Whittaker R. J. , 2004: Future climate change of the subtropical North Atlantic: Implications for the cloud forests of Tenerife. Climatic Change, 65 , 103123.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stadtmüller, T., 1987: Cloud Forests in the Humid Tropics: A Bibliographic Review. United Nations University Press, 82 pp.

  • Thom, A. S., 1971: Momentum absorption by vegetation. Quart. J. Roy. Meteor. Soc., 97 , 414428.

  • Thom, A. S., 1975: Momentum, mass and heat exchange. Vegetation and the Atmosphere, J. L. Monteith, Ed., Academic Press, 57–109.

  • Thom, A. S., and Oliver H. R. , 1977: On Penman’s equation for estimating regional evaporation. Quart. J. Roy. Meteor. Soc., 103 , 345357.

  • Vermeulen, A. T., Wyers G. P. , Römer F. G. , Van Leeuwen N. F. M. , Draaijers G. P. J. , and Erisman J. W. , 1997: Fog deposition on a coniferous forest in The Netherlands. Atmos. Environ., 31 , 375386.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vogelmann, H. W., 1973: Fog precipitation in the cloud forests of eastern Mexico. Bioscience, 23 , 96100.

  • Walmsley, J. L., Schemenauer R. , and Bridgman H. A. , 1996: A method for estimating the hydrologic input from fog in mountainous terrain. J. Appl. Meteor., 35 , 22372249.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wrzesinsky, T., Thalmann E. , Burkard R. , Eugster W. , and Klemm O. , 2001: Fog deposition of nutrients and pollutants to a montane forest site. Proc. Second Int. Conf. on Fog and Fog Collection, St. John’s, NL, Canada, Meteorological Service of Canada, 169–172.

  • Wyslouzil, B. E., Whipple M. , Chatterjee Ch , Walcerz D. B. , Weathers P. J. , and Hart D. P. , 1997: Mist deposition onto hairy root cultures: Aerosol modeling and experiments. Biotechnol. Prog., 13 , 185194.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zadroga, F., 1981: The hydrological importance of a montane cloud forest area of Costa Rica. Tropical Agricultural Hydrology: Watershed Management and Land Use, R. Lal and E. W. Russell, Eds., John Wiley, 59–73.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    Location of (a) the experimental watershed and (b) plots. Distribution of (c) laurel forest types and (d) survey plots (marked with asterisk) within the watershed; isolines in (d) drawn every 5 m.

  • View in gallery

    Fog droplet size frequency distributions for eight liquid water content classes, after Klemm et al. (2005).

  • View in gallery

    Diagram of the strategy used to calculate fog water contribution to soil by E. arborea trees from fog water collected by artificial catchers (QFCs). Bold italic font indicate estimated parameters; subscripts “fc” and “v” refer to QFC and vegetation parameters, respectively.

  • View in gallery

    Scatterplots of the rain gauge and QFC hourly rainfall totals at the four plots for different wind speed intervals.

  • View in gallery

    Frequency histograms of data observed at the four measurement plots for rainfall (a) intensity and (b) duration, wind (c) speed and (d) direction, and (e) fog water collection rate. From left to right: bars correspond to plots P1145, P1185, P1230, and P1270.

  • View in gallery

    Monthly totals of rainfall and fog water collection at the four measurement sites during the 2-yr period. Rainfall data correspond to the average value (±std error) of the four plots. From left to right: bars correspond to plots P1145, P1185, P1230, and P1270.

  • View in gallery

    Mean monthly fog water collected by artificial fog catchersin the four plots. Bars indicate 95% confidence interval.

  • View in gallery

    Frequency histograms of estimates obtained with the impaction model for (a) liquid water content w and (b) fog water q collected by E. arborea tree. Bars represent average frequency values (±std error) computed assuming different fog droplet sizes (5, 10, 15, 25, 40, and 50 μm).

  • View in gallery

    Hourly evolution of micrometeorological variables measured in P1270 from February 2003 to January 2005: (a) totals of fog water captured by a single tree at each time of day for whole measurement period; (b) global radiation Rg; (c) linear regression obtained for Rg under foggy and fog-free conditions; (d) actual wind speed; (e) max wind speed; (f) air temperature; and (g) relative humidity. Dashed lines in (b) and (d)–(g) indicate hourly mean values (thick line) and ±std dev (thin lines) under fog-free conditions. Range of micrometeorological variables when fog was collected indicated by gray bands.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 593 323 26
PDF Downloads 500 279 17

Fog Water Collection in a Subtropical Elfin Laurel Forest of the Garajonay National Park (Canary Islands): A Combined Approach Using Artificial Fog Catchers and a Physically Based Impaction Model

View More View Less
  • 1 Instituto Canario de Investigaciones Agrarias, Tenerife, Spain
  • | 2 Applied Botany, University of Duisburg-Essen, Essen, Germany
Full access

Abstract

Fog precipitation has long been assumed as an additional water source in the relic laurel ecosystems of the Canary Islands, located at 500–1400 m MSL. However, to what extent fog water can contribute to the laurel forest water balance is not yet clear. Combining data from artificial fog catchers and a physically based impaction model, the authors evaluated the potential contribution of fog water captured by needle-leafed Erica arborea L. trees in a selected watershed of the Garajonay National Park (La Gomera Island) for a 2-yr period (February 2003–January 2005). Fog water collection was measured with artificial catchers at four micrometeorological stations placed at 1145, 1185, 1230, and 1270 m MSL. Average fog water collection was only significant at the highest measurement site (one order of magnitude greater than at lower altitudes), totaling 496 L m−2 yr−1 during the 2-yr period. The average fog water yield in the first and second annual periods ranged between 0.2–5.0 and 0.1–2.1 L m−2 day−1, respectively. Rainfall exhibited seasonality, distinguishing between rainy and dry seasons, while fog water collection was distributed more evenly throughout the year. Regarding fog water captured by the vegetation, the impaction model predicted a significant amount of fog water potentially collected by a single E. arborea tree, on the order of 1810–2090 L yr−1. Taking tree population density into account, the yearly average water contribution to the soil surface by wind-driven fogs was 251–281 mm, whereas annual rainfall was 635 and 1088 mm, respectively. The hourly course of micrometeorological variables shows a 58% reduction in global radiation under foggy conditions and a concomitant 3°–6°C mean temperature decrease compared to fog-free periods. Thus, limiting evapotranspiration may also be a relevant effect of fog in this subtropical elfin cloud forest.

Corresponding author address: A. Ritter, Instituto Canario de Investigaciones Agrarias (ICIA), Apdo. 60 La Laguna, Tenerife 38200, Spain. Email: aritter@icia.es

Abstract

Fog precipitation has long been assumed as an additional water source in the relic laurel ecosystems of the Canary Islands, located at 500–1400 m MSL. However, to what extent fog water can contribute to the laurel forest water balance is not yet clear. Combining data from artificial fog catchers and a physically based impaction model, the authors evaluated the potential contribution of fog water captured by needle-leafed Erica arborea L. trees in a selected watershed of the Garajonay National Park (La Gomera Island) for a 2-yr period (February 2003–January 2005). Fog water collection was measured with artificial catchers at four micrometeorological stations placed at 1145, 1185, 1230, and 1270 m MSL. Average fog water collection was only significant at the highest measurement site (one order of magnitude greater than at lower altitudes), totaling 496 L m−2 yr−1 during the 2-yr period. The average fog water yield in the first and second annual periods ranged between 0.2–5.0 and 0.1–2.1 L m−2 day−1, respectively. Rainfall exhibited seasonality, distinguishing between rainy and dry seasons, while fog water collection was distributed more evenly throughout the year. Regarding fog water captured by the vegetation, the impaction model predicted a significant amount of fog water potentially collected by a single E. arborea tree, on the order of 1810–2090 L yr−1. Taking tree population density into account, the yearly average water contribution to the soil surface by wind-driven fogs was 251–281 mm, whereas annual rainfall was 635 and 1088 mm, respectively. The hourly course of micrometeorological variables shows a 58% reduction in global radiation under foggy conditions and a concomitant 3°–6°C mean temperature decrease compared to fog-free periods. Thus, limiting evapotranspiration may also be a relevant effect of fog in this subtropical elfin cloud forest.

Corresponding author address: A. Ritter, Instituto Canario de Investigaciones Agrarias (ICIA), Apdo. 60 La Laguna, Tenerife 38200, Spain. Email: aritter@icia.es

1. Introduction

Fog water deposition has been widely recognized as an important ecological factor in mountain forests frequently immersed in dense fog (Azevedo and Morgan 1974; Stadtmüller 1987; Cavelier and Goldstein 1989; Schemenauer and Cereceda 1994; Walmsley et al. 1996; Bruijnzeel 2001). This hydrological process, also known as fog precipitation, occurs when atmospheric water contained in wind-driven fogs impinges upon obstacles in their path. The small droplets impacting on an obstacle’s surface merge into bigger drops that flow down over the obstacle and finally drip to the ground. The role of fog precipitation as an additional water source in tropical and subtropical montane cloud forests has been often emphasized (Kämmer 1974; Zadroga 1981; Ingraham and Matthews 1988; Bruijnzeel and Proctor 1995; Hutley et al. 1997; Chang et al. 2002; Gutiérrez et al. 2008). Additionally, fog may also play an important role not only as an auxiliary water pool but as a mechanism to reduce solar radiation fluxes, that is, evaporation (Eugster et al. 2006) or as an income of dissolved nutrients into the soil (Hafkenscheid 2000). Conversely, continuous moistening of leaf surfaces by fog may reduce photosynthetic activity because of the reduced diffusivity of CO2 in water as compared to air, such that plants may have developed repellency strategies to reduce leaf wetness, thus favoring water dripping to the soil (Holder 2007) and increasing incident sunlight on the leaf surface by the lensing effect of merged water droplets (Brewer et al. 1991).

Laurel ecosystems are evergreen subtropical forests that resemble the Tertiary, which due to climatic changes have largely disappeared from southern Europe and North Africa. These forests survived the Quaternary ice age, seeking refuge in the favorable climatic conditions of the Macaronesian regions (Axelrod 1975; Gioda et al. 1995). Microclimatological conditions in these relic forests are unique. The persistent northeast trade winds in this area are forced up by the mountain slopes of the islands, thus leading to cooling, condensation, and the formation of an almost permanent layer of clouds, which is prevented from rising up on the north side of the islands because of the existence of a temperature inversion between 900 and 1500 m (Höllermann 1981; Sperling et al. 2004). This northside cloud belt, locally known as “mar de nubes” (sea of clouds), provides low evaporation and favorable fog formation conditions. This may help to sustain these evergreen cloud forests in windward areas at middle elevations in an otherwise arid environment like that of the Canary Islands situated at 28°N off the Sahara coast. However, this hypothesis has not been thoroughly investigated, and it has become a poorly proved fact that fog water captured by vegetation is responsible for the survival of the laurel forest. Different authors have commented on the significance of fog occurrence for the laurel forest in the Canary Islands. For instance, Kämmer (1974) concluded that fog precipitation was of minor ecological importance for most plant species of the laurel forest ecosystem, being only significant during the dry season at the edge of the stand or on very steep slopes, but on a small spatial scale. Lately Santana (1986), in support of these observations, reported that single isolated laurel trees captured more fog water than those in dense forest areas or located in ravines. Conversely, Aboal Viñas (1998) claimed that no relevant fog precipitation occurred in the Agua García laurel forest in Tenerife during his observation period. On the other hand, Höllermann (1981) suggested that the reduction of incoming solar radiation by fog may, in fact, play a much more important ecological role than the interception of fog droplets. Without taking into account the complexity of the fog collection phenomena, it is not yet clear to what extent fog water can significantly contribute to the laurel forest.

The amount of water produced by fog precipitation is highly site dependent (Walmsley et al. 1996), based partly on vegetation properties, climatic factors, and terrain characteristics (Bruijnzeel and Proctor 1995). Vegetation factors include (i) the height, size, and structure of the canopy; (ii) the arrangement and shape of leaves; (iii) the tree location and population density; and (iv) the presence and morphology of epiphytes. The climatic factors mainly refer to (i) the wind speed and direction, (ii) the fog frequency, and (iii) the fog properties (particularly the droplet size distribution and liquid water content). Orography, elevation, and orientation are terrain-related characteristics that also influence fog precipitation (Stadtmüller 1987; Cavelier and Goldstein 1989; Cavelier et al. 1996). Attempts to measure fog water interception have been conducted for many years and in different places around the world with variable results being reported (Goodman 1982; Bruijnzeel and Proctor 1995; Bruijnzeel 2001). Quantifying fog water captured by vegetation is, however, not an easy task. Traditionally, two basic methods have been established for measuring fog precipitation: artificial fog catchers and throughfall measurements. However, both methods exhibit problems concerning implementation and data interpretation. Artificial fog catchers provide an indication of the frequency and amount of fog that can be potentially captured. The problem using artificial fog catchers is that each forest canopy represents a unique situation that cannot be fully characterized with a fog water collector. This method has therefore been mainly used for site characterization (Bruijnzeel and Proctor 1995; Bruijnzeel 2001). The throughfall measurement approach measures the amount of water dripping inside the stand. This is, in most cases, a better method because it takes into account the amount of water intercepted and evaporated from the wetted canopy. Large numbers of gauges are needed to obtain frequent and spatially representative data (e.g., Lloyd et al. 1988), although a “roving” strategy may help to minimize the number of gauges necessary to quantify net precipitation (Holwerda et al. 2006). Other alternative methodologies have also been used for quantifying fog deposition, such as the eddy covariance measuring systems (Vermeulen et al. 1997; Wrzesinsky et al. 2001) or the monitoring of weight changes of a known mass of living mossy epiphytes suspended below the canopy (Chang et al. 2002). By taking into account the different concentrations of Na, 2H, and 18O isotopes between rain and fog water, either a sodium or isotope mass balance approach remains as another method to quantify the relative contribution of fog, with some limitations (Bruijnzeel 2001; Fischer and Still 2007). Other methods rely on measurements of fog liquid water content and fog droplet distributions, determined either directly (Eugster et al. 2006) or indirectly from horizontal visibility determinations (Klemm et al. 2005). Whereas the former requires expensive high-power supply equipment that is difficult to install in unattended remote areas, the latter suffers from the large uncertainty of reported visibility versus air liquid water content relationships (Klemm et al. 2005; Eugster et al. 2006).

In view of these difficulties, the use of models for determining fog water collection efficiency by vegetation based on physical principles may well be an alternative to the more traditional approaches discussed above (Bruijnzeel 2001; Herckes et al. 2002; Klemm et al. 2005). These models can be elaborated to include assumptions about dimensions and geometry of both fog droplets and capturing elements (leaves), collection efficiency, vertical gradient of wind speed, both orientation and sheltering effects of collecting elements, tree population density, or terrain topography (Shuttleworth 1977; Goodman 1982; Lovett 1984; Joslin et al. 1990; Müller 1991; Müller et al. 1991; Padilla et al. 1996; Walmsley et al. 1996).

In this paper, we follow a novel hybrid approach between traditional methods and modeling techniques, whereby artificial fog catcher data are used to derive fog characteristics, which are then combined with a physically based impaction model and terrain topography to quantify fog collection efficiency of a subtropical elfin cloud forest. We focus on fog precipitation in the most extensive and best conserved laurel ecosystem in the Canary Islands (Pérez de Paz 1990): the subtropical cloud forests of the Garajonay National Park on La Gomera Island, which was added to the UNESCO World Heritage List in 1986 (online at http://reddeparquesnacionales.mma.es/parques/garajonay/index.htm). First, we study the micrometeorological variables and the fog water collection measured with artificial fog catchers distributed within a selected forested watershed. Second, we use a physically based impaction model to compute the amount of fog water captured by the vegetation in the watershed and estimate the quantity that may contribute to the soil surface.

2. Materials and methods

a. Experimental site

The study area is a 43.7-ha watershed within the Garajonay National Park (Fig. 1a). It is located between 1090 and 1300 m MSL and is exposed to a stratocumulus cloud layer that produces fog as it reaches the central high plateau of the island. Orography (Figs. 1b and 1d) is characterized by steep slopes up to 40%. The evergreen forest vegetation within the watershed (Fig. 1c) can be basically classified into three main groups (Golubic 2001): 1) “valley laurel forest” is characterized by a tall (10–25 m high) and dense tree layer mainly composed of Laurus azorica (Seub.) Franco, Persea indica (L.) K. Spreng, and Ilex canariensis Poir.; 2) “hillside laurel forest” represents a transitional vegetation covering the midslopes between the bottom of the valleys and the mountain crests. The 10–15-m high tree layer is dominated by Laurus azorica, Ilex canariensis, and Myrica faya Ait, with some scarce tall Erica arborea L. trees partially covered by epiphytes. The well-developed shrub layer consists mainly of young laurel trees; and 3) the wax myrtle-tree heath (fayal-brezal) is the most drought-adapted vegetation, located at crests and upper slopes. It is characterized by low (7–12 m height), thin, and shrubby E. arborea stands with a high abundance of terricolous and epiphytic mosses and lichens. Ten survey plots identified by Golubic (2001) within the studied watershed (Fig. 1d) provide data about the tree population density for each of the three forest types discussed above. Four scaffolded towers located at different sites within the watershed were used in this study. Each site or plot was characterized by elevation, aspect, and vegetation type (Table 1; Figs. 1c and 1d). The plots were denoted as P1145, P1185, P1230, and P1270 according to their altitude, expressed in m MSL after each letter. The scaffolded towers were instrumented above the canopy to measure rainfall, fog water collection, and micrometeorological variables for computing wet canopy evaporation. From February 2003 to January 2005, data were measured every 3 min and stored as average and totals at 15-min intervals.

b. Wet canopy evaporation rate

The wet canopy potential evaporation rate (i.e., surface resistance set to zero) at 15-min intervals was estimated with the Penman–Monteith equation (Allen et al. 1998), which can be split into an energy and aerodynamic term as
i1525-7541-9-5-920-e1
where λ is the latent heat of vaporization (J kg−1); Δ is the slope of the saturation vapor pressure curve at ambient temperature (Pa K−1); γ is the psychrometric constant (Pa K−1); Rn describes the net radiation for the vegetation cover (W m−2); G is the soil heat flux (W m−2), which was approximated during daylight and nighttime periods as 10% and 50% of Rn, respectively; ρ is the air density (kg m−3); cp is the heat capacity of air (J kg−1 K−1); es and ea are the saturated and actual vapor pressure (Pa); and ra is the aerodynamic resistance (s m−1) computed after Thom and Oliver (1977) with the following:
i1525-7541-9-5-920-e2
with zu and ze being the heights at which wind speed (u) and humidity are measured. The vapor roughness length is given by z0υ = 0.1z0, where z0 is the momentum roughness length of the forest (m) computed as (Thom 1971)
i1525-7541-9-5-920-e3
with h (m) being the average height of the canopy. From Brutsaert (1975), the above expression may be simplified assuming d = 2h/3 for forest conditions; that is, z0 = 0.123h. Air temperature, relative humidity, wind speed (mean and maximum), and solar radiation were measured above the canopy using the HMP45C temperature/relative humidity probe, the A100R switching anemometer (Campbell Scientific Ltd.), and the SKS 1110 pyranometer (Skye Instruments Ltd.), respectively. The pyranometer data were used for computing net radiation, according to Allen et al. (1998), as the difference between the incoming net shortwave and the net outgoing longwave radiation using an albedo of 0.14 (Aschan 1998). The wet canopy potential evaporation rate was computed with the ETPM_H program (available online at http://www.icia.es/gh/software.html).

c. Correction of rainfall measurements

Rainfall data were obtained at each scaffolded tower with a Rain-O-Matic Professional (0.2-mm resolution) spoon tipping rain gauge (Pronamic Bekhøi International Trading Engineering Co. Ltd.). When windy conditions prevail, actual rainfall incident to the forest may be underestimated if this is assumed to be equal to the rain gauge measurements (Sevruk 2005). Therefore, these were corrected for possible 1) inclined rainfall falling on a sloping terrain and 2) wind-induced losses due to wind field distortion at the rim of the rain gauge. According to Førland et al. (1996), a correction factor (kf) was first computed based on wind speed (u) and rainfall intensity (R) so as to calculate a wind-corrected rainfall intensity, Rc (mm h−1) = kfR:
i1525-7541-9-5-920-e4
To carry out the correction due to inclined rainfall incident to a sloping surface, the wind-corrected rainfall intensity Rc (mm h−1) was used to compute the median raindrop diameter D50 (mm) (Laws and Parsons 1943):
i1525-7541-9-5-920-e5
The terminal fall velocity VD (m s−1) of the raindrops was then calculated from the median droplet size such that (Gunn and Kinzer 1949)
i1525-7541-9-5-920-e6
Next, the mean rainfall inclination angle θ (°) may be computed from
i1525-7541-9-5-920-e7
From this inclination angle, a trigonometric model developed by Sharon (1980) results in a second correction factor to be applied to the wind-corrected rainfall to obtain the actual rainfall incident to the forest canopy.

d. Fog water measurements

Each scaffolded tower was instrumented for collecting fog water above the canopy with an artificial fog catcher (QFC), consisting of a 0.5 m × 0.5 m screen with a single layer of polypropylene Raschel-type mesh with a 65% shade coefficient. Each artificial fog catcher was attached to an additional spoon tipping Rain-O-Matic rain gauge such that fog droplets, impinging on the mesh elements and joining to form larger drops, fall under the influence of gravity into a tilted gutter below the mesh, from which they are conveyed to the rain gauge. Under fog conditions evaporation from the mesh was considered negligible. All QFCs were oriented in the direction of the predominant trade winds (northeastward).

e. The physical impaction model

The theoretical amount of fog water droplets captured by impaction on a cylindrical element or obstacle under wind-driven fog conditions can be estimated with the following expression (Goodman 1982; Walmsley et al. 1996):
i1525-7541-9-5-920-e8
where q is the rate of collected fog water (l h−1), w is the liquid water content (g m−3), A is the cross-sectional area (m2) of the obstacle, and u is wind speed (m s−1). To express q in common l h−1 units, a conversion factor of 3.6 (s h−1)(l g−1) is included in Eq. (8). Here η (−) is the fog capture efficiency due to impaction with the obstacle although other fog collection mechanisms such as direct interception, Brownian diffusion, and gravitational sedimentation, are also possible (Friedlander 2000). Inertial impaction occurs when fog water droplets encounter an obstacle in the direct path of the wind-driven airstreams. Additionally, a fog water droplet may not impact directly on the target, but may pass within one radius off the edge of the obstacle. In this instance, the droplet may be captured by a process called (direct) interception (Monteith and Unsworth 1995). Interception becomes an important mechanism of fog water capture for droplets with a diameter comparable with or larger than the obstacle that they approach. By contrast, collection by Brownian motion of fog water droplets is only significant for small droplet diameters (<0.1 μm). Finally, gravitational sedimentation is also a potential mechanism of fog deposition. In general, however, owing to the low settling velocity (υt) of essentially all water droplets with less than 80-μm diameter, gravitational settling may be considered as an insignificant fog capture mechanism (Walmsley et al. 1996). Considering the Stokes (Stk) and Reynolds (Re) numbers, wind velocity, and droplet and obstacle sizes relevant in this study, interception and diffusion may also be discarded as important fog capture mechanisms, in terms of their low efficiency compared with that of inertial impaction (Table 2). Interception efficiency is defined as (Wyslouzil et al. 1997)
i1525-7541-9-5-920-e9
where Dg and Dυ are the fog droplet and obstacle diameter, respectively, and Ku = 2 − ln(Re) is the Kuwabara number. The interception efficiency ηint is always ≤6% (Table 2). Diffusion efficiency is given by (Wyslouzil et al. 1997; Friedlander 2000)
i1525-7541-9-5-920-e10
where Pe is the Peclet number, which is never greater than 1% for typical fog droplet diameters (5 ≤ Dg ≤ 50 μm) and frequent wind velocities (1 and 3 m s−1) (Table 2). Additionally, the low (<7 cm s−1) settling velocities of fog droplets indicate the negligible contribution of the sedimentation process (cf. Fig. 10.1 in Monteith and Unsworth 1995). With respect to the efficiency of the fog capturing process due to impaction, ηimp, this depends on Stk, which is the ratio of the stopping distance of a particle (e.g., droplet) to the radius of the obstacle. The stopping distance indicates how long a particle will travel on its original path after the airstream is turned abruptly around an obstacle. Thus, ηimp becomes larger with increasing Stk. In addition, ηimp is also a function of the ratio ϕ = Re2/Stk. Both ϕ and Stk are influenced by u. The diameters of the cylindrical obstacle and fog droplets are also required. Values of ηimp for different Stk and ϕ can be obtained from Friedlander (2000). A simple form is that proposed by Lagmuir and Blodgett (1946) and Brunn et al. (1955) for inviscid flow and ϕ ≤ 10, verified numerically by McComber and Touzot (1981):
i1525-7541-9-5-920-e11
From (11), it follows that impaction efficiency is generally >60% for Dg ≥ 10 μm (Table 2). Thus, impaction may be considered the most important mechanism of fog water droplet deposition in our study; therefore, needle-type leaves, with diameters on the order of magnitude of the fog droplets, may be considered efficient in capturing fog by impaction (Shuttleworth 1977). In the following section, we shall concentrate on the model (8) with the collection efficiency given by (11), neglecting all other minor contributions from the remaining fog capture mechanisms.

f. Uncertainty in the estimates of the impaction model

The impaction model [Eqs. (8) and (11)] contains both linear and nonlinear terms. The q estimate uncertainty due to the linear terms (A and u) is simply proportional to the measurement errors associated with the obstacle’s normal cross-sectional area (A) and wind speed (u). Regarding the nonlinear terms (w and ηimp), the parameterization of the fog droplet size distribution proposed by Klemm et al. (2005) for eight w classes (Fig. 2) shows the interrelationship between w and Dg. Thereby, the sensitivity of the q estimates to droplet size distribution, and thus the liquid water content, was investigated for droplet diameters Dg = 5, 10, 15, 25, 40, and 50 μm. Since ηimp is affected by Dg through the Stokes number [Eq. (11)], uncertainty in ηimp was also taken into account by considering the above Dg range.

g. Application of the impaction model

Figure 3 is a diagram illustrating the procedure for estimating the contribution of fog water to the soil by E. arborea trees from the fog water collected by the artificial catcher. The mesh in the catcher can be envisaged as a combination of cylindrical elements oriented both vertically and horizontally. Thereby, following Eqs. (8) and (11), fog collector data were used to estimate the air liquid water content w. This was applied during fog-only conditions (i.e., no precipitation) because, when rainfall occurs, the artificial fog catcher may also collect rainwater (Juvik and Nullet 1995). Since the artificial fog catcher was oriented to the northeast, we calculated an effective wind speed, ueff, perpendicular to the mesh surface at each 15-min interval, which takes into account changes in wind direction. This was computed from the fog catcher’s orientation (α), the wind speed (u), and the wind direction (β) data, such that ueff = ucos(βα). Wind direction was measured with a W200P potentiometer windvane sensor (Campbell Scientific Ltd.). Collection efficiencies were calculated for fog droplet diameters 5, 10, 15, 25, 40, and 50 μm (Goodman 1982; Monteith and Unsworth 1995) and a mean diameter of the mesh elements in the fog catcher, Dfc = 1.4 mm.

Assuming that needle-type leaves may be envisaged as cylindrical elements, the impaction model described in Eqs. (8) and (11) was then applied to estimate the fog water captured by an individual tree with cylindrical vegetative elements from the above estimated air liquid water content. This is the case for the “fayal-brezal” vegetation type, which is composed of a significant proportion of E. arborea (needle-leaved trees). The application of Eqs. (8) and (11) to the vegetation uses the wind speed at the top of the canopy. For the sake of simplicity, constant wind speed and liquid water content with decreasing height within the canopy is assumed. The wind speed at the top of the canopy was computed from measurements at the top of the tower and the logarithmic wind speed profile assumption in the surface boundary layer (Thom 1975):
i1525-7541-9-5-920-e12
where uz is the wind speed (m s−1) at canopy height z (m), k = 0.41 is the von Kármán constant, u* the friction velocity, d the zero plane displacement height, and z0 is the momentum roughness length of the forest, as described above.

The capture efficiency was obtained from the wind speed data, fog droplet diameters, and the measured average leaf diameter Dυ = 0.25 ± 0.02 mm. We calculated the total normal cross-sectional area of all needlelike leaves in the canopy (Aυ = 2.77 m2 tree−1) from leaf area (m2 ha−1) and stem density (trees ha−1). This value was reduced by an orientation factor (0.64) that accounts for the canopy being composed of randomly oriented cylindrical elements (Shuttleworth 1977). In addition, the normal cross-sectional area was divided by a shelter factor, which tends to unity as the density of the vegetative elements decreases. Following Shuttleworth (1977), we assumed a shelter factor similar to those for momentum and water vapor, that is, 2.25. Tree population density was computed from Golubic (2001), as described in Ritter et al. (2007).

Thereby, from the artificial catcher measurements, fog water captured by an E. arborea canopy was computed at 15-min intervals. Additionally, the amount of water that does not evaporate from the canopy surface (8.70 m2 tree−1) was assumed to be incorporated into the soil by dripping.

3. Results and discussion

a. Rain gauge measurement corrections

Corrections of rain gauge measurements for wind-induced losses and the effect of inclined rainfall falling on a sloping surface under windy conditions are advisable (Sevruk 2005). The QFC can be envisaged as a nonideal rainfall collector, but with the particularity that it can also collect rainfall falling obliquely, because of the shield effect of the mesh. Hence, when precipitation occurs, the QFC tips are a consequence of both the drops falling directly into the gutter and those impacting on the mesh being conveyed into the gutter below. In an attempt to verify if the rainfall data were affected by our wind and rainfall intensity site-specific conditions, we compared the rain gauge measurements with the rainfall collected by the QFC. Figure 4 shows scatterplots of the rain gauges and the QFC hourly rainfall totals at the four sites for different wind speed intervals. For low wind speeds (≤1 m s−1), a linear relationship may be suggested, although this departs from the 1:1 line (Fig. 4a), showing the nonideal design of the QFC as a rainfall measuring device. Furthermore, as wind speed increases (Figs. 4b–f), data scatters further from the 1:1 line, thus indicating that the rain gauges were more affected by the windy conditions. This data bias toward the OY axes shows that rain gauge hourly totals are underestimated. Most precipitation differences >1 mm between the rain gauge and the QFC cannot be attributed to fog water collected by the QFC during fog and rainfall concurrent events because of the small rates at which fog droplets are captured. The approach followed here takes advantage of the noncovered design of the QFCs, given that using a protective cover does not guarantee elimination of all wind-driven rain (Juvik and Nullet 1995).

Motivated by these results, we corrected rain gauge measurements as described in section 2c. After correcting for both wind-induced losses and inclination angle, rainfall in the four plots was on average higher than rain gauge measurements (8 ± 4% and 5 ± 2% for February 2003–January 2004 and February 2004–January 2005, respectively). The mean correction factor for wind-induced losses for the whole measurement period was 1.068 ± 0.064. Similarly, the mean correction factor due to the effect of inclined rainfall falling on a sloping surface was 1.024 ± 0.08.

b. Micrometeorology within the watershed

Rainfall above the forest canopy did not vary considerably within the four plots. Rainfall intensities <2 mm h−1 were observed in 69 ± 1% of the events during the measurement period (Fig. 5a). Storm durations of 15 and 30 min represent 72 ± 1% of the events recorded (Fig. 5b). Figures 5c and 5d show the differences in wind speed and wind direction between the four plots. Although, in general, wind speeds <2 m s−1 were frequent for all plots, at P1270 higher wind speeds values (up to 12 m s−1) were observed (Fig. 5c). No predominant wind direction could be determined at P1185; however, data obtained at the other locations show wind blowing mainly from east and northeast at P1145 and P1230 and from north and northeast at P1270 (Fig. 5d) as a consequence of the predominant direction of trade winds. Data obtained from the artificial catchers indicate that fog water was collected at the four locations mainly at low intensities (<0.25 L h−1 m−2 horizontal area) (Fig. 5e). Figure 6 summarizes monthly totals of rainfall (as the mean value of the four observation sites ± standard errors) and fog water collected by each catcher during the whole measurement period, which may be split into two periods denoted as “Year1” (February 2003–January 2004) and “Year2” (February 2004–January 2005). In both periods, seasonal variation in rainfall can be observed from October to May (rainy season) and from June to September (dry season). The amount of rainfall in Year2, computed as the average value of the four plots (1088 ± 59 mm yr−1), was significantly different (p < 0.05; normality assumption checked with the Shapiro–Wilk’s test) and almost double that of Year1 (635 ± 26 mm yr−1). Evaporation of intercepted water from the canopy depends on the potential evaporation rate (E), which in this case resulted in similar yearly totals for Year1 and Year2 at the four plots (Table 3). These high values of potential wet canopy evaporation (E > 2500 mm) are a consequence of both high incoming radiant energy values (typical of the Canary Islands) and the more efficient aerodynamic vapor exchange of taller vegetation compared to shorter crops (Shuttleworth 1993). In this study, the average aerodynamic vapor exchange [second term in Eq. (1)] contribution to the total wet canopy evaporation was 61 ± 3% (obtained as the average of the four plots).

c. Fog water collection measured by artificial catchers

Fog collection by catchers was observed at the four locations throughout the observation period; however, the amounts measured at P1270 were significantly higher (one order of magnitude) than those obtained at lower elevations. Figure 7 and Table 3 summarize the differences between the four plots in relation to fog water collection. Compared to P1230, P1270 yielded 6 times more fog water. This value increases up to 13 and 19 times if the latter is compared to the plots located at lower altitudes (P1145 and P1185). No satisfactory time series cross-correlations (<0.3) were found between the amount of water captured by the fog catchers and the corresponding wind speeds. This was also the case when computing cross-correlations between wind speed data from the four locations. Despite this, average maximal wind speed data (Table 3) suggest that fog collection is affected by wind speed, which partly depends on site elevation. These results confirm that elevation is an important factor that influences fog precipitation (Cavelier and Goldstein 1989). However, additional aspects should be considered to explain the differences observed between P1270 and P1230. Hillside orientation could have an effect on wind canalization and acceleration that may improve fog capture (Stadtmüller 1987). The large amount of water obtained in P1270 may be a consequence of the location of this plot on an upper hillside, being thus exposed to higher wind speeds and higher air liquid water content. The upward movement of air from the lower to upper sites produces an increase in air saturation, wind speed, droplet size, and liquid water (Riehl 1979). Cavelier et al. (1996) also reported an increase in fog precipitation with increasing altitude as being more important on ridges than on slopes and valleys.

Focusing on P1270, differences between the first and the second period (Fig. 6) were observed. Fog water collection in Year2 is distributed more evenly throughout the year. This is not the case for Year1 when the amounts captured in the first semester (February 2003–July 2003) are higher than in the second semester (August 2003–January 2004). Despite this, both periods yielded similar amounts of fog water (561 and 432 l m−2 yr−1, respectively). Average fog water yield in both periods ranged between 0.2–5.0 l m−2 day−1 and 0.1–2.1 l m−2 day−1, respectively. These results are comparable to those found elsewhere. Vogelmann (1973) reported values between 0.4–1.6 l m−2 day−1 for the Sierra Madre (Mexico) from measurements performed at different elevations (1330–1900 m MSL). Juvik and Ekern (1978) obtained 1.9 and 2.1 l m−2 day−1 at 2530 and 1580 m MSL, respectively, for Mauna Loa (Hawaii). Dohrenwend (1979) also measured 2.1 l m−2 at about 3500 m MSL in Cerro de Buenavista (Costa Rica). Cáceres (1981) carried out experiments at 1300 m MSL in Balalaica (Costa Rica) and reported 4.0 l m−2 day−1. Results obtained by Cavelier and Goldstein (1989) ranged between 0.2–2.2 l m−2 day−1 from measurements in El Zumbador (Venezuela; 3100 m MSL), Cerro Copey (Venezuela; 987 m MSL), Cerro Santa Ana (Venezuela; 815 m MSL), and Serranía de Maguira (Colombia; 865 m MSL). Bruijnzeel et al. (1993) reported values of 0.4 l m−2 day−1 in Gunung Silam (Malaysia; 884 m MSL).

d. Water captured by vegetation

Since the amounts of water collected at plots P1145, P1185, and P1230 were not significant, we applied the impaction model described in section 2e at P1270 only. In addition, vegetation in these three lower plots is mainly characterized by trees with broad and large leaves, which are less efficient in capturing fog than needle-like leaves with diameters on the order of magnitude of the fog droplets because collection efficiency increases with decreasing obstacle diameter (Goodman 1982). We verified the sensitivity of the impaction model [(8) and (11)] to different fog droplet diameters for Dg = 5, 10, 15, 25, 40, and 50 μm. Figure 8a shows the average frequency distribution of calculated liquid water content for the 2-yr period. Liquid water contents computed using the above droplet sizes yielded mostly values w < 0.1 g m−3 (70 ± 7%) and less frequent values (13 ± 2%) of w = 0.1–0.2 g m−3. Similarly, Fig. 8b shows the frequency distribution of Dg average fog water collected by an E. arborea tree, with rates of q = 0.05–0.5 L h−1 tree−1 predominating during the measurement period and with relatively small error bars (72 ± 2%). These results suggest that the model [(8) and (11)] is robust, being little affected by the uncertainty in Dg. Furthermore, based on the parameterization of the fog droplet size distribution proposed by Klemm et al. (2005) for eight w classes (Fig. 2), it follows that for w = 0.025–0.1 g m−3 and w = 0.1–0.2 g m−3 the maximum droplet diameter frequencies are placed at Dg = 10 and 15 μm, respectively. In addition, Eugster et al. (2006) show how for w < 0.1 g m−3 fog droplet diameters were mainly below 20 μm. Furthermore, they show that diameters below 7 μm and above 30 μm do not contribute to fog water collection. Thus, in our case, Dg = 10 and 15 μm were the relevant diameters for the most frequently (83 ± 5%) computed w values (<0.2 g m−3; Fig. 8a). The yearly totals of computed fog water captured by a single E. arborea tree are given in Table 4, as the average of the results obtained using Dg = 10 and 15 μm. The wet canopy potential evaporation was subtracted from collected fog water at 15-min intervals, assuming that water that does not evaporate drips onto the soil. Thus, a single E. arborea tree may potentially contribute a considerable amount of fog water (>1500 L) to the annual soil water input. However, this result may be viewed as a fog collection upper bound. Single isolated trees have higher capture efficiency (Santana 1986; Schemenauer and Cereceda 1994), and the amount of fog precipitation may decrease with distance from the windward edge because the first line of trees captures a high percentage of the larger fog droplets, acting as a barrier to diminish the wind speed downstream (Goodman 1982). All of these effects deserve further investigation. When tree population density (trees per square meter of soil surface) is taken into account, the yearly contribution to the soil surface by the fog water captured by E. arborea trees may be considered, at best, 281 and 251 mm for Year1 and Year2, respectively.

e. Hourly evolution of climatic variables under foggy and fog-free conditions

The hourly evolution of micrometeorological variables measured in P1270 was also studied for the whole measurement period. Figure 9a shows the distribution of fog water captured by an E. arborea tree at each time of the day from February 2003 to January 2005. Fog collection occurs mostly during the night until early morning and in the late afternoon. Figures 9b–g show how global radiation, wind speed, temperature, and relative humidity change during the day for both fog-only and fog-free conditions. This illustrates how the presence of fog affects, or is associated with, these climatic variables. During foggy conditions, global radiation is reduced considerably (Fig. 9b). In addition, the relationship between global radiation measured during fog-free and foggy conditions at the same time of day was found to be linear (R2 = 0.993), indicating that global radiation is reduced by 58% in the presence of fog (Fig. 9c). Giambelluca and Nullet (1991) reported 36% attenuation of global radiation by fog on a Hawaiian forest under the influence of a trade wind temperature inversion. Recently, Eugster et al. (2006) found a 48 ± 2% reduction in global radiation under foggy conditions in a Puerto Rican elfin cloud forest. Fog occurrence is related to higher wind speeds (Figs. 9d–e); thus, during the day, both mean and maximum wind speeds are higher under foggy conditions compared to those values obtained when no fog is collected (4.4 ± 2.0 versus 2.4 ± 1.7 m s−1 and 7.0 ± 2.7 versus 4.2 ± 2.4 m s−1, respectively). Air temperature and relative humidity varied during the day (Fig. 9f,g) with higher values occurring from early morning until late afternoon. During the measurement period, these variables exhibited a range of around 10°C and 60% for fog-free conditions. The presence of fog is associated with very high (>98%) relative humidity values (inset Fig. 9g) and lower temperatures, which varied between 6.5° and 12.5°C. Eugster et al. (2006) found a similar temperature trend in their study, claiming that higher air temperatures may promote fog droplet evaporation, leading to fog dissipation during midday. All of these results indicate that fog may play a role not only as a source of additional water to the watershed budget but, more importantly, as a limiting factor in both evaporation from the canopy and plant physiology. Both carbon assimilation rate and stomatal conductance are significantly reduced below 15°C in laurel species (González-Rodríguez et al. 2002), such that the low temperatures (<12.5°C) observed here concomitant with fog conditions may reduce canopy transpiration and photosynthesis. Furthermore, the carbon assimilation rate is diminished as photosynthetic photon flux density decreases, with laurel leaves being light saturated at about 25%–37.5% of full sunlight (254–382 W m−2) (GonzálezRodríguez et al. 2002). Thus, according to Figures 9a,b, foggy conditions may significantly affect carbon assimilation but only during early morning and in the evening. Thereby, the potentially limiting effect on plant transpiration and photosynthesis because of light attenuation by fog is buffered by the shade tolerance of laurel forest tree species and the low probability of fog occurrence around midday.

4. Conclusions

The importance of the water supplied by wind-driven fogs in the Garajonay National Park was found to be limited to well-exposed high altitude areas. From the measurements of micrometeorological variables in four locations within an elfin laurel forest watershed, it follows that at the highest measurement site (1270 m MSL), the average yearly amount of water collected by artificial fog catchers was about 500 L m−2. Using a physically based impaction model and a simple hypotheses about the canopy fog capture process, the yearly amount of fog water provided by needle-leaved trees such as E. arborea was estimated to be ∼250 and 280 mm in the two years studied. This water input is distributed throughout the year, whereas rainfall exhibits seasonality. Compared to precipitation, the contribution of fog to the forest water balance is still small. However, an indirect effect on the water balance is that of fog on evapotranspiration. The hourly course of micrometeorological variables derived from our results shows characteristic trends concomitant with the presence of fog, such that fog occurrence may limit wet canopy evaporation and plant transpiration. Pushing this situation further, climate change models predict a downward shift of the trade wind inversion in the Canary Islands during the dry season by the end of the twenty-first century, leading to an increase in both vapor pressure deficit and radiation above the current upper cloud forest border (Sperling et al. 2004). Thereby, current evapotranspiration constraints imposed by fog conditions in the upper areas of the park may be reduced in the future owing to the more frequent fog-free situation that is expected. The combination of information provided in this study of an intensively instrumented watershed with a detailed botanical survey and topographical maps could be extended to other areas of the Garajonay National Park, with important implications for vegetation survival and conservation policies. The novel hybrid approach presented here, which combines a physically based fog droplet impaction model with artificial fog catcher data, may well serve as a low-cost alternative to other more sophisticated fog characteristic measuring methods.

Acknowledgments

This work was financed with funds from the INIA-Programa Nacional de Recursos y Tecnologías Agroalimentarias (Project RTA2005-228). A. Ritter acknowledges financial help from the European Social Fund. The authors thank Thomas Keller (Up Gmbh, Germany) for installing the instrumentation in the field. We also thank A. Fernández and L.A. Gómez (Garajonay National Park) for their support.

REFERENCES

  • Aboal Viñas, J., 1998: Los flujos netos hidrológicos y químicos asociados de un bosque de laurisilva en Tenerife (The hydrological and chemical net fluxes associated with a laurel forest in Tenerife). Ph.D. thesis, University of La Laguna, 184 pp. [Available online at http://dialnet.unirioja.es/servlet/tesis?codigo=637.].

  • Allen, R. G., Pereira L. S. , Raes D. , and Smith M. , 1998: Crop evapotranspiration: Guidelines for computing crop water requirements. FAO Irrigation and Drainage Paper 56, Food and Agriculture Organization of the United Nations, 300 pp.

  • Aschan, G., 1998: Microclimate, Energy Budget and Water Balance of Tropical and Extratropical Forests. (in German). Biology Series, Vol. 159, Tectum-Verlag, 256 pp.

    • Search Google Scholar
    • Export Citation
  • Axelrod, D. I., 1975: Evolution and biogeography of Madrean-Tethyan sclerophyll vegetation. Ann. Mo. Bot. Gard., 62 , 280334.

  • Azevedo, J., and Morgan D. L. , 1974: Fog precipitation in coastal California forests. Ecology, 55 , 11351141.

  • Brewer, C. A., Smith W. K. , and Vogelmann T. C. , 1991: Functional interaction between leaf trichomes, leaf wettability and the optical properties of water droplets. Plant Cell Environ., 14 , 955962.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Bruijnzeel, L. A., 2001: Hydrology of tropical montane cloud forests: A reassessment. Land Use Water Resour. Res., 1 , 1.11.18.

  • Bruijnzeel, L. A., and Proctor J. , 1995: Hydrology and biochemistry of tropical montane cloud forests: What do we really know? Tropical Montane Cloud Forests, L. S. Hamilton, J. O. Juvik, and F. N. Scatena, Eds., Ecological Studies, Vol. 110, Springer-Verlag, 38–78.

    • Search Google Scholar
    • Export Citation
  • Bruijnzeel, L. A., Waterloo M. J. , Proctor J. , Kuiters A. T. , and Kotterink B. , 1993: Hydrological observations in montane rain forests on Gunung Silam, Sabah, Malaysia, with special reference to the “Massenerhebung” effect. J. Ecol., 81 , 145167.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Brunn, R. J., Lewis W. , Perkins P. J. , and Serafini J. S. , 1955: Impingement of cloud droplets on a cylinder and procedure for measuring liquid-water content and droplet sizes in supercooled clouds by rotating multicylinder method. NASA Rep. 1215, 183 pp.

  • Brutsaert, W., 1975: Comments on surface roughness parameters and the height of dense vegetation. J. Meteor. Soc. Japan, 53 , 9697.

  • Cáceres, G., 1981: Hydrological importance of horizontal interception in a humid premontane forest in Balalaica, Turrialba, Costa Rica (in Spanish). M.S. thesis, Centro Agronómico Tropical de Investigación y Enseñanza, University of Costa Rica, 98 pp.

  • Cavelier, J., and Goldstein G. , 1989: Mist and fog interception in elfin cloud forests in Colombia and Venezuela. J. Trop. Ecol., 5 , 309322.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cavelier, J., Solis D. , and Jaramillo M. A. , 1996: Fog interception in montane forests across the Central Cordillera of Panamá. J. Trop. Ecol., 12 , 357369.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Chang, S-C., Lai I-L. , and Wu J-T. , 2002: Estimation of fog deposition on epiphytic bryophytes in a subtropical montane forest ecosystem in northeastern Taiwan. Atmos. Res., 64 , 159167.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dohrenwend, R. E., 1979: Hydrologic behavior at the top of a tropical mountain. Ford Forestry Center Research Note 29, Michigan Technological University, 14 pp.

  • Eugster, W., Burkard R. , Holwerda F. , Scatena F. N. , and Bruijnzeel L. A. , 2006: Characteristics of fog and fogwater fluxes in a Puerto Rican elfin cloud forest. Agric. For. Meteor., 139 , 288306.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Fischer, D. T., and Still C. J. , 2007: Evaluating patterns of fog water deposition and isotopic composition on the California Channel Islands. Water Resour. Res., 43 .W04420, doi:10.1029/2006WR005124.

    • Search Google Scholar
    • Export Citation
  • Førland, E. J., and Coauthors, 1996: Manual for operational correction of Nordic precipitation data. Norwegian Meteorological Institute Rep. 24/96, KLIMA, 66 pp.

  • Friedlander, S. K., 2000: Smoke, Dust, and Haze: Fundamentals of Aerosol Dynamics. Oxford University Press, 407 pp.

  • Giambelluca, T. W., and Nullet D. , 1991: Influence of the trade-wind inversion on the climate of a leeward mountain slope in Hawaii. Climate Res., 1 , 207216.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gioda, A., Maley J. , Guasp R. E. , and Baladón A. A. , 1995: Some low elevation fog forests of dry environments: Applications to African paleoenvironments. Tropical Montane Cloud Forests, L. S. Hamilton, J. O. Juvik, and F. N. Scatena, Eds., Ecological Studies, Vol. 110, Springer-Verlag, 156–164.

    • Search Google Scholar
    • Export Citation
  • Golubic, I., 2001: Vegetation analysis of laurel forest and similar stands within the research framework of the water balance in the Garajonay National Park, La Comera (Canary Islands) (in German). Diploma thesis, Department of Plant Sociology and Ecology, University of Duisburg-Essen, 107 pp.

  • González-Rodríguez, A. M., Morales D. , and Jiménez M. S. , 2002: Leaf gas exchange characteristics of a Canarian laurel forest tree species [Persea indica (L.) K. Spreng.] under natural conditions. J. Plant Physiol., 159 , 695704.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Goodman, J., 1982: Water potential from advection fog. Department of Meteorology Progress Rep. 1, San Jose State University, 45 pp.

  • Gunn, R., and Kinzer G. D. , 1949: The terminal velocity of fall for water droplets in stagnant air. J. Meteor., 6 , 243248.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Gutiérrez, A. G., and Coauthors, 2008: Regeneration patterns and persistence of the fog-dependent Fray Jorge forest in semiarid Chile during the past two centuries. Global Change Biol., 14 , 161176.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hafkenscheid, R. L. L. J., 2000: Hydrology and biogeochemistry of montane rain forests of contrasting stature in the Blue Mountains of Jamaica. Ph.D. thesis, Vrije Universiteit, 302 pp.

  • Herckes, P., Mirabel P. , and Wortham H. , 2002: Cloud water deposition at a high-elevation site in the Vosges Mountains (France). Sci. Total Environ., 296 , 5975.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holder, C. D., 2007: Leaf water repellency of species in Guatemala and Colorado (USA) and its significance to forest hydrology studies. J. Hydrol., 336 , 147154.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Höllermann, P., 1981: Microenvironmental studies in the laurel forest of the Canary Islands. Mt. Res. Dev., 1 , 193207.

  • Holwerda, F., Scatena F. N. , and Bruijnzeel L. A. , 2006: Throughfall in a Puerto Rican lower montane rain forest: A comparison of sampling strategies. J. Hydrol., 327 , 592602.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hutley, L. B., Doley D. , Yates D. J. , and Boonsaner A. , 1997: Water balance of an Australian subtropical rainforest at altitude: The ecological and physiological significance of intercepted cloud and fog. Aust. J. Bot., 45 , 311329.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ingraham, N. L., and Matthews R. A. , 1988: Fog drip as a source of groundwater recharge in northern Kenya. Water Resour. Res., 24 , 14061410.

  • Joslin, J. D., Mueller S. F. , and Wolfe M. H. , 1990: Tests of models of cloudwater deposition to forest canopies using artificial and living collectors. Atmos. Environ., 24A , 30073019.

    • Search Google Scholar
    • Export Citation
  • Juvik, J. O., and Ekern P. C. , 1978: A climatology of mountain fog on Mauna Loa, Hawaii Island. Water Resources Research Center Tech. Rep. 118, University of Hawaii, 63 pp.

  • Juvik, J. O., and Nullet D. , 1995: Comments on “A proposed standard fog collector for use in high-elevation regions”. J. Appl. Meteor., 34 , 21082110.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kämmer, F., 1974: Klima und vegetation auf Tenerife, besonders in Hinblick auf den nebelniederschlag. Scr. Geobot., 7 , 178.

  • Klemm, O., Wrzesinsky T. , and Scheer C. , 2005: Fog water flux at a canopy top: Direct measurement versus one-dimensional model. Atmos. Environ., 39 , 53755386.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Langmuir, I., and Blodgett K. B. , 1946: A mathematical investigation of water droplet trajectories. The Collected Works of Irving Langmuir, V. J. Schaefer, Ed., Pergamon Press, 348–393.

    • Search Google Scholar
    • Export Citation
  • Laws, J. O., and Parsons D. A. , 1943: The relation of raindrop-size to intensity. Eos, Trans. Amer. Geophys. Union, 24 , 452460.

  • Lloyd, C. R., Gash J. H. C. , Shuttleworth W. J. U. , Marques F. , and De O F. , 1988: The measurement and modelling of rainfall interception by Amazonian rain forest. Agric. For. Meteor., 43 , 277294.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lovett, G. M., 1984: Rates and mechanisms of cloud water deposition to a subalpine balsam fir forest. Atmos. Environ., 18 , 361371.

  • McComber, P., and Touzot G. , 1981: Calculation of the impingement of cloud droplets in a cylinder by the finite-element method. J. Atmos. Sci., 38 , 10271036.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Monteith, J., and Unsworth M. , 1990: Principles of Environmental Physics. 2nd ed. Edward Arnold, 291 pp.

  • Müller, S. F., 1991: Estimating cloud water deposition to subalpine spruce-fir forests. I: Modifications to an existing model. Atmos. Environ., 25A , 10931104.

    • Search Google Scholar
    • Export Citation
  • Müller, S. F., Joslin J. D. , and Wolfe M. H. , 1991: Estimating cloud water deposition to subalpine spruce-fir forests. II: Model testing. Atmos. Environ., 25A , 11051122.

    • Search Google Scholar
    • Export Citation
  • Padilla, H., Belmont R. , Torres M. C. , Garcia R. , and Baez A. , 1996: A field comparison of cloud water collectors in a mountainous region under low wind speed conditions in eastern Mexico. Atmósfera, 9 , 189199.

    • Search Google Scholar
    • Export Citation
  • Pérez de Paz, P. L., 1990: Garajonay National Park: World Heritage. (in Spanish). National Institute for the Conservation of Nature, Excmo. Cabildo Insular de La Gomera, 367 pp.

    • Search Google Scholar
    • Export Citation
  • Riehl, H., 1979: Climate and Weather in the Tropics. Academic Press, 611 pp.

  • Ritter, A., Regalado C. M. , and Aschan G. , 2007: An impaction model for estimating fog water collection in a subtropical laurel cloud forest of the Garajonay National Park. Proc. Fourth Int. Conf. on Fog, Fog Collection and Dew, La Serena, Chile, Pontificia Universidad Católica de Chile, 355–358.

  • Santana, L., 1986: Study of Precipitation Fog in Tenerife. (in Spanish). National Institute for the Conservation of Nature, 97 pp.

  • Schemenauer, R. S., and Cereceda P. , 1994: The role of wind in rainwater catchment and fog collection. Water Int., 19 , 7076.

  • Sevruk, B., 2005: Rainfall measurement: Gauges. Encyclopedia of Hydrological Sciences, M. G. Andersen, Ed., Vol. 1, Hydrometeorology, John Wiley, 529–536.

    • Search Google Scholar
    • Export Citation
  • Sharon, D., 1980: Distribution of hydrologically effective rainfall incident on sloping ground. J. Hydrol., 46 , 165188.

  • Shuttleworth, W. J., 1977: The exchange of wind-driven fog and mist between vegetation and the atmosphere. Bound.-Layer Meteor., 12 , 463489.

  • Shuttleworth, W. J., 1993: Evaporation. Handbook of Hydrology, D. R. Maidment, Ed., McGraw-Hill, 4.1–4.53.

  • Sperling, F. G., Washington R. , and Whittaker R. J. , 2004: Future climate change of the subtropical North Atlantic: Implications for the cloud forests of Tenerife. Climatic Change, 65 , 103123.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Stadtmüller, T., 1987: Cloud Forests in the Humid Tropics: A Bibliographic Review. United Nations University Press, 82 pp.

  • Thom, A. S., 1971: Momentum absorption by vegetation. Quart. J. Roy. Meteor. Soc., 97 , 414428.

  • Thom, A. S., 1975: Momentum, mass and heat exchange. Vegetation and the Atmosphere, J. L. Monteith, Ed., Academic Press, 57–109.

  • Thom, A. S., and Oliver H. R. , 1977: On Penman’s equation for estimating regional evaporation. Quart. J. Roy. Meteor. Soc., 103 , 345357.

  • Vermeulen, A. T., Wyers G. P. , Römer F. G. , Van Leeuwen N. F. M. , Draaijers G. P. J. , and Erisman J. W. , 1997: Fog deposition on a coniferous forest in The Netherlands. Atmos. Environ., 31 , 375386.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Vogelmann, H. W., 1973: Fog precipitation in the cloud forests of eastern Mexico. Bioscience, 23 , 96100.

  • Walmsley, J. L., Schemenauer R. , and Bridgman H. A. , 1996: A method for estimating the hydrologic input from fog in mountainous terrain. J. Appl. Meteor., 35 , 22372249.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wrzesinsky, T., Thalmann E. , Burkard R. , Eugster W. , and Klemm O. , 2001: Fog deposition of nutrients and pollutants to a montane forest site. Proc. Second Int. Conf. on Fog and Fog Collection, St. John’s, NL, Canada, Meteorological Service of Canada, 169–172.

  • Wyslouzil, B. E., Whipple M. , Chatterjee Ch , Walcerz D. B. , Weathers P. J. , and Hart D. P. , 1997: Mist deposition onto hairy root cultures: Aerosol modeling and experiments. Biotechnol. Prog., 13 , 185194.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Zadroga, F., 1981: The hydrological importance of a montane cloud forest area of Costa Rica. Tropical Agricultural Hydrology: Watershed Management and Land Use, R. Lal and E. W. Russell, Eds., John Wiley, 59–73.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

Location of (a) the experimental watershed and (b) plots. Distribution of (c) laurel forest types and (d) survey plots (marked with asterisk) within the watershed; isolines in (d) drawn every 5 m.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 2.
Fig. 2.

Fog droplet size frequency distributions for eight liquid water content classes, after Klemm et al. (2005).

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 3.
Fig. 3.

Diagram of the strategy used to calculate fog water contribution to soil by E. arborea trees from fog water collected by artificial catchers (QFCs). Bold italic font indicate estimated parameters; subscripts “fc” and “v” refer to QFC and vegetation parameters, respectively.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 4.
Fig. 4.

Scatterplots of the rain gauge and QFC hourly rainfall totals at the four plots for different wind speed intervals.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 5.
Fig. 5.

Frequency histograms of data observed at the four measurement plots for rainfall (a) intensity and (b) duration, wind (c) speed and (d) direction, and (e) fog water collection rate. From left to right: bars correspond to plots P1145, P1185, P1230, and P1270.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 6.
Fig. 6.

Monthly totals of rainfall and fog water collection at the four measurement sites during the 2-yr period. Rainfall data correspond to the average value (±std error) of the four plots. From left to right: bars correspond to plots P1145, P1185, P1230, and P1270.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 7.
Fig. 7.

Mean monthly fog water collected by artificial fog catchersin the four plots. Bars indicate 95% confidence interval.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 8.
Fig. 8.

Frequency histograms of estimates obtained with the impaction model for (a) liquid water content w and (b) fog water q collected by E. arborea tree. Bars represent average frequency values (±std error) computed assuming different fog droplet sizes (5, 10, 15, 25, 40, and 50 μm).

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Fig. 9.
Fig. 9.

Hourly evolution of micrometeorological variables measured in P1270 from February 2003 to January 2005: (a) totals of fog water captured by a single tree at each time of day for whole measurement period; (b) global radiation Rg; (c) linear regression obtained for Rg under foggy and fog-free conditions; (d) actual wind speed; (e) max wind speed; (f) air temperature; and (g) relative humidity. Dashed lines in (b) and (d)–(g) indicate hourly mean values (thick line) and ±std dev (thin lines) under fog-free conditions. Range of micrometeorological variables when fog was collected indicated by gray bands.

Citation: Journal of Hydrometeorology 9, 5; 10.1175/2008JHM992.1

Table 1.

Information of the plots selected within the watershed.

Table 1.
Table 2.

Fog collection efficiencies due to interception ηint [Eq. (9)], diffusion ηdiff [Eq. (10)], and impaction ηimp [Eq. (11)]; sedimentation settling velocities (υt) and the Stokes (Stk) and Reynolds (Re) numbers for different droplet sizes (Dg) and wind speeds (u).

Table 2.
Table 3.

Fog water collection (q), wind speed* [actual (u) and maximum (umax)], and potential wet canopy evaporation (E) observed in the two periods.

Table 3.
Table 4.

Yearly total of estimated fog water captured by E. arborea at P1270, Year 1: February 2003–January 2004 and Year 2: February 2004–January 2005.

Table 4.
Save