Fog precipitation in the Sierra de las Minas Biosphere Reserve,Guatemala

Fog precipitation occurs when fog droplets are filtered by the forest canopy and coalesce on the vegetative surfaces to form larger water droplets that drip to the forest floor. This study examines the quantity of throughfall compared with incident precipitation produced by the canopy of a lower montane rain forest (2100 m) and an upper montane cloud forest (2550 m) in the Sierra de las Minas Biosphere Reserve, Guatemala. Fog precipitation was measured with throughfall and precipitation gauges from 23 July 1995 to 7 June 1996. Fog precipitation occurred during sampling periods when throughfall exceeded incident precipitation. Fog precipitation contributed <1% of total water inputs in the cloud forest at 2100 m during the 44‐week period, whereas fog precipitation contributed 7·4% at 2550 m during the same period. The depth equivalent of fog precipitation was greater at 2550 m (203·4 mm) than at 2100 m (23·4 mm). The calculation of fog precipitation in this study is underestimated. The degree of underestimation may be evident in the difference in apparent rainfall interception between 2100 m (35%) and 2550 m (4%). Because the apparent interception rate at 2550 m is significantly lower than 2100 m, the canopy probably is saturated for longer periods as a result of cloud water contributions. Data show a seasonal pattern of fog precipitation most evident at the 2550 m site. Fog precipitation represented a larger proportion of total water inputs during the dry season (November to May). Because cloud forests generate greater than 1 mm day^?1 of fog precipitation in higher elevations of the Sierra de las Minas, the conservation of the cloud forest may be important to meet the water demands of a growing population in the surrounding arid lowlands. Copyright © 2003 John Wiley & Sons, Ltd.     

Estimating fog deposition at a Puerto Rican elfin cloud forest site: comparison of the water budget and eddy covariance methods

The deposition of fog to a wind‐exposed 3 m tall Puerto Rican cloud forest at 1010 m elevation was studied using the water budget and eddy covariance methods. Fog deposition was calculated from the water budget as throughfall plus stemflow plus interception loss minus rainfall corrected for wind‐induced loss and effect of slope. The eddy covariance method was used to calculate the turbulent liquid cloud water flux from instantaneous turbulent deviations of the surface‐normal wind component and cloud liquid water content as measured at 4 m above the forest canopy. Fog deposition rates according to the water budget under rain‐free conditions (0·11 ± 0·05 mm h^?1) and rainy conditions (0·24 ± 0·13 mm h^?1) were about three to six times the eddy‐covariance‐based estimate (0·04 ± 0·002 mm h^?1). Under rain‐free conditions, water‐budget‐based fog deposition rates were positively correlated with horizontal fluxes of liquid cloud water (as calculated from wind speed and liquid water content data). Under rainy conditions, the correlation became very poor, presumably because of errors in the corrected rainfall amounts and very high spatial variability in throughfall. It was demonstrated that the turbulent liquid cloud water fluxes as measured at 4 m above the forest could be only ～40% of the fluxes at the canopy level itself due to condensation of moisture in air moving upslope. Other factors, which may have contributed to the discrepancy in results obtained with the two methods, were related to effects of footprint mismatch and methodological problems with rainfall measurements under the prevailing windy conditions. Best estimates of annual fog deposition amounted to ～770 mm year^?1 for the summit cloud forest just below the ridge top (according to the water budget method) and ～785 mm year^?1 for the cloud forest on the lower windward slope (using the eddy‐covariance‐based deposition rate corrected for estimated vertical flux divergence). Copyright © 2006 John Wiley & Sons, Ltd.     

Calculated and measured values of liquid water content in clean and polluted environments

The relationship between liquid water content (LWC) and visibility (VIS) in a fog was examined by a field measurement of LWC values at a clean location of the meteorological observatory Sodankylä (Finland) and in a polluted region at the meteorological observatory Mile?ovka (Czech Republic). Furthermore, the use of a simple regression fog model to determine LWC from VIS is examined by comparing well-known relationships with the measurement results. The results show that the verbal characterization of environment (clean, mild polluted, polluted) is insufficient to calculate the LWC values from visibility. It is necessary to establish an additional criterion based on quantitative fog characteristics or to determine the site-related equation from the previous LWC and VIS measurement.     

How much does dry‐season fog matter? Quantifying fog contributions to water balance in a coastal California watershed

The seasonally‐dry climate of Northern California imposes significant water stress on ecosystems and water resources during the dry summer months. Frequently during summer, the only water inputs occur as non‐rainfall water, in the form of fog and dew. However, due to spatially heterogeneous fog interaction within a watershed, estimating fog water fluxes to understand watershed‐scale hydrologic effects remains challenging. In this study, we characterized the role of coastal fog, a dominant feature of Northern Californian coastal ecosystems, in a San Francisco Peninsula watershed. To monitor fog occurrence, intensity, and spatial extent, we focused on the mechanisms through which fog can affect the water balance: throughfall following canopy interception of fog, soil moisture, streamflow, and meteorological variables. A stratified sampling design was used to capture the watershed's spatial heterogeneities in relation to fog events. We developed a novel spatial averaging scheme to upscale local observations of throughfall inputs and evapotranspiration suppression and make watershed‐scale estimates of fog water fluxes. Inputs from fog water throughfall (10–30 mm/year) and fog suppression of evapotranspiration (125 mm/year) reduced dry‐season water deficits by 25% at watershed scales. Evapotranspiration suppression was much more important for this reduction in water deficit than were direct inputs of fog water. The new upscaling scheme was analyzed to explore the sensitivity of its results to the methodology (data type and interpolation method) employed. This evaluation suggests that our combination of sensors and remote sensing allows an improved incorporation of spatially‐averaged fog fluxes into the water balance than traditional interpolation approaches.     

The impact of climate change on water fluxes in a Macaronesian cloud forest

Fog phenomena and their associated meteorological variables were continuously monitored during 4 years in an evergreen laurisilva cloud forest of the Anaga Massif Biosphere Reserve (Tenerife, Canary Islands), in order to establish its current dynamics. Fog was more frequent during night through early morning and in the afternoon, and particularly from May until September, coincidental with a frequent immersion of the 1025 m a.s.l. experimental site in the cloud layer of wind‐driven stratocumulus. The concomitant meteorological conditions during different fog regimes, characterized according to visibility (Ω) ranges, were compared with those when fog was absent. The presence of fog was associated with a significant reduction in global solar radiation, R_g, increased wind speed, and lower and more stable ambient temperatures. The foggy versus fog‐free hourly medians of R_g were found to be linearly related, whereas the proportion of median R_g reduction due to fog varied logarithmically with Ω. However, foggy versus fog‐free extreme values of the hourly R_g distributions departed from such a linear trend. By contrast, hourly temperatures during foggy versus fog‐free periods behaved linearly for most of the Ω range, except for very dense fog, Ω ≤ 100 m. Transpiration of the canopy, intermittently wetted due to interception of both rain and fog water droplets, was determined by quantifying the water balance at leaf scale with a mathematical model for the two representative hypostomatous species present at the site: the arboreal shrub Erica platycodon, with needle‐like leaves, and the laurophyll tree Myrica faya. Both tree transpiration and evaporation of the intercepted fog water were predictively higher during summer. By contrast, transpiration was reduced during February, in agreement with a 1 year period of sap velocity measurements, and was not appreciably affected by soil moisture content. The consequences of an anticipated downward shift of the stratocumulus cloud layer and of various projected Representative Concentration Pathways (RCPs) scenarios in the Macaronesian area were simulated, yielding in all cases a significant rise in transpiration for both species. Particularly, the simulated RCPs scenarios implied 29%–73% increments in transpiration from the actual values. Because fog is concomitant with lower temperatures and vapour pressure deficit, the modification of its current distribution as a consequence of climate change may have a direct effect on such associated meteorological variables, and therefore a meaningful impact in the water relations of the laurel cloud forests.     

Rainfall interception by an isolated evergreen oak tree in a Mediterranean savannah

Redistribution of ground‐level rainfall and interception loss by an isolated Quercus ilex tree were measured over 2 years in a Mediterranean oak savannah. Stemflow, meteorological variables and sap flow were also monitored. Rainfall at ground level was measured by a set of rain‐gauges located in a radial layout centred on the tree trunk and extending beyond the crown limits. Interception loss was computed as the difference between the volume of rainwater that would reach the ground in the absence of the tree and the volume of water that actually fell on the ground sampling area (stemflow included). This procedure provided correct interception loss estimates, irrespective of rainfall inclination. Results have shown a clear non‐random spatial distribution of ground‐level rainfall, with rainwater concentrations upwind beneath the crown and rain‐shadows downwind. Interception loss amounted to 22% of gross rainfall, per unit of crown‐projected area. Stand interception loss, per unit of ground area, was only 8% of gross rainfall and 28% of tree evapotranspiration. These values reflect the low crown cover fraction of the stand (0·39) and the specific features of the Mediterranean rainfall regime (predominantly with few large storms). Nevertheless, it still is an important component of the water balance of these Mediterranean ecosystems. Copyright © 2005 John Wiley & Sons, Ltd.     

Contribution of occult precipitation to the water balance of páramo ecosystems in the Colombian Andes

Páramos are high‐altitudinal neotropical ecosystems located in the upper regions of the northern Andes. Their hydrology is characterized by an extraordinarily high run‐off ratio. One major contributing mechanism is thought to be fog occurrence, which is common in the páramos and occurs by the cooling of near‐surface moist air, as it is forced to higher elevations by topography. However, field‐based observations and quantification of this flux are rare. We present results of monitoring of occult precipitation, understood as the combination of fog and drizzle inputs, combined with meteorological and soil moisture monitoring for periods between 7 to 17 months in 6 sites distributed over 3 páramos catchments in Colombia: three sites in Romerales (Quindío), two in Chingaza (Cundinamarca), and one in Belmira (Antioquia). Occult precipitation inputs were measured with cylindrical fog gauges with a cover on top. We estimate occult precipitation inputs to add between 7% and 28% to rainfall inputs in the study sites. Our results also show that occult precipitation has a large temporal and spatial variability, both within one site and between sites, which make it difficult to upscale and quantify at a catchment scale. Nevertheless, occult precipitation can be important for downstream water supply given that these inputs are especially concentrated during periods with low rainfall. Lastly, we also find evidence for an increase in soil moisture related to occult precipitation during a dry period in Romerales páramo.     

Rainfall partitioning into throughfall,stemflow, and interception within a single beech (Fagus sylvatica L.) canopy: influence of foliation,rain event characteristics,and meteorology

While the hydrological balance of forest ecosystems has often been studied at the annual level, quantitative studies on the factors determining rainfall partitioning of individual rain events are less frequently reported. Therefore, the effect of the seasonal variation in canopy cover on rainfall partitioning was studied for a mature deciduous beech (Fagus sylvatica L.) tree over a 2‐year period. At the annual level, throughfall amounted to 71% of precipitation, stemflow 8%, and interception 21%. Rainfall partitioning at the event level depended strongly on the amount of rainfall and differed significantly (p < 0·001) between the leafed and the leafless period of the year. Therefore, water fluxes of individual events were described using a multiple regression analysis (r_a² > 0·85, n = 205) with foliation, rainfall characteristics and meteorological variables as predictor variables. For a given amount of rainfall, foliation significantly increased interception and decreased throughfall and stemflow amounts. In addition, rainfall duration, maximum rainfall rate, vapour pressure deficit, and wind speed significantly affected rainfall partitioning at the event level. Increasing maximum hourly rainfall rate increased throughfall and decreased stemflow generation, while higher hourly vapour pressure deficit decreased event throughfall and stemflow amounts. Wind speed decreased throughfall in the growing period only. Since foliation and the event rainfall amount largely determined interception loss, the observed net water input under the deciduous canopy was sensitive to the temporal distribution of rainfall. Copyright © 2007 John Wiley & Sons, Ltd.     

Summary of a 4-Year Fog Field Study in Northern Nanjing, Part 1: Fog Boundary Layer

Comprehensive fog field observations were conducted during the winters of 2006–2009 at the Nanjing University of Information Science and Technology to study the macro and micro-physical structures and the physical–chemical processes of dense fogs in the area. The observations included features of the fog boundary layer, characteristics of fog water, the particle spectrum, the chemical composition of atmospheric aerosols, radiation and heat components, turbulence, meteorological elements (air temperature, pressure, wind speed, wind direction), and environmental monitoring. The fogs observed were divided into four types: radiation fog, advection–radiation fog, advection fog, and precipitation fog, according to the mechanisms and primary factors of the fog processes. Fog boundary-layer structures of different types and their corresponding characteristics were then studied. Fog boundary-layer features, temperature structures, wind fields, and fog maintenance are discussed. The results show that radiation fog had remarkable diurnal variation and formed mostly at sunset or midnight, and lifted after sunrise or at noon, and that advection–radiation fog and advection fog were of very long duration. Extremely dense fogs occurred only in radiation-related cases. Inversion in radiation fog was short-lived, disappearing 1 or 2 hours after sunrise or at noon, faster than that in advection–radiation fog. When wind direction reversed from easterly to westerly or from southerly to northerly, the fog became an extremely dense fog. Low-level jet at times impeded fog development, whereas at other times it encouraged fog continuance. The deep inversion was merely an essential condition for a thick fog layer; sufficient vapor supply was advantageous to the formation and maintenance of a deep fog layer.     

Spherical gauges for improving the accuracy of rainfall measurements

Rain‐gauge catch efficiencies are affected by wind. Wind makes raindrops fall at an angle of inclination and the effective diameter of the rain gauge orifice smaller than if raindrops fall into the gauge vertically. Two spherical and two semi‐spherical orifices were designed to modify standard gauges and others in use today. The two spherical orifices catch rain with an effective diameter always equal to the actual diameter regardless of wind speed and direction. The semi‐spherical orifices, used side‐by‐side with a standard gauge, correct 50% of catch deficiencies made by the standard gauge. Tests based on 115 storms show that the four new gauges caught more rainfall than the standard gauge, with an average catch increase ranging from 8% to 16%. Compared with the pit gauge, average deficiency in catch ranged from ?1% (spherical rain gauge orifice with cylinders) to 4%, whereas the deficiency for the standard gauge was ?10%. Percentage deficiencies of the new gauges were positively affected by wind speed, raindrop inclination and rainfall intensity. Although the new gauges tended to underestimate the standard gauge in small storms (<0·25 cm) and overestimated the pit gauge under strong winds, their deviations are small. Underestimates for small storms could be improved by using gauge materials that reduce surface temperature, evaporation and water retention. The gauges are simple in design, easy to operate and inexpensive. In order to maintain a historically consistent set of rainfall data, a dual‐gauge (standard gauge + spherical gauge) is recommended for existing rainfall stations. The new rain gauge orifices are suitable for large‐scale applications. Copyright © 2001 John Wiley & Sons, Ltd.     

Winter rainfall interception by two mature open‐grown trees in Davis,California

A rainfall interception measuring system was developed and tested for open‐grown trees. The system includes direct measurements of gross precipitation, throughfall and stemflow, as well as continuous collection of micrometeorological data. The data were sampled every second and collected at 30‐s time steps using pressure transducers monitoring water depth in collection containers coupled to Campbell CR10 dataloggers. The system was tested on a 9‐year‐old broadleaf deciduous tree (pear, Pyrus calleryana ‘Bradford’) and an 8‐year‐old broadleaf evergreen tree (cork oak, Quercus suber) representing trees having divergent canopy distributions of foliage and stems. Partitioning of gross precipitation into throughfall, stemflow and canopy interception is presented for these two mature open‐grown trees during the 1996–1998 rainy seasons. Interception losses accounted for about 15% of gross precipitation for the pear tree and 27% for the oak tree. The fraction of gross precipitation reaching the ground included 8% by stemflow and 77% by throughfall for the pear tree, as compared with 15% and 58%, respectively, for the oak tree. The analysis of temporal patterns in interception indicates that it was greatest at the beginning of each rainfall event. Rainfall frequency is more significant than rainfall rate and duration in determining interception losses. Both stemflow and throughfall varied with rainfall intensity and wind speed. Increasing precipitation rates and wind speed increased stemflow but reduced throughfall. Analysis of rainfall interception processes at different time‐scales indicates that canopy interception varied from 100% at the beginning of the rain event to about 3% at the maximum rain intensity for the oak tree. These values reflected the canopy surface water storage changes during the rain event. The winter domain precipitation at our study site in the Central Valley of California limited our opportunities to collect interception data during non‐winter seasons. This precipitation pattern makes the results more specific to the Mediterranean climate region. Copyright © 2000 John Wiley & Sons, Ltd.     

The role of fog,orography, and seasonality on precipitation in a semiarid,tropical island

Isotopes of water (²H/¹H and ¹⁸O/¹⁶O) are commonly used to trace hydrological processes such as moisture recycling, evaporation loss, and moisture source region and often vary temporally in a given region. This study provides a first‐ever characterization of temporally variable precipitation mechanisms of San Cristóbal Island, Galápagos. We collected fog, rain, and throughfall samples over three field seasons to understand the mechanisms driving seasonal‐ and event‐based variability in the isotopic composition of precipitation in Galápagos. We establish that fog is a common phenomenon in San Cristóbal, especially during the dry season, and we found that fog, compared with cocollected rainfall, is consistently enriched. We further suggest that the relative contribution of fog formed via different mechanisms (orographic, advective, radiation) varied seasonally. We found that the source region is the most dominant control of the isotopic composition of rainfall in the Galápagos at both the seasonal and event scales, but subcloud evaporative processes (the nontraditional manifestation of the amount effect) became a dominant control on the isotopic composition of rainfall during the dry season. Overall, our findings suggest that understanding seasonally variable water‐generating mechanisms is required for effective water resource management in San Cristóbal Island and other semiarid island ecosystems under current and future regimes of climate change.     

The Relation Between Humidity and Liquid Water Content in Fog: An Experimental Approach

Microphysical measurements of orographic fog were performed above a montane cloud forest in northeastern Taiwan (Chilan mountain site). The measured parameters include droplet size distribution (DSD), absolute humidity (AH), relative humidity (RH), air temperature, wind speed and direction, visibility, and solar short wave radiation. The scope of this work was to study the short term variations of DSD, temperature, and RH, with a temporal resolution of 3?Hz. The results show that orographic fog is randomly composed of various air volumes that are intrinsically rather homogeneous, but exhibit clear differences between each other with respect to their size, RH, LWC, and DSD. Three general types of air volumes have been identified via the recorded DSD. A statistical analysis of the characteristics of these volumes yielded large variabilities in persistence, RH, and LWC. Further, the data revealed an inverse relation between RH and LWC. In principle, this finding can be explained by the condensational growth theory for droplets containing soluble or insoluble material. Droplets with greater diameters can exist at lower ambient RH than smaller ones. However, condensational growth alone is not capable to explain the large observed differences in DSD and RH because the respective growth speeds are too slow to explain the observed phenomena. Other mechanisms play key roles as well. Possible processes leading to the large observed differences in RH and DSD include turbulence induced collision and coalescence, and heterogeneous mixing. More analyses including fog droplet chemistry and dynamic microphysical modeling are required to further study these processes. To our knowledge, this is the first experimental field observation of the anti-correlation between RH and LWC in fog.     

Quantification of cloud water interception in the canopy vegetation from fog gauge measurements

With changes in climate looming, quantifying often‐overlooked components of the canopy water budget, such as cloud water interception (CWI), is increasingly important. Commonly, CWI quantification requires detailed continuous measurements, which is extremely challenging, especially when throughfall is included. In this study, we propose a simplified approach to estimate CWI using the Rutter‐type interception model, where CWI inputs in the canopy vegetation are proportional to fog interception measured by an artificial fog gauge. The model requires the continuous acquisition of meteorological variables as input and calibration datasets. Throughfall measurements below the forest are used only for calibration and validation of the model; thus, CWI estimates can be provided even after the cessation of throughfall monitoring. This approach provides an indirect and undemanding way to quantify CWI by vegetation and allows the identification of its controlling factors, which could be useful to the comparison of CWI in contrasting land covers. The method is applied on a 2‐year dataset collected in an endemic highland forest of San Cristobal Island (Galapagos). Our results show that CWI reaches 21% ± 6% of the total water input during the first year, and 9% ± 2% during the second one. These values represent 32% ± 10% and 17% ± 5% of water inputs during the cool foggy season of the first and second year, respectively. The difference between seasons is attributed to a lower fog liquid water during the second season.     

Field assessments on the accuracy of spherical gauges in rainfall measurements

In an effort to reduce wind effect on rainfall catch to a minimum level, Chang and Flannery (2001. Hydrological Processes 15 : 643–654) designed two spherical orifices to modify the standard gauge and other gauges in use today. Because of the spherical shape, the two orifices will catch rain with an effective diameter always equal to the actual diameter, regardless of wind speed and direction. This report covers the testing of spherical gauges at two different locations, one at the City Landfill, Nacogdoches, TX, and the other at the NWS Forecast Office, Shreveport, LA. Based on 131 storms at Nacogdoches and 94 storms at Shreveport, observed between May 1998 and February 2001, the results showed: (1) spherical gauges recorded an average 6–9% greater than standard gauge and 3–4% less than pit gauge, only 1–2% less than reported in the original study; (2) the catch of spherical gauges was not significantly affected by three gauge heights at 0·91, 1·83, and 2·74 m above the ground, but catch by the standard gauge decreased with increasing gauge height; (3) improvements of the spherical gauges were most significant for larger storms and for winds at higher speeds; (4) the spherical gauge with cylinders recorded 1–2% more rainfall than the spherical gauge with vanes; and (5) correlation coefficients between catch deficiencies and wind speed were low and weak because of the distance and height of the existing wind sensor. Owing to greater surface wetting and evaporation loss, the spherical gauges may underestimate rainfall catch by standard gauge for small storms (generally less than 5·0 mm), especially on hot summer afternoons and for smaller storms. However, the underestimates do not overshadow the merits of spherical gauges, because the differences are too small to be of hydrologic significance. Using polyethylene or other synthesized materials to construct spherical orifices may improve the catch for small storms. The results of the study agreed with the previous claims that spherical gauges are effective in reducing wind effects on rainfall measurements. The spherical gauges could greatly improve the accuracy of hydrologic simulations and the efficiency on the designs and management of water resources. They are suitable for large‐scale applications. Copyright © 2004 John Wiley & Sons, Ltd.     

Roundhouse (RND) Mountain Top Research Site: Measurements and Uncertainties for Winter Alpine Weather Conditions

The objective of this work is to better understand and summarize the mountain meteorological observations collected during the Science of Nowcasting Winter Weather for the Vancouver 2010 Olympics and Paralympics (SNOW-V10) project that was supported by the Fog Remote Sensing and Modeling (FRAM) project. The Roundhouse (RND) meteorological station was located 1,856 m above sea level that is subject to the winter extreme weather conditions. Below this site, there were three additional observation sites at 1,640, 1,320, and 774 m. These four stations provided some or all the following measurements at 1 min resolution: precipitation rate (PR) and amount, cloud/fog microphysics, 3D wind speed (horizontal wind speed, U _h; vertical air velocity, w _a), visibility (Vis), infrared (IR) and shortwave (SW) radiative fluxes, temperature (T) and relative humidity with respect to water (RH_w), and aerosol observations. In this work, comparisons are made to assess the uncertainties and variability for the measurements of Vis, RH_w, T, PR, and wind for various winter weather conditions. The ground-based cloud imaging probe (GCIP) measurements of snow particles using a profiling microwave radiometer (PMWR) data have also been shown to assess the icing conditions. Overall, the conclusions suggest that uncertainties in the measurements of Vis, PR, T, and RH can be as large as 50, >60, 50, and >20 %, respectively, and these numbers may increase depending on U _h, T, Vis, and PR magnitude. Variability of observations along the Whistler Mountain slope (~500 m) suggested that to verify the models, model space resolution should be better than 100 m and time scales better than 1 min. It is also concluded that differences between observed and model based parameters are strongly related to a model’s capability of accurate prediction of liquid water content (LWC), PR, and RH_w over complex topography.     

The Impact of Vertical Resolution in the Explicit Numerical Forecasting of Radiation Fog: A Case Study

Numerical experiments are performed with a comprehensive one-dimensional boundary layer/fog model to assess the impact of vertical resolution on explicit model forecasts of an observed fog layer. Two simulations were performed, one using a very high resolution and another with a vertical grid typical of current high-resolution mesoscale models. Both simulations were initialized with the same profiles, derived from observations from a fog field experiment. Significant differences in the onset and evolution of fog were found. The results obtained with the high-resolution simulation are in overall better agreement with available observations. The cooling rate before the appearance of fog is better represented, while the evolution of the liquid water content within the fog layer is more realistic. Fog formation is delayed in the low resolution simulation, and the water content in the fog layer shows large-amplitude oscillations. These results show that the numerical representation of key thermo-dynamical processes occurring in fog layers is significantly altered by the use of a grid with reduced vertical resolution.     

POTENTIAL OF EXTENDING THE RAINFALL INTENSITY–DURATION–FREQUENCY RELATIONSHIP TO NON-RECORDING RAIN GAUGES

Maximum rainfall intensity–duration–frequency (IDF) curves are commonly applied to determine the design rainfall in water resource projects. Normally, the IDF relationship is derived from recording rain gauges. As the network of non-recording rain gauges (daily rainfall) in Taiwan has a higher density than recording rain gauges, attempts were made in this study to extend the IDF relationship to non-recording rain gauges. Eighteen recording rain gauges and 99 non-recording rain gauges over the Chi-Nan area in Southern Taiwan provide the data sets. The regional IDF formulae were generated for ungauged areas to estimate rainfall intensity for various return periods and rainfall durations larger than or equal to one hour. For rainfall durations less than one hour, a set of adjustment formulae were applied to modify the regional IDF formulae. The method proposed in this study had reasonable application to non-recording rain gauges, which was concluded from the verification of four additional recording rain gauges. © 1997 John Wiley & Sons, Ltd.     

A multiplier-based method of generating stochastic areal rainfall from point rainfalls

Catchment modelling for water resources assessment is still mainly based on rain gauge measurements as these are more easily available and cover longer periods than radar and satellite-based measurements. Rain gauges however measure the rain falling on an extremely small proportion of the catchment and the areal rainfall obtained from these point measurements are consequently substantially uncertain. These uncertainties in areal rainfall estimation are generally ignored and the need to assess their impact on catchment modelling and water resources assessment is therefore imperative. A method that stochastically generates daily areal rainfall from point rainfall using multiplicative perturbations as a means of dealing with these uncertainties is developed and tested on the Berg catchment in the Western Cape of South Africa. The differences in areal rainfall obtained by alternately omitting some of the rain gauges are used to obtain a population of plausible multiplicative perturbations. Upper bounds on the applicable perturbations are set to prevent the generation of unrealistically large rainfall and to obtain unbiased stochastic rainfall. The perturbations within the set bounds are then fitted into probability density functions to stochastically generate the perturbations to impose on areal rainfall. By using 100 randomly-initialized calibrations of the AWBM catchment model and Sequent Peak Analysis, the effects of incorporating areal rainfall uncertainties on storage-yield-reliability analysis are assessed. Incorporating rainfall uncertainty is found to reduce the required storage by up to 20%. Rainfall uncertainty also increases flow-duration variability considerably and reduces the median flow-duration values by an average of about 20%.