Energy balance comparison of the Hartheim forest and an adjacent grassland site during the HartX experiment

Summary Energy balance components over a grassland surface were compared to those obtained above an adjacent, uniform Scots pine plantation during a five-day period of fine, sunny, spring weather. Soils were judged to contain ample water. Shortwave and total radiation flux densities were measured at both sites with pyranometers and total pyrradiometers. Soil heat flux densities were measured with heat flux plates at both sites, and additional storage changes were estimated for air and canopy at the forest site. The forest gained more shortwave energy than the grassland during daytime because of its lower albedo, but it lost more longwave radiation at night. The turbulent fluxes of sensible and latent energy were evaluated with the Bowen ratio energy balance (BREB) method at both sites. Temperature and humidity gradients were measured with fixed psychrometers at the grassland site, and with interchanging psychrometers at the forest site. Mean daily evapotranspiration (ET) averaged 2.26 mm over the five days for the Scots pine, or only 57 percent of the 3.94 mm measured at the grassland site. The mean Bowen ratios were 2.6 and 0.8, respectively.An error analysis was carried out for the BREB estimates of latent heat flux at the two sites. For a given error in latent heat flux and at a specified Bowen ratio the demands on accuracy of dry- and wet-bulb temperature gradients above the rough forest canopy was found to be 10 times higher than above the smoother grassland. If additionally the observed differences in transpiration rates between the two sites were taken into account, the precision for temperature gradient measurements above the slowly transpiring forest becomes fortyfold greater than required above the rapidly transpiring grass. At present, BREB precision requirements for gradients above rougher, drier canopies appear achievable only through use of specialized instrumentation, such as measurement systems that incorporate interchangeable psychrometers into their design.With 9 Figures     

Continuous measurement of aerodynamic and alfalfa canopy resistances using the Bowen ratio-energy balance and Penman-Monteith methods

To investigate the alfalfa crop response to environmental factors, a Bowen ratio-energy balance method was used to evaluate short-term alfalfa canopy resistance. Continuous evapotranspiration (ET _a ) and the aerodynamic resistance (r _a ) for an alfalfa crop in each 20-min interval were calculated. Using the calculated ET _a and r _a and the Penman-Monteith approach, the bulk stomatal or actual canopy resistance (r _c ) was evaluated. The continuous 20-min resistances were computed for clear and partially cloudy sky conditions, and different average crop heights. The results show that this technique can satisfactorily be used to study the manner in which the aerodynamic and canopy resistances respond to short-term variations in weather elements such as photosynthetically active radiation (PAR), wind speed and atmospheric saturation vapor deficit.Research Assistant Professor and Assistant Utah State Climatologist, Research Associate Professor and Research Assistant, respectively.     

Aerodynamic and canopy resistance of short-rotation forest in relation to leaf area index and climate

The aerodynamic and canopy resistances of a willow short-rotation stand were estimated during the course of a growing season on the basis of micrometeorological measurements. The normalized roughness length (z ₀/h) decreased from about 0.10 at a leaf area index of one to 0.05 at a leaf area index of seven. This implies that the aerodynamic resistance at peak leaf area index is more than twice the value at zero leaf area index, all other variables unchanged. The canopy resistance depended strongly on air water concentration deficit and on leaf area index. The Lohammar equation showed good agreement between estimated and measured canopy resistances over the whole course of leaf development. The stand was well-coupled to the atmosphere only for very small values of leaf area indices, less than one, and it was practically de-coupled for leaf area indices above two. From the point of view of factors controlling evaporation, this type of stand acts as a traditional forest at the beginning and end of the season and as an agricultural crop in the middle of the season.     

Evapotranspiration and canopy resistance of grass in a Mediterranean region

Summary A simple method for estimating actual evapotranspiration (ET) could become a suitable tool for irrigation scheduling. Resistance models can be useful if data on canopy resistance to water vapor flow (rc) and on aerodynamic resistance (ra) are available. These parameters are complex and hard to obtain. In this studyrc is analysed for a reference crop (grass meadow). Canopy resistance is dependent on climate, weather (radiation, atmospheric vapor pressure deficit, aerodynamic resistance), agronomic practices (irrigation, grass cutting) and time scale (hour, day). Anrc model, proposed by Katerji and Perrier (KP model), using some meteorological parameters as inputs, is presented. Canopy resistance calculated according to the KP model was used to estimate a referenceET _ref on hourly and daily time scales.TheET _ref estimated using the KP model on a daily time scale was compared with a model proposed by Allen, Jensen, Wright and Burman (AJWB model) — in whichrc depends on leaf area index only — and with direct measurements from a weighing lysimeter. The results show an underestimation of 18% for the AJWB model against an underestimation of 2% for the KP model. Since the hypotheses are the same for both models and aerodynamic resistance plays a secondary role, the better results obtained by the KP model are due torc modelling.With 11 Figures     

Practical use of Phragmites australis to study evapotranspiration in a wetland zone of Lake Balaton (southwest Hungary)

In the present study, evapotranspiration (ET) data from a common reed-dominated wetland and its meteorological controls was analysed using measured ET (ET _m) in compensation evapotranspirometers. Six seasons in the time period between 2003 and 2012 were assessed with the objective of converting theoretical observations into long-term practical use. They reveal the effects of annual fluctuations and allow for a more exact understanding of the results of ET losses, which remain an elusive and substantial part of the hydrologic budget particularly in wetland habitats. Daily measured ET rates were strongly influenced by weather variables causing considerable variation of ET characteristics between the two distinguished season types. The results of multiple stepwise regression analysis showed that the major meteorological elements impacting the sum of seasonal ET was much higher in the warm growing seasons (857 mm), due to increased available energy for ET, than in the cool season (385 mm). The sum of average ET totalled 778.6 mm over measurements. A simplified water budget analysis confirmed that adequate water volume, caused by precipitation, entered the Kis-Balaton wetland (KBW) area during the cool season. Conversely, in warm seasons, only 21.5 % of total ET resulted from rainfall, accentuating its seasonality in wetland. This information about annual variability of long-term ET values would assist in finding an ideal solution for determining the proper water level needed. The current balance of habitat types in wetland should be permanently assessed by selection of the suitable water level in order to sustain the most appropriate wetland ecological conditions.

     

Sensitivity of ADOM dry deposition velocities to input parameters: A comparison with measurements for SO2 and NO2 over three land use types

Abstract

Dry deposition velocity measurements of SO₂ and NO₂ over a deciduous forest, a carrot field and a snow surface are compared with estimates obtained from the dry deposition module in the regional Eulerian Acid Deposition and Oxidant Model (ADOM). The comparison with measurements taken in the fall and winter shows large model overestimates, sometimes as large as a factor of 5. The NO₂ estimates are particularly poor and support existing evidence that models that employ the constant flux assumption for NO₂ are inadequate. The canopy and the snow surface resistances are the largest contributors to the total resistances for SO₂ and NO₂, except for situations in which some of the snow turns into liquid water, when the aerodynamic resistance becomes important.

Increasing the magnitudes, taken from measurements, of the ADOM original values for the stomatal, cuticle, ground and snow resistances and decreasing the NO₂ mesophyll resistance and the Leaf Area Index (LAI) yield improved model results, particularly for SO₂, reducing the error by almost a factor of 5 at times. The new estimates compare favourably with those from a model that includes Wesely's canopy resistance parametrization. Over snow the NO₂ estimates are improved by as much as a factor of 6. Observed deposition velocities for SO₂ vary from 0 to 0.65 cm s^?2 over a deciduous forest, 0 to 0.60 cm s^?2 over a carrot field and are generally less than 0.05 cm s^?2 over snow.     

The May–October energy budget of a Scots pine plantation at Hartheim,Germany

Summary This paper describes measurements of the Hartheim forest energy budget for the 157-day period of May 11 – Oct. 14, 1992. Data were collected as 30-min means. Energy available to the forest was measured with net radiometers and soil heat flux discs; sensible heat exchange between the canopy and atmosphere was measured with two One-Propeller Eddy Correlation (OPEC) systems, and latent energy (evapotranspiration orET) was determined as a residual in the surface energy balance equation. Net rediation, change in thermal storage, and sensible heat flux were verified by independent measurements during the Hartheim Experiment (HartX, May 11–12), and again during the HartX2 experiment over 20 days late in the summer (Sep. 10–29). Specifically, sensible heat estimates from the two adjacent OPEC sensor sets were in close agreement throughout the summer, and in excellent agreement with measurements of sonic eddy correlation systems in May and September. The eddy correlation/energy balance technique was observed to overestimate occurrence of dew, leading to an underestimate of dailyET of about 5%. After taking dew into account, estimates of OPECET totaled 358 mm over the 5.1-month period, which is in quite good agreement with an ET estimate of 328 mm from a hydrologic water balance. An observed decrease in forestET in July and August was clearly associated with low rainfall and increased soil water deficit. The OPEC system required only modest technical supervision, and generated a data yield of 99.5% over the period DOY 144–288. The documented verification and precision of this energy budget appears to be unmatched by any other long-term forest study reported to date.With 9 Figures     

A lagrangian random-walk model for simulating water vapor,CO2 and sensible heat flux densities and scalar profiles over and within a soybean canopy

An integrated canopy micrometeorological model is described for calculating CO₂, water vapor and sensible heat exchange rates and scalar concentration profiles over and within a crop canopy. The integrated model employs a Lagrangian random walk algorithm to calculate turbulent diffusion. The integrated model extends previous Lagrangian modelling efforts by employing biochemical, physiological and micrometeorological principles to evaluate vegetative sources and sinks. Model simulations of water vapor, CO₂ and sensible heat flux densities are tested against measurements made over a soybean canopy, while calculations of scalar profiles are tested against measurements made above and within the canopy. The model simulates energy and mass fluxes and scalar profiles above the canopy successfully. On the other hand, model calculations of scalar profiles inside the canopy do not match measurements.The tested Lagrangian model is also used to evaluate simpler modelling schemes, as needed for regional and global applications. Simple, half-order closure modelling schemes (which assume a constant scalar profile in the canopy) do not yield large errors in the computation of latent heat (LE) and CO₂ (F _c ) flux densities. Small errors occur because the source-sink formulation of LE andF _c are relatively insensitive to changes in scalar concentrations and the scalar gradients are small. On the other hand, complicated modelling frames may be needed to calculate sensible heat flux densities; the source-sink formulation of sensible heat is closely coupled to the within-canopy air temperature profile.     

Die effektive Temperatur unter dem Aspekt „Schwüle” in 2 m und 175 m über Grund

Zusammenfassung Aus zehnjährigen kontinuierlichen Messungen der Lufttemperatur und der relativen Luftfeuchtigkeit in Karlsruhe-Rheinhafen (2 m ü. Gr.) und Karlsruhe-Turmberg (175 m ü. Gr.) werden mittlere Tages- und Jahresgänge der effektiven TemperaturET berechnet.Die Auszählungen der Fälle mitET24,0°C (Schwülebereich) ergeben in beiden Höhen die größten Werte im Juli; die tageszeitlichen Maxima liegen dabei zwischen 15 und 18 Uhr und die Minima zwischen 3 und 6 Uhr. Von November bis März wurde dagegen je Dreistundenintervall die durchET24,0°C festgelegte Schwülegrenze nie überschritten.

---

The effective temperature from the point of sultriness in 2 and 175 m over ground

Summary For every three hours the mean effective temperatureET is computed from the air temperature and the relative air humidity which are measured at Karlsruhe-Rheinhafen (2 m o. g.) and at Karlsruhe-Turmberg (175 m o. g.). From theseET-data mean diurnal and annual variations ofET are computed for ten years (1963 to 1972).At the two heights the greatest frequencies ofET24.0°C (region of sultriness) are in July with maximum between 15 and 18 o'clock and minimum between 3 and 6 o'clock. From November to March however the sultriness-boundary determined byET24.0°C never occured.

---

---

Mit 2 Abbildungen     

Comparison of the Bowen ratio-energy balance and stability-corrected aerodynamic methods for measurement of evapotranspiration

Summary The Bowen ratio-energy balance (BREB) and the stability-corrected aerodynamic method were used to estimate turbulent fluxes of sensible and latent heat at an irrigated alfalfa site in a semi-arid valley in northern Utah, U.S.A., during August and September of 1991. Despite inclusion of a generalized stability factor, the aerodynamic method underestimated the daytime (sunrise-sunset) sensible and latent heat fluxes by approximately 30% in comparison with the BREB method. The sum of the aerodynamic estimates of sensible and latent heat seldom balanced the energy avaiable from net radiation and change in storage. Wind speed was low during the experiment (averaging 1.6 m s^–1), and so a second analysis was run for data from daytime, non-rainy, turbulent conditions (wind > 1.5 m s^–1). This showed that sensible and latent heat were still underestimated by approximately 30% in comparison with the BREB approach. This suggests that underestimation of sensible and latent heat fluxes by the aerodynamic method was not related to the wind speed conditions during the experiment. These results show that the stability-corrected aerodynamic model did not agree with the Bowen ratio method in this experiment. It appears unlikely that the discrepancies resulted from measurement errors. Perhaps the theoretical foundation of the similarity parameters (stability functions) in the aerodynamic model are not sufficiently generalized. The discrepancies found here confirm the necessity of calibration checks on the validity of aerodynamic estimates of the turbulent fluxes.With 7 Figures     

A comparison of bowen ratio and eddy correlation sensible and latent heat flux measurements above deciduous forest

Sensible and latent heat flux densities (H and E) were measured above a mature, 18 m deciduous forest during July and August, 1988, using the Bowen ratio-energy balance (BREB) and eddy correlation (EC) methods. EC estimates ofH and E underestimated day-time surface available energy by 11%. EC also partitioned available energy differently than BREB. for/L<0.0, EC favouredH and BREB favoured E. Practical and theoretical limitations of the BREB and EC methods above forests are discussed. The most plausible causes for the failure of EC to close the surface energy balance are a low frequency loss of flux and the failure of a single point measurement to account for the spatial dispersive flux. The most plausible causes of the EC-BREB energy partitioning anomaly are the invalidity of the BREB similarity assumption and the violation of flux-gradient diffusion assumptions in the near-field diffusion region.     

Modelling evaporation from wetland tundra

Evapotranspiration is a major component of both the energy and water balances of wetland tundra environments during the thaw season. Reliable estimates of evapotranspiration are required in the analysis of climatological and hydrological processes occurring within a wetland and in interfacing the surface climate with atmospheric processes. Where direct measurements are unavailable, models designed to accurately predict evapotranspiration for a particular wetland are used.This paper evaluates the performance, sensitivity and limitations of three physically-based, one-dimensional models in the simulation of evaporation from a wetland sedge tundra in the Hudson Bay Lowland near Churchill, Manitoba. The surface of the study site consists of near-saturated peat soil with a sparse sedge canopy and a constantly varying coverage of standing water. Measured evaporation used the Bowen ratio energy balance approach, to which the model results were compared. The comparisons were conducted with hourly and daily simulations.The three models are the Penman-Monteith model, the Shuttleworth-Wallace sparse canopy model and a modified Penman-Monteith model which is weighted for surface area of the evaporation sources.Results from the study suggest that the weighted Penman-Monteith model has the highest potential for use as a predictive tool. In all three cases, the importance of accurately measuring the surface area of each evaporation source is recognized. The difficulty in determining a representative surface resistance for each source and the associated problems in modelling without it are discussed.

List of Symbols

Models BREB Bowen ratio energy balance - P-M Penman-Monteith combination - S-W Shuttleworth-Wallace combination - W-P-M Weighted Penman-Monteith combination Other AE Available energy-all surfaces - AE _c Available energy-canopy (S-W, W-P-M) - AE _s Available energy-bare soil (S-W, W-P-M) - AE _w Available energy-open water (W-P-M) - C _p Specific heat of air - D Vapor pressure deficit - DAI Dead area index - FAI Foliage area index - LAI Leaf area index - Q ^* Net radiation - Q _e Latent heat flux-total - Q _ec Latent heat flux-canopy (S-W, W-P-M) - Q _es Latent heat flux-bare soil (S-W, W-P-M) - Q _ew Latent heat flux-open water (W-P-M) - Q _g ground heat flux - Q _h Sensible heat flux - S Proportion of area in bare soil - W Proportion of surface in open water - r _a Aerodynamic resistance (P-M, W-P-M) - r _c Canopy resistance - r _s Generalized optimized surface resistance - r _st Stomatal resistance - r _c ^a Bulk boundary layer resistance (S-W) - r _s ^a Aerodynamic resistance below mean canopy level (S-W) - r _s ^s Soil surface resistance (S-W, W-P-M) Greek Bowen ratio - Psychrometer constant - Air density - Slope of saturation vapour pressure vs temperature curve     

Estimating sensible heat flux from radiometric temperature over crop canopy

The model devised by Lhommeet al. (1988) allows one to calculate the sensible heat flux over a homogeneous crop canopy from radiometric surface temperature by adding a so-called canopy aerodynamic resistance to the classical aerodynamic resistance calculated above the canopy. This model is reformulated in order to simplify the mathematical procedure needed to calculate this additional resistance. Analytical expressions of micrometeorological profiles within the canopy are introduced. Assuming a constant leaf area density, an analytical expression of canopy aerodynamic resistance is inferred, which is a function of wind velocity, inclination angle of the radiometer and crop characteristics such as crop height, leaf area index, inclination index of the foliage and leaf width. Sensitivity of this resistance to the different parameters is investigated. The most significant are wind velocity and LAI. Finally, the predictions of the model are tested against two sets of measurements obtained for two different crops, potato and maize.     

Daytime variations of ozone eddy fluxes to maize

The vertical fluxes of ozone, momentum and heat in the atmospheric surface layer have been measured by eddy correlation above both mature and senescent maize canopies. Aerodynamic formulae are applied to find that the bulk canopy surface resistancer _c to ozone uptake and destruction varies between 4.0 and 0.5 s cm⁻¹ during the daytime. Apparently, surface properties tend to control the removal of ozone at the surface of the earth. For a lush canopy, the stomatal diffusion resistance is the most important property, while changes in surface temperature have little effect. Destruction at the soil and exterior plant surfaces appears to account for 20–50% of the total loss if leaf mesophyll resistances are assumed to be very small. Free water at leaf surfaces may at times inhibit ozone removal by both senescent and healthy plants.     

Transpiration and evaporation from heather Moorland

The energy balance of an upland heath dominated by heather (Calluna vulgaris) was measured in dry and wet weather. Median values of both transpiration and evaporation rates were ca. 2 mm hr^-1. The median Bowen ratio for the dry canopy was 2.0 and for the wet canopy 0.6. On dry days the median value of the saturation deficit was only 3.8 mb and that of the climatological resistance was 30 s m^-1. The bulk stomatal resistance increased from ca. 50 s m ^-2 in the morning to over 290 s m^-1 in the afternoon with an overall median value of 110 s m^-1. Transpiration from the dry canopy was controlled by a combination of small saturation deficits and large stomatal resistances. The median value of the boundary-layer resistance of the canopy was 22 s m^-1 and was low partly because of a large low-level drag coefficient. Saturation deficits on wet days were close to zero and evaporation of intercepted water proceeded at close to the equilibrium rate, being largely limited by the low fluxes of available energy. The water loss from heather was compared with simulated losses from coniferous forest, herbaceous crops and grassland in the same conditions to evaluate the effects of vegetation on water loss from catchments.Laboratorio de Ecologia, Dept. de Biologia Vegetal, Universidade de Brasilia, 70 910 — Brasila — DF, Brazil.     

The conductance of a maize crop and the underlying soil to ozone under various environmental conditions

Flux measurements of ozone and water vapour employing the eddy correlation technique were used to determine the surface conductance and canopy conductance to ozone. In the surface conductance to ozone, all surfaces at which ozone is destroyed and the transport process to these surfaces are included. The canopy conductance to ozone represents the ozone uptake of transpiring plant parts. The surface conductance to ozone of the maize crop and the underlying soil was generally larger than the canopy conductance to ozone. This means that beside the uptake by stomata, there was another important ozone sink. Under wet soil surface conditions, the surface conductance and the canopy conductance to ozone coincided. This indicates that the resistance of wet soil and the remaining plant parts (cuticle) to ozone was much larger than the stomatal or soil resistance. On the other hand, under dry soil conditions the conductances differ, largely caused by a variation in the transport process to the soil. The transport of ozone to soil increased with increasing friction velocity (u _*) and decreased with increasing atmospheric stability, leaf area index (LAI) or crop height (h). These effects for midday (unstable) conditions were parameterized with an in-crop aerodynamic resistance,r _inc in a very straightforward way;r _inc=13.9 LAIh/u _*+67 (cc.=0.77). If the ozone flux in air pollution models is described with a simple resistance model (Big Leaf model), the extra destruction at the soil should be modelled using an in-crop aerodynamic resistance. For these measurements the ozone flux to the soil was 0–65% of the total ozone flux measured above the crop. Under wet soil conditions, this was less than 20%; under dry soil conditions, this was 30–65%.     

Dew formation and the drying process within a maize canopy

The amount of dew-fall and dew-rise to a maize canopy during one night, and the drying time during the early morning, were estimated by using the Bowen ratio energy budget technique (BREB) and the soil diffusivity technique (SD).In addition, the distribution of the free liquid water within the plant community was measured at sunrise by using Leick plates and blotting papers. The course of the drying process within the plant canopy was also monitored by the blotting paper technique until the received night-time free liquid water within the canopy was evaporated. The calculated total amount of dew, estimated with the BREB and SD techniques, agreed with the measured amount to within 6%, though the measured drying process lagged about 1 hour behind the calculated one. Possible causes of this difference are the dripping effect and drainage down the stems, which were not included in the calculations. Initially, the drying process was faster in the upper layers of the canopy. Later, when the moisture distribution had become more or less even through the whole canopy, the drying rate became approximately uniform.     

A Measurement Based Sensitivity Analysis of the Penman-Monteith Actual Evapotranspiration Model for Crops of Different Height and in Contrasting Water Status

Summary This paper presents a study of the sensibility of the Penman-Monteith evapotranspiration model to climatic (available energy and vapour pressure deficit) and parametric (aerodynamic and canopy resistances, r _a and r _c respectively) factors in a semi-arid climate, for crops in contrasting water status (well irrigated and under water stress) and of different heights. Three experiments were carried out in southern Italy on reference grass (≈ 0.1 m), grain sorghum (≈ 1 m) and sweet sorghum (≈ 3 m). For this analysis the sensitivity coefficients, taken as hourly means, were evaluated during the growth season when the crops completely covered the soil. The relative errors on evapotranspiration were also evaluated for r _a and r _c . The results showed that, for reference grass, available energy and aerodynamic resistance play a major role. For crops under water stress the most important term to evaluate is canopy resistance. For a tall crop, as sweet sorghum, the role of the vapour pressure deficit is fundamental, both when the crop is in good water status and under water stress. Received July 14, 1997 Revised February 5, 1998     

Daily reference evapotranspiration estimates by the Penman-Monteith equation in Southern Italy. Constant vs. variable canopy resistance

Summary ?The performance of the Penman-Monteith (PM) equation to estimate daily reference evapotranspiration (ET_O) was investigated by attributing three distinct features to the canopy resistance (r _c): (i) r _c constant at 70 s m⁻¹ (Allen et al., 1998; FAO Irrigation and Drainage Paper n. 56), (ii) r _c variable as linear function of a critical resistance r _c, depending on weather variables and empirical parameters relating r _c to r ^* (Katerji and Perrier, 1983; Agronomie, 3[6]: 513–521) and (iii) r _c variable as a mechanistic function of weather variables only (Todorovic, 1999; J. Irrig. Drainage Eng., ASCE, 125[5]: 235–245). Daily weather and grass lysimeter data, measured for a period of seven years at Policoro (Southern Italy), were used. The results confirmed the relative robustness of the PM method with constant r _c while better estimates were obtained only when variable r _c was used. The mechanistic approach of Todorovic (1999) provided the best estimates, while the approach of Katerji and Perrier (1983), with empirically derived parameters, has shown to be not conservative enough to be extended to different locations without calibration. Received January 2, 2002; revised October 31, 2002; accepted December 7, 2002     

A mixed-layer model for regional evaporation

A simple mixed-layer model is developed to describe evaporation into a convective planetary boundary layer (PBL). The model comprises volume budget equations for temperature and humidity, equations to describe transport through the surface layer which is treated as part of the lower boundary, and equations to describe entrainment at the top of the PBL. The ground surface is modelled as a canopy resistance. The model was integrated with canopy resistance, surface-layer resistance and available energy, (R _n – G), input as given functions of time, and the simulated PBL was allowed to grow into an atmosphere with known temperature and humidity profiles.Two variants of the mixed-layer model were tested using data from the KNMI tower site at Cabauw in the Netherlands. These variants differed only in the formulation of entrainment: one used a formulation developed by Driedonks (1982) while the other was a simpler formulation. Simulated evaporation agreed very well with observations irrespective of which entrainment formulation was used, despite discrepancies between simulated and observed PBL height growth which were sometimes quite large for the simpler formulation. Sensitivity analysis of the model confirms that good PBL height-growth predictions are not always a prerequisite for good evaporation predictions.