A smog chamber for studies of the reactions of terpenes and alkanes with ozone and OH

The design and performance of a smog chamber for the study of photochemical reactions under simulated environmental conditions is described. The chamber is thermostated for aerosol experiments, and it comprises a gas chromatographic sample enrichment system suitable for monitoring hydrocarbons at the ppbv level. By irradiating NO _x /alkane-mixtures rate constants for the reaction of OH radicals with n-alkanes are determined from n-pentane to n-hexadecane to be (k±2)/10^–12 cm³ s^–1=4.29±0.16, 6.2±0.6, 7.52 (reference value), 8.8±0.3, 10.2±0.3, 11.7±0.4, 13.7±0.3, 15.1±0.5, 17.5±0.6, 19.3±0.7, 22.3±1.0, and 25.0±1.3, respectively at 312 K. Rate constants, (k±2)/10^–17 cm³ s^–1, for the reaction of ozone with trans-2-butene (21.2±1.0), cis-3-methylpentene-(2) (47.2±1.7), cyclopentene (62.4±3.5), cyclohexene (7.8±0.5), cycloheptene (28.3±1.5), -pinene (8.6±1.3), and -pinene (1.4±0.2) are determined in the dark at 297 K using cis-2-butene (13.0) as reference standard.     

Henry's law constant and hydrolysis of peroxyacetyl nitrate (PAN)

The effect of temperature on the solubility of PAN and on its hydrolysis rate in near-neutral and slightly acidic water were studied in a bubble column apparatus. The results obtained are a Henry's law coefficient H=10^–9.04±0.6 exp[(6513±376)/T] M atm^–1, and a first-order hydrolysis rate constant k=10^6.60±1.0 exp[(–6612±662)/T] s^-1, which was independent of pH in the range 3.2pH6.7. The products formed are nitrite and nitrate in approximately equal proportions under near-neutral conditions. At a pH<4, nitrite is oxidized in a secondary reaction, and nitrate becomes the only product at low pH. Previously measured deposition velocities of PAN on stagnant water surfaces are shown to be hydrolysis rate limited.     

The behaviour of aerosol optical depth during clear summer days in Athens,Greece

Summary The integral aerosol optical depths (k ) at the hour of 08:20 Local Standard Time (LST), are compared with those calculated previously at 11:20 and 14:20 LST, for clear days during summer in Athens over the period 1962–1988. The mean values at 08:20 LST were consistently lower than the values at 11:20 and 14:20 LST. The influence of the vertical wind profile on the values ofk was also investigated. A comparison was made of the wind profiles at 02:00 and 14:00 LST, for days in which the 11:20 and 14:20 LST values ofk were 0.200 andk 0.350, respectively. The corresponding bulk wind shear s was also found for the period 1980–1988. The most significant results occurred with the first category of days. The resultant wind velocities from the surface to the 900 hPa level, in each hour were higher by 2–4 m·s^–1 with respect to the corresponding values for the second category. At 02:00 LST the bulk wind shear showed a considerable difference (1.8) between the two categories of days in the surface to 700 hPa layer at 02:00 LST. Finally, the associated weather conditions that appear to initiate a period of low values ofk (k 0.200) at 11:20 and 14:20 LST were examined for the period 1980–1988. Fifteen such cases were identified and it was found that they all occurred after the passage of weak cold fronts.With 6 Figures     

A generalization of the Kolmogorov-von Kármán relationship and some further implications on the values of the constants

A general relation between the Kolmogorov (₁) and von Kármán (k) constants appropriate to flows in the surface boundary layer is obtained by including stability and divergence dependent terms in the development. The result is used to show that the author's earlier inference of the valuek = 0.36 drawn from the observed value ,k ^4/3=0.141 cannot have been underestimated due to the neglect of stability or divergence effects. It is concluded that if k is to be 0.4, then ₁ must be 0.5.Work performed under the auspices of the U.S. Energy Research and Development Administration.     

Radiometrically determined skin temperature and scalar roughness to estimate surface heat flux part II: Performance of parameterized scalar roughness for the determination of sensible heat

Several formulations and proposals to determine the value of the radiometric scalar roughness for sensible heatz _0h,r are tested with respect to their performance in the estimation of the sensible heat flux by means of the profile equations derived from Monin-Obukhov similarity theory. The equations are applied to the data set of spatially averaged surface skin temperature and profiles of wind speed and temperature observed in a pasture field during a growing season. The use of a physical model developed for a dense canopy to estimate scalar roughness for sensible heatz _0h,r produced sensible heat fluxH with a correlation coefficientr=0.884, the ratio of means being H _s /H=1.19 in a comparison with reference values ofH _s . In comparison, a proposal for a fixed value ofz _0h yieldedr=0.887, H _s /H=0.879. In both cases, the validity ofz _0h =z _0h,r was assumed. All expressions derived to estimatez _0h,r from a multiple linear regression with such predictors as leaf area index, solar radiation and the ratio of solar radiation to extraterrestrial radiation, were found to produce a better result, withr better than 0.90 and H _s /H around 1.0. However, when the constantsc andf of a linear regression equationHs=cH+f are used to evaluate the equations, a marked difference in performance of each formulation appeared. In general, equations with smaller numbers of predictors tend to produce a biased result, i.e., an overestimation ofH at largeH _s . These values ofH are used in conjunction with the energy balance equation to derive values of the latent heat fluxLE, which are shown to be in good agreement with the reference valuesLE _s , withr greater than 0.97.     

On the cumulative frequency distributions ofk-day period amounts of precipitation for any station in The Netherlands,while k≧30

Summary A simple method of representing cumulative frequency distributions ofk-day period amounts of precipitation (30k1800), deduced from observations at Hoofddorp during 87 years and from observations at Winterswijk during 73 years, by means of cumulative Poisson distributions, had been developed. Poisson's constant appeared to depend linearly onk only, while the Poisson unitq appeared to be only proportional to the annual variation.In this paper it is shown that the representation of observed cumulative frequency distributions with the aid of Poisson distributions is true for any station in The Netherlands in the interval 30k1800. A map with isopleths of -values fork=360 presents a simple picture, so that for any station an -value fork=360 can be estimated. For the process of representation the meank-day period amount of precipitation is also necessary. This value may be obtained from a map for each of the 12 months, each map covered with isohyetal lines, which give a distribution of the monthly precipitation.

---

Zusammenfassung Es wurde eine Methode entwickelt, um in einfacher Weise Summenfunktionen vonk-tägigen Niederschlagsmengen (30k1800), welche aus Beobachtungen in Hoofddorp während 87 Jahren und in Winterswijk während 73 Jahren angestellt worden sind, durch kumulative Poissonsche Häufigkeitsverteilungen darzustellen. Dabei ergab sich, daß die Poissonsche Konstante nur linear vonk abhängig ist, während die Poissonsche Einheitq sich als proportional zum Jahresgang erwies.In der vorliegenden Untersuchung wird nun gezeigt, daß die Darstellung der beobachteten kumulativen Häufigkeitsverteilungen mit Hilfe von Poisson-Verteilungen für irgendeine Station in den Niederlanden im Bereich von 30k1800 gültig ist. Eine Karte mit Isoplethen der -Werte fürk=360 ergibt ein einfaches Bild, aus dem für jede beliebige Station ein -Wert fürk=360 abgelesen werden kann. Für eine solche Darstellung muß auch die Niederschlagsmenge derk-tägigen Perioden bekannt sein; diese Werte erhält man aus monatlichen Isohyeten-Karten, die die monatliche Niederschlagsverteilung wiedergeben.

---

Résumé On avait développé une méthode pour représenter d'une façon simple les distributions des sommes cumulées dek jours (30k1800) des quantités de précipitations déduites des observations faites à Hoofddorp pendant 87 ans et à Winterswijk pendant 73 ans. On avait démontré que la constante de Poisson ne dépend dek que linéairement et que l'unité de Poissonq est proportionnelle à la variation annuelle.De la présente étude ressort que, dans l'intervalle de 30k1800, la représentation des distributions des fréquences cumulées observées au moyen des formules de Poisson est valable pour une station quelconque aux Pays-Bas. Une carte d'isoplèthes de la valeur pourk=360 donne une image suffisamment simple pour que la valeur d'une station quelconque aux Pays-Bas puisse être déterminée sans difficulté. Pour une telle représentation il faut connaître également les moyennes des quantités des périodes dek jours de la précipitation. Ces valeurs peuvent se déduire de cartes mensuelles des isohyètes moyennes.

---

---

With 1 Figure

---

Dedicated to Dr.Anders K. Ångström on the occasion of his 70th birthday.     

A new theory of the energy spectrum

In a recent paper, the author introduced a new viscous boundary layer, called the mesolayer, in turbulent shear flow. Its importance stems from its location between the inner and outer regions which are controlled by the law of the wall and Reynolds number similarity, respectively. This intrusion prevents the classical overlap assumption which appears to be fundamental in the derivation of the classical logarithmic behavior. The mesolayer has a thickness proportional to Taylor's microscale . This, and the analogy between the energy equation for the spectrum function of isotropic turbulence and the momentum equation for shear flow, suggest the existence of a similar region in wavenumber space with wavenumber k ~ ^-1. This mesoregion separates the inner region k ~ k _s(where k _s^-1 and is the Kolmogorov length) and the outer region k k _e(where k _e ^-1 and l is the energy-containing eddy size) and again invalidates the overlap assumption which appears to be fundamental in the derivation of the classical k ^-5/3-behavior of the inertial subrange.Incorporation of the mesoregion into the argument leads to a new theory with k ^-5/3-behavior in two regions (^-1 k k _s) and (k _e k ^-1) although with two different coefficients of proportionality (Kolmogorov constants). This leads to a wandering of the spectrum curve about the classical k ^-5/3 line similar to a wandering in turbulent shear flow about the logarithmic curve. This is clearly indicated by the data for the variation of the Kolmogorov constant.Other data support the new theory. In particular, the location of the point k _mwhere the curve of the nonlinear energy-transfer function goes through zero shows agreement with the theory, i.e., k _m^-1.     

Aerodynamic roughness of vegetated surfaces

Available experimental results indicate that as the density of roughness elements over a horizontally homogeneous surface is varied, the roughness length, z ₀, varies in a manner that exhibits a maximum at intermediate density values. In an attempt to explain this behaviour, the available analytical solutions for the wind profile inside dense homogeneous canopies were reviewed. The review indicated that the variation of z ₀ with density depends on the interrelationship between the leaf density, a, and the mixing length, l. In view of this finding, a numerical model was devised based on a simple rule for constructing mixing-length profiles in the canopy. The rule states that the actual value of l is the maximum possible under the two constraints: l l _i and ¦dl/dz¦ k, where k is the von Karman constant and the intrinsic mixing length, l _i, is a function of the local internal structure of the canopy. The model which ensures a smooth transition from dense to thin canopy, was used to reproduce the observed maximum of z ₀. The model is also capable of handling vertically non-homogeneous canopies.     

The Influence of NaBr/NaCl Ratio on the Br--Catalysed Production of Halogenated Radicals

The activation of Br^- and Cl^- to atomic Br and Cl in sea-spray aerosol was investigated in smog-chamber experiments. In the presence of O₃, hydrocarbons and NaCl aerosol alone no activation was observed. By adding Br^- to the aerosol, the chain reaction: Br + O₃ BrO, BrO + HO₂ HOBr, HOBr HOBr(aq), HOBr(aq) + H⁺ + Br^- Br₂ (6), HOBr(aq) + H⁺ + Cl^- BrCl (7) was verified. The step from reaction (6) to (7) is accompanied by a decrease of the Br^-/Cl^- ratio from 1/600 to less than 1/2000. In the absence of sulphate, the chain is initiated by the reaction of OH(aq) with Br^-. The pH value decreases to less than 2 during the first minutes of the experiment and later on to almost 1 (in the absence of NO_x or SO₂). This is caused by the formation of oxalic acid from alkanes and toluene. In stopped flow experiments, the reduction of Br₂ by oxalic acid was observed to occur through a two-step mechanism: HC₂O₄ ^- + Br₂ Br^- + BrC₂O₄H (k₂₂, k_-22), BrC₂O₄H Br^- + H⁺ + 2 CO₂ (23) with the following rate constants and ratios of rate constants, k ± 2: k₂₂k_-23 / k_-22 = (2.9 ± 0.3) · 10^-4 s^-1, k_-22 / k_-23 = 7000 ± ₃₀₀₀ ¹³⁰⁰⁰ M^-1, k₂₂ = 2 ±_-1 ⁴ M^-1 s^-1, and k_-23 > 0.1 s^-1, k_-22 > 600 M^-1 s^-1. Oxalic acid may be responsible for the inhibition of the chain reaction observed at the end of the experiments.     

Sea surface winds derived from cloud motion vectors over the tropical Atlantic

The relationship between satellite-derived low-level cloud motion, surface wind and geostrophic wind vectors is examined using GATE data. In the trades, surface wind speeds can be derived from cloud motion vectors by the linear relation: V = 0.62 V _s + 1.9 m s^–1 with a mean scatter of ±1.3 m s^–1. The correlation coefficient between surface and satellite wind speed is 0.25. Considering baroclinicity, i.e., the influence of the thermal wind, the correlation coefficient does not increase, because of the uncertainty of the thermal wind vectors. The ratios of surface to geostrophic wind speed and surface to satellite wind speed are 0.7 and 0.8, respectively, with a statistical uncertainty of ±0.3. Calculations of the ratio of surface to geostrophic wind speed on the basis of the resistance law yield V/V _g = 0.8 ± 0.2, in agreement with experimental results. The mean angle difference between the surface and the satellite wind vectors amounts to - 18 °, taking into account baroclinicity. This value is in good agreement with the mean ageostrophic angle - 25 °.     

The extent of the unstable Monin-Obukhov layer for temperature and humidity above complex hilly grassland

Potential temperature, specific humidity and wind profiles measured by radiosondes under unstable but windy conditions during FIFE in northeastern Kansas were analyzed within the framework of Monin-Obukhov similarity. Around 86% of these profiles were found to have a height range over which the similarity, formulated in terms of the Businger-Dyer functions, is valid and for which the resulting surface fluxes are in good agreement with independent measurements at ground stations. When scaled with the surface roughness z ₀ = 1.05 m and the displacement height d ₀ = 26.9 m, for the potential temperature this height range was 45 (±31) (z – d ₀ )/z ₀ 104 (±54) and the comparison of the profile-derived surface fluxes with the independent measurements gave a correlation coefficient of r = 0.96. For the specific humidity these values are 42 (±29) (z – d ₀ )/z ₀ 96 (±38) and r = 0.94. In terms of the height of the bottom of the inversion H _i , in the morning hours the upper limit of (z – d ₀ ) in the Monin-Obukhov layer is approximately 0.3H _i , whereas for a fully developed ABL it is closer to 0.1H _i . Probably, as a result of the short sampling times and perhaps also of the small gradients under the windy conditions, the exact height range of validity was difficult to establish from a mere inspection of these profiles.     

On the attenuation of direct solar irradiance in Athens due to aerosols

Summary In this study the trend of the time sequence of the integral aerosol optical depth (k _a), as proposed by Unsworth and Monteith, was determined for clear days in summer for the period 1962–1988 in Athens. The trend was found by fitting a third degree polynomial curve and it was concluded that (k _a) showed a considerable increase (i.e. from the value of 0.18 to 0.31) in the period 1962–1976 and remained approximately constant until 1979, after which it started decreasing again slowly until 1988. The increase of (k _a) in the period 1962–1976 is likely attributable to the rapid development of the city in this period, while the decrease of (k _a) after 1979 probably reflects the efficiency of some restrictions which were imposed on the pollutant emissions during this period. In addition, an analysis of the percentage frequency distribution found that while 95% of the values of (k _a) ranged from 0.100 to 0.400 in the beginning of the period (1964–1967), in recent years (1984–1987) the same percentage of the values of (k _a) ranged from 0.100 to 0.500.With 3 Figures     

Estimation of the sensible heat flux of a semi-arid area using surface radiative temperature measurements

In the framework of an international field program for the study of semi-arid areas, observations were done in the region called La Crau in southern France. In this paper, the use of the surface radiative temperature for the determination of the sensible heat flux is addressed. We found that, once proper values of the roughness length of momentum (z ₀) and heat (z ₀h) are set, the sensible heat flux can be reliably predicted with a one-layer resistance model using standard observations of wind speed and air temperature, together with the surface temperature. The latter quantity has to be known with a precision better than ±2°C. From our observations, the value of the parameterB ^–1k ^–1 In (z ₀ z ₀h) was found to be 9.2, which falls between values quoted by Brutsaert (1982) for grass and bluff bodies.     

A comparison of airborne eddy correlation and bulk aerodynamic methods for ocean-air turbulent fluxes during cold-air outbreaks

Four bulk schemes (LKB, FG, D and DB), with the flux-profile relationships of Liuet al. (1979), Francey and Garratt (1981), Dyer (1974), and Dyer and Bradley (1982), are derived from the viscous interfacial-sublayer model of Liuet al. These schemes, with stability-dependent transfer coefficients, are then tested against the eddy-correlation fluxes measured at the 50 m flight level above the western Atlantic Ocean during cold-air outbreaks. The bulk fluxes of momentum (), sensible heat (H), and latent heat (E) are found to increase with various von Kármán constants (k _M for k _H forH, andk _E forE). Except that the LKB scheme overestimates by 28% (46Wm^–2), on the average, the fluxes estimated by the four bulk schemes appear to be in fairly good agreement with those of the eddy correlation method (magnitudes of biases within 10% for , 17% forH, and 13% forE). The results suggest that the overall fluxes and surface-layer scaling parameters are best estimated by FG and thatk _H <k _E . On the average, the FG scheme underestimates by 10% (0.032N m^–2) andE by 4% (12Wm^–2), and overestimatesH by 0.3% (0.5W m^–2). The equivalent neutral transfer coefficients at 10 m height of the FG scheme compare well with some schemes of those tested by Blanc (1985).The relative importance of various von Kármán constants, dimensionless gradients and roughness lengths to the oceanic transfer coefficients is assessed. The dependence of transfer coefficients on wind speeds and roughness lengths is discussed. The transfer coefficients for andE agree excellently between LKB and FG. However, the ratio of the coefficient forH of LKB to that of FG, increasing with decreasing stability, is very sensitive to stability at low winds, but approaches the neutral value of 1.25 at high winds.     

Analytical investigations of horizontal meso-scale momentum equations with Newtonian nudging

Summary This study presents an analytical investigation of the local behaviour of the solution to a mesoscale model with Newtonian nudging when observed winds are time varying. The analysis examines each Fourier component of the time series of observed winds. Unlike the case with a constant observed wind, the nudged wind vector does not asymptotically approach the observed wind. In response to sinusoidal oscillation of the observed wind, the nudged wind vector is always on a half circle connecting the vector ends of the observed and un-nudged modelled winds. When nudging parameter 0, the nudged wind vector approaches the un-nudged wind; when , the nudged wind vector approaches the observed wind. For commonly used values of nudging parameter , the modelled wind field always carries errors.A target nudging scheme is devised in this study in order to ensure the model result is identical to observed winds with sinusoidal oscillation. Investigation shows that such a target wind exists for a finite value of , and the magnitude of the target-nudging term is about the same as that of a normal nudging term if f, wheref is the Coriolis parameter and is the frequency of the wind oscillation.With 7 Figures     

Equatorial disturbances,monsoons, and 1982–1983 ENSO

Summary Interannual modes are described in terms of three-month running mean anomaly winds (u,v), outgoing longwave radiation (OLR), and sea surface temperature (T _* ). Normal atmospheric monsoon circulations are defined by long-term average winds (u _n,v _n) computed every month from January to December. Daily winds are grouped into three frequency bands, i.e., 30–60 day filtered winds (u _L,v _L); 7–20 day filtered winds (u _M,v _M); and 2–6 day filtered winds (u _S,v _S). Three-month running mean anomaly kinetic energy (signified asK _L , K _M , andK _S , respectively) is then introduced as a measure of interannual variation of equatorial disturbance activity. Interestingly, all of theseK _L , K _M , andK _S perturbations propagate slowly eastward with same phase speed (0.3 ms^–1) as ENSO modes. Associated with this eastward propagation is a positive (negative) correlation between interannual disturbance activity (K _L , K _M , K _S ) and interannualu (OLR) modes. Namely, (K _L , K _M , K _S ) becomes more pronounced than usual nearly simultaneously with the arrival of westerlyu and negativeOLR (above normal convection) perturbutions. In these disturbed areas with (K _L , K _M , K _S >0), upper ocean mixing tends to increase, resulting in decreased sea surface temperature, i.e.T _* 0. Thus, groups (not individual) of equatorial disturbances appear to play an important role in determiningT _* variations on interannual time scales. HighestT _* occurs about 3 months prior to the lowestOLR (convection) due primarily to radiational effects. This favors the eastward propagation of ENSO modes. The interannualT _* variations are also controlled by the prevailing monsoonal zonal windsu _n, as well as the zonal advection of sea surface temperature on interannual time scales. Over the central Pacific, all of the above mentioned physical processes contribute to the intensification of eastward propagating ENSO modes. Over the Indian Ocean, on the other hand, some of the physical processes become insignificant, or even compensated for by other processes. This results in less pronounced ENSO modes over the Indian Ocean.With 10 FiguresContribution No. 89-6, Department of Meteorology, University of Hawaii, Honolulu, Hawaii.     

On the similarity functions A and B as determined from the ‘VOVES’ experiment

Data collected during the french VOVES-1977 experiment are used for the determination of the functionsA() andB() of the stability parameter . It is found that, although pertaining to different geographic conditions, they lead to the same kind of dependency ofA andB upon as found previously considered data. The influence of baroclinicity onA andB is also studied and it is found in that, statistically, consideration of this parameter does not lead either to a decrease of the large scatter of experimental points or to a significant improvement of the similarity theory. It is also shown that a satisfactory estimation of the surface geostrophic wind can be made from the network of meteorological stations, to compute such functions.     

Photochemical aspects of tropospheric iodine behaviour

The I-atom sensitised decomposition of ozone in air at 1 atm pressure and ambient temperature has been investigated. Iodine atoms were produced by photolysis of I₂ using visible light or of CH₃I using ultraviolet light. In both cases, the quantum yield for O₃ decomposition was 1.25 (±0.11) per I atom. An important role is proposed for the self-reaction of IO radicals leading to higher oxides of iodine, IO+IO(+M)I₂O₂(+M)higher oxides, which predominated over the bimolecular reaction leading to regeneration of I atoms, IO+IO2I+O₂, with k _2a/k _2b4. Simple computer modelling calculations indicate that reaction (2a) may be important in determining the fate of photolabile iodine species in the atmosphere. The consequences for the behaviour of radioiodine releases are also discussed.     

An alternative analysis of flux-gradient relationships at the 1976 ITCE

An extensive micrometeorological data set from the 1976 International Turbulence Comparison Experiment (ITCE) is analysed to determine flux-gradient relationships in an unstable atmosphere for momentum, sensible heat and water vapour transfers. The data are first analysed for internal consistency, resulting in the rejection of some data. Following a least-square fit to the remaining data in the form /k = (1 – z/L)^-/k, rounded-off values of k, , and are selected for each form of transfer consistent with the statistical accuracy of the measurements. The equations finally adopted are _M = (1 – 28z/L)^-1/4 and _{H, W} = (1 – 14z/L)^-1/2 with k _M = k_H = k_W = 0.40.These expressions fit the averaged observations to within a few per cent in the stability range of the experiment (-4 < z/L < -0.004).     

The influence of surface cover and climate on energy partitioning and evaporation in a subarctic wetland

Energy partitioning and evaporation were measured over three wetland surfaces in a subarctic coastal marsh during pre-growing and growing periods. These surfaces included an alder/willow woodland, a sedge marsh and a raised backshore sedge meadow. A combination model analysis was used to assess the relative importance of surface resistance and meteorological conditions on the magnitude of the Bowen ratio, , during the growing period.Overall, the three surfaces experienced important site-to-site and seasonal differences in and evaporation, Q _E. During the non-foliated period, Q _E was largest and was smallest for the open water marsh, while the dry backshore site experienced the smallest Q _E and largest . The non-foliated woodland assumed intermediate values of and Q _E. After the vegetation covers were established, the woodland assumed the smallest and largest Q _E flux. It was also found that at the marsh site increased with the presence of a vegetation cover.Wind direction was always an important factor in determining Q _E and at all sites. was substantially larger and Q _E was smaller for onshore winds (i.e., originating from James Bay) than for offshore winds. The combination model analysis showed that canopy resistance at all sites was largest during warm offshore winds, which were associated with large saturation deficits. However, the effect of increased canopy resistance on during offshore winds was offset by a large climatological resistance, resulting in small values and large Q _E. When winds originated from James Bay, canopy resistance was smaller than for offshore winds, but the climatological resistance also was much smaller, resulting in larger and small Q _E. The results have important implications for changes in land cover and climate on the regional water balance.