Attenuation Correction Procedures for Water Vapour Fluxes from Closed-Path Eddy-Covariance Systems

Evapotranspiration is a source of water vapour to the atmosphere, and as a crucial indicator of landscape behaviour its accurate measurement has widespread implications. Here we investigate errors that are prevalent and systematic in the closed-path eddy-covariance measurement of latent heat flux: the attenuation of fluxes through dampened cospectral power at high frequencies. This process is especially pronounced during periods of high relative humidity through the adsorption and desorption of water vapour along the tube walls. These effects are additionally amplified during lower air temperature conditions. Here, we quantify the underestimation of evapotranspiration by a closed-path system by comparing its flux estimate to simultaneous and adjacent measurements from an open-path sensor. We apply models relating flux loss to relative humidity itself, to the lag time of the cross-correlation peak between the water vapour and vertical wind velocity signals, and to models of cospectral attenuation relative to the cospectral power of simultaneous sensible heat-flux measurements. We find that including the role of temperature in modifying the attenuation–humidity relationship is essential for unbiased flux correction, and that physically based cospectral attenuation methods are effective characterizers of closed-path instrument signal loss relative to the unattenuated flux value.     

New Transfer Functions for Correcting Turbulent Water Vapour Fluxes

We address the problem of the high-frequency correction of water vapour fluxes measured by eddy covariance with a closed-path infrared gas analyser (IRGA). Different transfer functions are compared and evaluated at a forested (Vielsalm, Belgium) and an agricultural (Lonzée, Belgium) site. Classical functions, usually applied to correct CO₂ fluxes (Gaussian, Lorentzian), are found to be unsuited to water vapour cospectral corrections, being characterised by too sharp a decrease at high frequency. Two other functions characterised by a lower decreasing slope are found to better fit experimental transfer functions. They were calibrated and validated on experimental transfer functions and their dependency on air humidity is parameterised. On this basis, new correction coefficients are estimated. The coefficients are found to be larger than those based on the classical functions, even when the dependency of the latter on air humidity is taken into account. The difference amounts to 10% at the forested site and to 5% larger at the crop site. The study highlights the necessity of characterising the water transfer function shape and taking it into account in the correction factor at each site equipped with a closed path IRGA.     

A Wavelet-Based Correction Method for Eddy-Covariance High-Frequency Losses in Scalar Concentration Measurements

Eddy-covariance (EC) scalar-flux measurements suffer from unavoidable biases introduced by high-frequency losses in the sampled scalar concentration fluctuations. This bias alone leads to an underestimation of scalar fluxes by as much as 20% in some cases, especially when a closed-path gas analyzer is used to sample concentration far from the inlet location. A novel method that directly corrects for these high-frequency losses using only the sampled scalar-concentration time series is proposed and tested. The sampled concentration fluctuation time series is adjusted, point-by-point, in the wavelet half-plane for each EC averaging interval (??30?min). Similarity between scalars (and temperature) is not necessary and a pre-defined theoretical shape of the cospectrum is not required, making this method attractive at meteorologically non-ideal sites. When closed-path gas analyzers are used to measure H₂O concentration fluctuations, the method is shown to reproduce the dependence of the attenuation on air relative humidity. Nevertheless, the method is not able to account for excessively large spectral attenuation that occurs close to the spectral peak, as might be the case with long tubes and high relative humidity. Since the method corrects the original scalar concentration time series and not the cospectrum, other flow statistics??such as variances and integral time scales??are also adjusted. The proposed method can be used synergistically with conventional high-frequency cospectral correction methods given the differences in assumptions and approaches among these methods. When the conventional and the proposed methods agree, added confidence to the estimate of the high frequency correction is gained, and vice versa.     

Eddy correlation measurement of CO2 flux using a closed-path sensor: Theory and field tests against an open-path sensor

We have examined the potential of using a closed-path sensor to accurately measure eddy fluxes of CO₂. Five inlet tubeflow configurations were employed in the experimental setup. The fluxes of CO₂ were compared against those measured with an open-path sensor. Sampling air through an intake tube causes a loss of flux, due to the attenuation of CO₂ density fluctuations. Adjustments need to be made to correct for this loss and to account for density effects due to the simultaneous transfer of heat and water vapor. Theory quantifying these effects is discussed.The raw CO₂ flux measured with the closed-path sensor was smaller than that measured with the open-path sensor by about 15% (on average) for the turbulent tubeflow configurations with a short (3 m) intake tube, by 31% for turbulent tubeflow with a longer (6 m) intake tube and by 24% for laminar tubeflow. The difference was, in part, caused by tube attenuation of the CO₂ density fluctuations and inadequate sensor time response. The elimination of the flux adjustment for the simultaneous transfer of sensible heat (i.e., the attenuation of ambient temperature fluctuations in the intake tube) generally accounted for the rest of this difference.The raw flux measured with the closed-path sensor was corrected for frequency response and density effects. Except in the case of laminar tubeflow, the corrected closed-path flux agreed consistently with the corrected open-path flux within a few percent (<5%). These results suggest that closed-path sensors, with appropriate corrections, can be used to measure CO₂ flux accurately. Recommendations are included on selecting an optimum flow configuration to minimize the effect of sampling air through a tube.Published as Paper No. 9938, Journal Series, Nebraska Agricultural Research Division.     

Practical applicability of high frequency correction theories to CO<Subscript>2</Subscript> flux measured by a closed-path system

The theoretical correction of CO₂ fluxes for high frequency attenuation in closed-path systems was re-summarized and its applicability examined using both measurements obtained at an Asiaflux forest site and empirical transfer functions used in previous studies. For our measurement system, the theoretical transfer function was applicable to high frequency correction, even when condensation occurred in the sampling line. Further, in respect to some measurement systems described in previous studies, it was found that the theoretical function was potentially applicable along with the empirical functions used. Meanwhile, in some systems significant errors could not be resolved by re-estimation of the theories. In these systems, because of undefined buffering effects, the actual response lag time decided by the maximum covariance method or by measurement of the system response time using tracer gas was significantly different from the lag time calculated from the tube dimensions and the measured flow rate. If the average flow rate calculated by the actual lag time was used to determine the theoretical function, the theoretical function became closer to, and sometimes agreed with, the empirical function. Any remaining deviation from each function might be associated with pressure fluctuations, but this problem was unable to be examined here. The results suggested that an empirical formulation for each site is considered applicable rather than a theoretical approach, although the theories are being developed to practical application.     

Comparison of eddy-covariance measurements of CO2 fluxes by open- and closed-path CO2 analysers

Eddy fluxes of CO₂ estimated using a sonic anemometer and a closed-path analyser were, on average, 16% lower than those obtained with the same anemometer and an adjacent open-path CO₂ analyser. Covariances between vertical windspeed and CO₂ density from the closed-path analyser were calculated using data points for CO₂ that were delayed relative to anemometer data by the time required for a parcel of air to travel from the tube inlet to the CO₂ sensor. Air flow in the intake tube was laminar. Densities of CO₂ that had been corrected for spurious fluctuations arising from fluctuations in temperature and humidity were used in the flux calculations. Corrections for the cross-sensitivity of CO₂ analysers to water vapour were also incorporated. Spectral analysis of the corrected CO₂ signal from the closed-path analyser showed that damping of fluctuations in the sampling tube at frequencies f > 0.1 Hz caused the apparent loss in flux. The measured losses can be predicted accurately using theory that describes the damping of oscillations in a sampling tube. High-frequency response of the closed-path system can be improved substantially by ensuring turbulent flow in the tube, using a combination of high volumetric flow rate and small tube diameter. The analysis of attenuation of turbulent fluctuations in flow through tubes is applicable to the measurement of fluxes of other minor atmospheric constituents using the eddy covariance method.     

Eddy-covariance CO2 flux measurements using open- and closed-path CO2 analysers: Corrections for analyser water vapour sensitivity and damping of fluctuations in air sampling tubes

Methods of calibrating infrared CO₂ analysers for sensitivity to CO₂ and water vapour are described. Equations to correct eddy covariance CO₂ flux measurements are presented for: (i) analyser cross-sensitivity to water vapour and the effects of density fluctuations arising from atmospheric fluxes of water vapour and sensible heat, (ii) flux losses caused by signal processing and limited instrument frequency response for open- and closed-path CO₂ analysers, and (iii) flux losses resulting from damping of concentration fluctuations in a tube used to sample air for closed-path CO₂ analysers. Examples of flux corrections required for typical instruments are presented.     

Comparative Measurements Of Co2flux Over A Forest Using Closed-Path And Open-Path Co2analysers

An eddy covariance system using a closed-path CO₂analyser was constructed for long-term CO₂flux measurements above a forest, and its total frequency response was valuated experimentally. The amplitude and phase responses of the system wereexamined through a preliminary test, in which a prescribed pattern of CO₂fluctuation was input to the system. The result showed that the amplitude of the output from the system was attenuated as frequency increased, with a half-power frequency of 0.3 Hz. The phase was delayed by the air sampling through a long tube, but the delay in phase decreased asfrequency increased. We then presented a new technique for the correction of flux loss due to the inadequate system response for the eddy covariance measurements of CO₂ flux. Using the present system and the correction technique, diurnal variations in CO₂ flux were measured over a temperate deciduous forest on three days in 1997. The results were compared with the CO₂fluxes measured with a fast response open-path gas analyser. The CO₂fluxes from the closed-path system agreed with those from the open-path system after the Webb, Pearman and Leuning correction was made for the latter. In the present test over a forest, the contribution of the frequency-response correction to the CO₂fluxes was small and its averaged percentage was only 3%in the daytime. However, the percentage would likely increase, if the system were applied to a shorter vegetation site where high frequency components are more important. The comparison confirmed that we can obtain correct measurements of CO₂flux using the present closed-path system and the correction technique.     

A Numerical Study on Effects of Land-Surface Heterogeneity from ‘Combined Approach’ on Atmospheric Process Part II: Coupling-Model Simulations

Two land surface schemes, one the standard Biosphere / Atmosphere Transfer Scheme Version le (B0Z) and the other B1Z based on B0Z and heterogeneously-treated by ‘combined approach’ , were coupled to the meso-scale model MM4, respectively. Through the calculations of equations from the companion paper, parameters representing land surface heterogeneity and suitable for the coupling models were found out. Three cases were simulated for heavy rainfalls during 36 hours, and the sensitivity of short-term weath-er modeling to the land surface heterogeneity was tested. Through the analysis of the simulations of the three heavy rainfalls, it was demonstrated that B1Z, compared with B0Z, could more realistically reflect the features of the land surface heterogeneity, therefore could more realistically reproduce the circulation and precipitation amount in the heavy rainfall processes of the three cases. This shows that even short-term weather is sensitive to the land surface heterogeneity, which is more obvious with time passing, and whose influence is more pronounced in the lower layer and gradually extends to the middle and upper layer. Through the analysis of these simulations with B1Z, it is suggested that the bulk effect of smaller-scale fluxes (i.e., the momentum, water vapor and sensible heat fluxes) near the significantly-heterogeneous Land surface is to change the larger-scale (i.e., meso-scale) circulation, and then to influence the development of the low-level jets and precipitation. And also, the complexity of the land-atmosphere interaction was shown in these simulations.     

Cross-Validation of Open-Path and Closed-Path Eddy-Covariance Techniques for Observing Methane Fluxes

Methane ( ${\mathrm {CH}}_{4}$ ) fluxes observed with the eddy-covariance technique using an open-path ${\mathrm {CH}}_{4}$ analyzer and a closed-path ${\mathrm {CH}}_{4}$ analyzer in a rice paddy field were evaluated with an emphasis on the flux correction methodology. A comparison of the fluxes obtained by the analyzers revealed that both the open-path and closed-path techniques were reliable, provided that appropriate corrections were applied. For the open-path approach, the influence of fluctuations in air density and the line shape variation in laser absorption spectroscopy (hereafter, spectroscopic effect) was significant, and the relative importance of these corrections would increase when observing small ${\mathrm {CH}}_{4}$ fluxes. A new procedure proposed by Li-Cor Inc. enabled us to accurately adjust for these effects. The high-frequency loss of the open-path ${\mathrm {CH}}_{4}$ analyzer was relatively large (11 % of the uncorrected covariance) at an observation height of 2.5 m above the canopy owing to its longer physical path length, and this correction should be carefully applied before correcting for the influence of fluctuations in air density and the spectroscopic effect. Uncorrected ${\mathrm {CH}}_{4}$ fluxes observed with the closed-path analyzer were substantially underestimated (37 %) due to high-frequency loss because an undersized pump was used in the observation. Both the bandpass and transfer function approaches successfully corrected this flux loss. Careful determination of the bandpass frequency range or the transfer function and the cospectral model is required for the accurate calculation of ${\mathrm {CH}}_{4}$ fluxes with the closed-path technique.     

Fresh water balance of the Gulf Stream system in a regional model study

 We investigate the dependence of surface fresh water fluxes in the Gulf Stream and North Atlantic Current (NAC) area on the position of the stream axis which is not well represented in most ocean models. To correct this shortcoming, strong unrealistic surface fresh water fluxes have to be applied that lead to an incorrect salt balance of the current system. The unrealistic surface fluxes required by the oceanic component may force flux adjustments and may cause fictitious long-term variability in coupled climate models. To identify the important points in the correct representation of the salt balance of the Gulf Stream a regional model of the northwestern part of the subtropical gyre has been set up. Sensitivity studies are made where the westward flow north of the Gulf Stream and its properties are varied. Increasing westward volume transport leads to a southward migration of the Gulf Stream separation point along the American coast. The salinity of the inflow is essential for realistic surface fresh water fluxes and the water mass distribution. The subpolar–subtropical connection is important in two ways: The deep dense flow from the deep water mass formation areas sets up the cyclonic circulation cell north of the Gulf Stream. The surface and mid depth flow of fresh water collected at high northern latitudes is mixed into the Gulf Stream and compensates for the net evaporation at the surface. Received: 19 September 2000 / Accepted: 5 February 2001     

用 IAP／LASG GOALS模式模拟CO2增加引起的东亚地区气候变化

Two simulations, one for the control run and another for the perturbation run, with a global coupled ocean-atmosphere-land system model (IAP / LASG GOALS version 4) have been carried out to study the global warming, with much detailed emphasis on East Asia. Results indicate that there is no climate drift in the control run and at the time of CO2 doubling the global temperature increases about 1.65℃. The GOALS model is able to simulate the observed spatial distribution and annual cycles of temperature and precipitation for East Asia quite well. But, in general, the model underestimates temperature and overestimates rainfall amount for regional annual average. For the climate change in East Asia, the temperature and precipitation in East Asia increase 2. l℃ and 5% respectively, and the maximum warming occurs at middle-latitude continent and the maximum precipitation increase occurs around 25°N with reduced precipitation in the tropical western Pacific.     

Comparison of flux measurements with open- and closed-path gas analyzers above an agricultural field and a forest floor

Comparison was made of the flux measurements of a closed-path CO₂/H₂O analyzer and an open-path H₂O analyzer above a clover field and the forest floor of a Douglas-fir stand. The attenuation of the gas concentration fluctuations caused by the sampling tube of the closed-path analyzer resulted in underestimation of the H20 flux above both surfaces. The degree of underestimation above the clover field depended on wind speed, but was smaller than that calculated from the transfer function for laminar flow in a circular tube and the scalar cospectrum in the neutral and unstable surface layer. Above the forest floor CO₂ fluctuations led those of H₂O by 0.7s. The implications of this are discussed regarding the determination of the time delay caused by the sampling tube of the closedpath analyzer. The day-time CO₂ efflux from the forest floor, averaged over three days, was 0.043 mg/(m²s).     

Ship-board measurements and estimations of air-sea fluxes in the western tropical pacific during TOGA COARE

Direct air-sea flux measurements were made on R/V Kexue #1 at 4 ° S, 156 ° E during the Tropical Ocean Global Atmosphere (TOGA) Coupled Ocean-Atmospheric Response Experiment (COARE) Intensive Observation Period (IOP). An array of six accelerometers was used to measure the motion of the anchored ship, and a sonic anemometer and Lyman-α hygrometer were used to measure the turbulent wind vector and specific humidity. The contamination of the turbulent wind components by ship motion was largely removed by an improvement of a procedure due to Shao based on the acceleration signals. The scheme of the wind correction for ship motion is briefly outlined. Results are presented from data for the best wind direction relative to the ship to minimize flow distortion effects. Both the time series and the power spectra of the sonic-measured wind components show swell-induced ship motion contamination, which is largely removed by the accelerometer correction scheme. There was less contamination in the longitudinal wind component than in the vertical and transverse components. The spectral characteristics of the surface-layer turbulence properties are compared with those from previous land and ocean results. Momentum and latent heat fluxes were calculated by eddy correlation and compared to those estimated by the inertial dissipation method and the TOGA COARE bulk formula. The estimations of wind stress determined by eddy correlation are smaller than those from the TOGA COARE bulk formula, especially for higher wind speeds, while those from the bulk formula and inertial dissipation technique are generally in agreement. The estimations of latent heat flux from the three different methods are in reasonable agreement. The effect of the correction for ship motion on latent heat fluxes is not as large as on momentum fluxes.     

Exploration of the remote sounding of infrared cooling rates due to water vapor

Summary We explore the feasibility of deriving atmospheric infrared cooling rates by direct inversion of radiances observed by satellites from space. In order to convert radiances to fluxes and achieve vertical profiling at the same time, we show that it is necessary to combine radiances from narrow channels with radiances averaged over spectral bands. We demonstrate that the vertical integral of the cooling rate in the spectral band, convolved with a kernel function associated with the narrow channel, can be related to a weighted sum of the channel and band radiances. The band radiance must be evaluated at a specific zenith angle, which is a result of use of the mean value theorem. With known kernel functions, the combined radiances may be inverted to obtain the cooling rate profile. These results are derived from use of a random model for the transmittance in its strong-and weak-line limits. The results are similar in the two limits leading us to conclude that there are expressions that are approximately valid over the entire range of transmittance. We show by numerical methods that this conclusion is correct and apply the retrieval technique successfully to get the cooling rate profile in the rotational band of water vapor.With 3 Figures     

The radiative influence of aerosol effects on liquid-phase cumulus and stratiform clouds based on sensitivity studies with two climate models

Aerosol effects on warm (liquid-phase) cumulus cloud systems may have a strong radiative influence via suppression of precipitation in convective systems. A consequence of this suppression of precipitation is increased liquid water available for large-scale stratiform clouds, through detrainment, that in turn affect their precipitation efficiency. The nature of this influence on radiation, however, is dependent on both the treatment of convective condensate and the aerosol distribution. Here, we examine these issues with two climate models—CSIRO and GISS, which treat detrained condensate differently. Aerosol–cloud interactions in warm stratiform and cumulus clouds (via cloud droplet formation and autoconversion) are treated similarly in both models. The influence of aerosol–cumulus cloud interactions on precipitation and radiation are examined via simulations with present-day and pre-industrial aerosol emissions. Sensitivity tests are also conducted to examine changes to climate due to changes in cumulus cloud droplet number (N _c); the main connection between aerosols and cumulus cloud microphysics. Results indicate that the CSIRO GCM is quite sensitive to changes in aerosol concentrations such that an increase in aerosols increases N _c, cloud cover, total liquid water path (LWP) and reduces total precipitation and net cloud radiative forcings. On the other hand, the radiative fluxes in the GISS GCM appear to have minimal changes despite an increase in aerosols and N _c. These differences between the two models—reduced total LWP in the GISS GCM for increased aerosols, opposite to that seen in CSIRO—appear to be more sensitive to the detrainment of convective condensate, rather than to changes in N _c. If aerosols suppress convective precipitation as noted in some observationally based studies (but not currently treated in most climate models), the consequence of this change in LWP suggests that: (1) the aerosol indirect effect (calculated as changes to net cloud radiative forcing from anthropogenic aerosols) may be higher than previously calculated or (2) lower than previously calculated. Observational constrains on these results are difficult to obtain and hence, until realistic cumulus-scale updrafts are implemented in models, the logic of detraining non-precipitating condensate at appropriate levels based on updrafts and its effects on radiation, will remain an uncertainty.     

一个压力坐标下的海洋环流模式

A new oceanic general circulation model in pressure coordinates is formulated. Since the bottom pressure changes with time, the vertical coordinate is actually a pressure-σ coordinate. The numerical solution of the model is based on an energy-conservation scheme of finite difference. The most important new feature of the model is that it is a truly compressible ocean model and it is free of the Boussinesq approxima tions. Thus, the new model is quite different from many existing models in the following ways: 1) the exact form of mass conservation, 2) the in-situ instantaneous pressure and the UNESCO equation of state to calculate density, 3) the in-situ density in the momentum equations, 4) finite difference schemes that conserve the total energy. Initial tests showed that the model code runs smoothly, and it is quite stable. The quasi-steady circulation patterns generated by the new model compare well with existing models, but the time evolution of the new model seems different from some existing models. Thus, the non-Boussinesq models may provide more accurate information for climate study and satellite observations.     

Measurement of the Sensible Eddy Heat Flux Based on Spatial Averaging of Continuous Ground-Based Observations

Using the standard eddy-covariance (EC) method to quantify mass and energy exchange at a single location usually results in an underestimation of vertical eddy fluxes at the surface. In order to better understand the reasons for this underestimation, an experimental set-up is presented that is based on spatial averaging of air temperature data from a network of ground-based sensors over agricultural land. For eight days during the 34-day observational period in May and June 2007, additional contributions to the sensible heat flux of more than 50Wm⁻² were measured in the lower surface layer by applying the spatial EC method as opposed to the standard temporal EC method. Smaller but still significant additional sensible heat fluxes were detected for four more days. The additional energy is probably transported in organised convective structures resulting in a mean vertical wind velocity unequal to zero at the tower location. The results show that convective transport contributes significantly to the surface energy budget for measurement heights as low as 2–3 m. Since these structures may be quasi-stationary, they can hardly be captured by a single-location measurement. The spatial EC set-up presented here is capable of quantifying contributions to the sensible heat flux from structures up to the scale of our spatial sensor network, which covered an area 3.5 × 3.5 km. For future experiments aiming at closing the energy balance, the spatial EC method should be employed to measure both the sensible and latent heat fluxes. Experimental determination of the horizontal advection of sensible and latent heat should also be considered, since such transport must occur due to convergence and divergence related to convection.     

Structure Function Analysis of Two-Scale Scalar Ramps. Part I: Theory and Modelling

Structure functions are used to study the dissipation and inertial range scales of turbulent energy, to parametrize remote turbulence measurements, and to characterize ramp features in the turbulent field. Ramp features are associated with turbulent coherent structures, which dominate energy and mass fluxes in the atmospheric surface layer. The analysis of structure functions to identify ramp characteristics is used in surface renewal methods for estimating fluxes. It is unclear how commonly observed different scales of ramp-like shapes (i.e., smaller ramps and spikes embedded in larger ramps) influence structure function analysis. Here, we examine the impact of two ramp-like scales on structure function analysis using artificially generated data. The range of time lags in structure function analysis was extended to include time lags typically associated with isotropic turbulence to those larger than the ramp durations. The Van Atta procedure (Arch Mech 29:161–171, 1977) has been expanded here to resolve the characteristics of two-scale ramp models. This new method accurately, and in some cases, exactly determines the amplitude and duration of both ramp scales. Spectral analysis was applied to the structure functions for a broad range of time lags to provide qualitative support for the expanded Van Atta procedure results. The theory reported here forms the foundation for novel methods of analyzing turbulent coherent structures.     

Variability of Sub-Canopy Flow,Temperature, and Horizontal Advection in Moderately Complex Terrain

We examine the space–time structure of the wind and temperature fields, as well as that of the resulting spatial temperature gradients and horizontal advection of sensible heat, in the sub-canopy of a forest with a dense overstorey in moderately complex terrain. Data were collected from a sensor network consisting of ten stations and subject to orthogonal decomposition using the multiresolution basis set and stochastic analyses including two-point correlations, dimensional structure functions, and various other bulk measures for space and time variability. Despite some similarities, fundamental differences were found in the space–time structure of the motions dominating the variability of the sub-canopy wind and temperature fields. The dominating motions occupy similar spatial, but different temporal, scales. A conceptual space–time diagram was constructed based on the stochastic analysis that includes the important end members of the spatial and temporal scales of the observed motions of both variables. Short-lived and small-scale motions govern the variability of the wind, while the diurnal temperature oscillation driven by the surface radiative transfer is the main determinant of the variability in the temperature signal, which occupies much larger time scales. This scale mismatch renders Taylor’s hypothesis for sub-canopy flow invalid and aggravates the computation of meaningful estimates of horizontal advective fluxes without dense spatial information. It may further explain the ambiguous and inconclusive results reported in numerous energy and mass balance and advection studies evaluating the hypothesis that accounting for budget components other than the change in storage term and the vertical turbulent flux improves the budget closure when turbulent diffusion is suppressed in plant canopies. Estimates of spatial temperature gradients and advective fluxes were sensitive to the network geometry and the spatial interpolation method. The assumption of linear spatial temperature gradients was not supported by the results, and leads to increased spatial and temporal variability of inferred spatial gradients and advection estimates. A method is proposed to estimate the appropriate minimum network size of wind and temperature sensors suitable for an evaluation of energy and mass balances by reducing spatial and temporal variability of the spatially sampled signals, which was estimated to be on the order of 200 m at the study site.