Lagrangian Stochastic Dispersion Model Research Articles

Contaminant dispersion prediction and source estimation with integrated Gaussian-machine learning network model for point source emission in atmosphere

Gas dispersion model is important for predicting the gas concentrations when contaminant gas leakage occurs. Intelligent network models such as radial basis function (RBF), back propagation (BP) neural network and support vector machine (SVM) model can be used for gas dispersion prediction. However, the prediction results from these network models with too many inputs based on original monitoring parameters are not in good agreement with the experimental data. Then, a new series of machine learning algorithms (MLA) models combined classic Gaussian model with MLA algorithm has been presented. The prediction results from new models are improved greatly. Among these models, Gaussian-SVM model performs best and its computation time is close to that of classic Gaussian dispersion model. Finally, Gaussian-MLA models were applied to identifying the emission source parameters with the particle swarm optimization (PSO) method. The estimation performance of PSO with Gaussian-MLA is better than that with Gaussian, Lagrangian stochastic (LS) dispersion model and network models based on original monitoring parameters. Hence, the new prediction model based on Gaussian-MLA is potentially a good method to predict contaminant gas dispersion as well as a good forward model in emission source parameters identification problem.

Impact of the electric vehicles on the air pollution from a highway

We have quantified the impact that an introduction of electric vehicles into the car fleet has on air quality (regarding NO and NO2) using a numerical dispersion model. An experimental campaign is conducted close to a highway in Milan, Italy. Meteorological parameters and chemical concentrations are measured along with the traffic emissions. We use a Lagrangian Stochastic Dispersion Model to create numerical simulations of the chemical reactions and dispersion involving pollutants from the highway. To evaluate the air pollution reductions, emission scenarios with different rates of electric vehicles introduction are simulated. We have found that only a significant replacement (50%) of non-electric vehicles with electric ones yields a remarkable reduction of the pollutant concentrations. However, even with lower electric vehicles introduction rates, the air quality improvements may be relevant during intense pollution episodes. The results provide useful information to decision makers and public administrators for planning measures to modify the car fleet composition aiming to improve the urban air quality.

A simple two-dimensional parameterisation for Flux Footprint Prediction (FFP)

Abstract. Flux footprint models are often used for interpretation of flux tower measurements, to estimate position and size of surface source areas, and the relative contribution of passive scalar sources to measured fluxes. Accurate knowledge of footprints is of crucial importance for any upscaling exercises from single site flux measurements to local or regional scale. Hence, footprint models are ultimately also of considerable importance for improved greenhouse gas budgeting. With increasing numbers of flux towers within large monitoring networks such as FluxNet, ICOS (Integrated Carbon Observation System), NEON (National Ecological Observatory Network), or AmeriFlux, and with increasing temporal range of observations from such towers (of the order of decades) and availability of airborne flux measurements, there has been an increasing demand for reliable footprint estimation. Even though several sophisticated footprint models have been developed in recent years, most are still not suitable for application to long time series, due to their high computational demands. Existing fast footprint models, on the other hand, are based on surface layer theory and hence are of restricted validity for real-case applications. To remedy such shortcomings, we present the two-dimensional parameterisation for Flux Footprint Prediction (FFP), based on a novel scaling approach for the crosswind distribution of the flux footprint and on an improved version of the footprint parameterisation of Kljun et al. (2004b). Compared to the latter, FFP now provides not only the extent but also the width and shape of footprint estimates, and explicit consideration of the effects of the surface roughness length. The footprint parameterisation has been developed and evaluated using simulations of the backward Lagrangian stochastic particle dispersion model LPDM-B (Kljun et al., 2002). Like LPDM-B, the parameterisation is valid for a broad range of boundary layer conditions and measurement heights over the entire planetary boundary layer. Thus, it can provide footprint estimates for a wide range of real-case applications. The new footprint parameterisation requires input that can be easily determined from, for example, flux tower measurements or airborne flux data. FFP can be applied to data of long-term monitoring programmes as well as be used for quick footprint estimates in the field, or for designing new sites.

Testing a mechanistic dispersal model against a dispersal experiment with a wind‐dispersed moss

Wind is the main dispersal agent for a wide array of species and for these species the environmental conditions under which diaspores are released can potentially modify the dispersal kernel substantially. Little is known about how bryophytes regulate spore release, but conditions affecting peristome movements and vibration of the seta may be important. We modelled airborne spore dispersal of the bryophyte species Discelium nudum (spore diameter 25 μm), in four different release scenarios, using a Lagrangian stochastic dispersion model and meteorological data. We tested the model predictions against experimental data on colonization success at five distances (5, 10, 30, 50 and 100 m) and eight directions from a translocated point source during seven two‐day periods. The model predictions were generally successful in describing the observed colonization patterns, especially beyond 10 m. In the laboratory we established spore release thresholds; horizontal wind speed sd > 0.25 m s−1 induced the seta to vibrate and in relative humidity < 75% the peristome was open. Our dispersal model predicts that the proportion of spores dispersing beyond 100 m is almost twice as large if the spores are released under turbulent conditions than under more stable conditions. However, including release thresholds improved the fit of the model to the colonization data only minimally, with roughly the same amount of variation explained by the most constrained scenario (assuming both vibration of the seta and an open peristome) and the scenario assuming random release. Model predictions under realised experimental conditions suggest that we had a low statistical power to rank the release scenarios due to the lack of measurements of the absolute rate of spore release. Our results hint at the importance of release conditions, but also highlight the challenges in dispersal experiments intended for validating mechanistic dispersal models.

Derivation of the Turbulent Time Scales and Velocity Variances from LES Spectral Data: Application in a Lagrangian Stochastic Dispersion Model

Turbulent time scales and velocity variances for a convective boundary layer are derived from large eddy simulation spectral data. Spectral peak frequencies obtained from LES data are used directly in expressions that allow establishing such times scales and velocity variances. These turbulent parameters were compared with those provided by experimental turbulence data. The comparison employing a stochastic dispersion model and observed concentration data shows that both parameterizations reproduce adequately the contaminant dispersion process in a convective boundary layer.

Lagrangian Stochastic Modelling of Dispersion in the Convective Boundary Layer with Skewed Turbulence Conditions and a Vertical Density Gradient: Formulation and Implementation in the FLEXPART Model

A correction for the vertical gradient of air density has been incorporated into a skewed probability density function formulation for turbulence in the convective boundary layer. The related drift term for Lagrangian stochastic dispersion modelling has been derived based on the well-mixed condition. Furthermore, the formulation has been extended to include unsteady turbulence statistics and the related additional component of the drift term obtained. These formulations are an extension of the drift formulation reported by Luhar et al. (Atmos Environ 30:1407–1418, 1996) following the well-mixed condition proposed by Thomson (J Fluid Mech 180:529–556, 1987). Comprehensive tests were carried out to validate the formulations including consistency between forward and backward simulations and preservation of a well-mixed state with unsteady conditions. The stationary state CBL drift term with density correction was incorporated into the FLEXPART and FLEXPART-WRF Lagrangian models, and included the use of an ad hoc transition function that modulates the third moment of the vertical velocity based on stability parameters. Due to the current implementation of the FLEXPART models, only a steady-state horizontally homogeneous drift term could be included. To avoid numerical instability, in the presence of non-stationary and horizontally inhomogeneous conditions, a re-initialization procedure for particle velocity was used. The criteria for re-initialization and resulting errors were assessed for the case of non-stationary conditions by comparing a reference numerical solution in simplified unsteady conditions, obtained using the non-stationary drift term, and a solution based on the steady drift with re-initialization. Two examples of “real-world” numerical simulations were performed under different convective conditions to demonstrate the effect of the vertical gradient in density on the particle dispersion in the CBL.

Use of open-path FTIR and inverse dispersion technique to quantify gaseous nitrogen loss from an intensive vegetable production site

An open-path Fourier transform infrared (OP-FTIR) spectroscopic technique in combination with a backward Lagrangian stochastic (bLS) dispersion model (WindTrax) can be used to simultaneously measure gaseous emissions of N2O, NH3, CH4 and CO2. We assessed the capability of this technique for measuring NH3 and N2O emissions following the application of calcium nitrate (Ca(NO3)2), Nitrophoska (NPK) and chicken manure on a celery farm at Boneo, Victoria, during April and May 2013. We found that the OP-FTIR/WindTrax method was able to measure the diurnal variation in NH3 flux from the field site following application of chicken manure with measured emissions ranging from approximately 0.1–9.8 kg NH3–N ha−1 day−1. The OP-FTIR/WindTrax method also detected a diurnal variation in N2O flux of 1.5–6.2 kg N2O–N ha−1 day−1 and N2O flux increased in response to application of the Ca(NO3)2. We concluded that the OP-FTIR/WindTrax technique can quantify gaseous N loss from vegetable production systems.

Do small spores disperse further than large spores?

In species that disperse by airborne propagules an inverse relationship is often assumed between propagule size and dispersal distance. However, for microscopic spores the evidence for the relationship remains ambiguous. Lagrangian stochastic dispersion models that have been successful in predicting seed dispersal appear to predict similar dispersal for all spore sizes up to -40 microm diameter. However, these models have assumed that spore size affects only the downwards drift of particles due to gravitation and have largely omitted the highly size-sensitive deposition process to surfaces such as forest canopy. On the other hand, they have assumed that spores are certain to deposit when the air parcel carrying them touches the ground. Here, we supplement a Lagrangian stochastic dispersion model with a mechanistic deposition model parameterized by empirical deposition data for 1-10 microm spores. The inclusion of realistic deposition improved the ability of the model to predict empirical data on the dispersal of a wood-decay fungus (aerodynamic spore size 3.8 microm). Our model predicts that the dispersal of 1-10 microm spores is in fact highly sensitive to spore size, with 97-98% of 1 microm spores but only 12-58% of 10-microm spores dispersing beyond 2 km in the simulated range of wind and canopy conditions. Further, excluding the assumption of certain deposition at the ground greatly increased the expected dispersal distances throughout the studied spore size range. Our results suggest that by evolutionary adjustment of spore size, release height and timing of release, fungi and other organisms with microscopic spores can change the expected distribution of dispersal locations markedly. The complex interplay of wind and canopy conditions in determining deposition resulted in some counterintuitive predictions, such as that spores disperse furthest under intermediate wind, providing intriguing hypotheses to be tested empirically in future studies.

Comparative study to evaluate three ground-based optical remote sensing techniques under field conditions by a gas tracer experiment

Today, ground-based optical remote sensing (ORS) has become an intensively used method for quantifying pollutant or greenhouse gas emissions from point or area sources, and for the validation of airborne or satellite remote sensing data. In this study, we present the results of a release experiment using acetylene (C2H2) as a tracer gas, where three ORS techniques were simultaneously tested for two main purposes: (1) the detection of emission sources and (2) the quantification of release rates. Therefore, passive and active open-path Fourier transform infrared spectroscopy (OP-FTIR) and open-path tunable diode laser absorption spectroscopy (TDLAS) were applied and evaluated. The concentration results of the active ORS methods are compared to those estimated by a Lagrangian stochastic atmospheric dispersion model. Our results reveal that passive OP-FTIR is a valuable tool for the rapid detection and imaging of emission sources and the spatial tracer gas distribution; while with active OP-FTIR and TDLAS, C2H2 concentrations in the sub-ppm range could be quantified that correlated well with the concentration data obtained by our modeling approach.

Measuring Greenhouse-Gas Emissions from a Synthetic Tracer Source

This paper describes a controlled tracer-release experiment in which methane was released from a synthetic source at known rates. An open-path Fourier transform infrared (OP-FTIR) spectroscopy system was used to measure line-averaged methane concentrations downwind of the source. A Lagrangian stochastic (LS) dispersion model was employed to infer emission rates from downwind gas concentrations. The main purpose of our study was to investigate the ability of our open-path FTIR system combined with the LS dispersion model to accurately measure greenhouse gas emissions. In our study, the average ratio of the estimated emissions to actual release rates QLS/Q for CH4 was about 0.86 (σQLS/Q = 0.2, n = 6) and 0.84 (σQLS/Q = 0.22, n = 3) after data fi ltering for a 15- and a 30- min period, respectively. Although there is a limited amount of data in this experiment, the results demonstrate the potential of the measurement system for accurate quantifi cation of greenhouse gas emissions.

A Velocity–Dissipation Lagrangian Stochastic Model for Turbulent Dispersion in Atmospheric Boundary-Layer and Canopy Flows

An extended Lagrangian stochastic dispersion model that includes time variations of the turbulent kinetic energy dissipation rate is proposed. The instantaneous dissipation rate is described by a log-normal distribution to account for rare and intense bursts of dissipation occurring over short durations. This behaviour of the instantaneous dissipation rate is consistent with field measurements inside a pine forest and with published dissipation rate measurements in the atmospheric surface layer. The extended model is also shown to satisfy the well-mixed condition even for the highly inhomogeneous case of canopy flow. Application of this model to atmospheric boundary-layer and canopy flows reveals two types of motion that cannot be predicted by conventional dispersion models: a strong sweeping motion of particles towards the ground, and strong intermittent ejections of particles from the surface or canopy layer, which allows these particles to escape low-velocity regions to a high-velocity zone in the free air above. This ejective phenomenon increases the probability of marked fluid particles to reach far regions, creating a heavy tail in the mean concentration far from the scalar source.

Derivation of third–order vertical velocity turbulence moment in the convective boundary layer from large eddy simulation data: an application to the dispersion modeling

A new formulation for the vertical turbulent velocity third statistical moment in a convective boundary layer is proposed. The parameterization is based directly on the definition of this higher order moment, with velocity skewness and variance being calculated from large eddy simulation data. The formulation, included in a Lagrangian stochastic dispersion model, has been tested and compared with expressions for the third moment obtained from experimental data and reported in the literature, using concentration data from field experiments. The application of a statistical evaluation shows that the proposed parameterization has one of the best overall adjustments to the data.

Inverse dispersion modelling highlights the efficiency of slurry injection to reduce ammonia losses by agriculture in the Po Valley (Italy)

The performance of two inverse dispersion models to quantify ammonia (NH3) fluxes has been evaluated. Ammonia emissions from cattle slurry spreading have been estimated following two different application techniques in northern Italy: (i) surface spreading with incorporation after 24h and (ii) injection into the soil. Passive diffusion samplers were used to measure atmospheric ammonia concentrations at the treated fields and at background sites. A sonic anemometer was used to characterise the atmospehric turbulence and provided the necessary inputs for the inverse models. Two inverse dispersion models were then employed: FIDES, based on the Philip (1959) analytical solution of the advection–diffusion equation and WindTrax, a Lagrangian stochastic dispersion model. Both models were parameterised with friction velocity, Obukhov length and roughness as input. WindTrax was also alternatively used with wind components statistics from the sonic anemometer. The two models agreed well in estimating ammonia emissions under neutral conditions but diverged under stable and unstable conditions because of different parameterisation of the diffusivity profile. The maximum difference between the two models was less than 32%. The uncertainty associated with the concentration measurements with the passive samplers, was lower than 11%. The direct injection of slurry into the soil was the most effective method for reducing N loss from slurry application, resulting in an emission of 2% of the total ammoniacal nitrogen (TAN) applied when surface application resulted in losses of 5.4% of TAN.

Effect of the Representative Source Area for Eddy Covariance Measuraments on Energy Balance Closure for Maize Fields in the Po Valley, Italy

This paper has as main objective to show the effect of the representative source area for eddy covariance measurements (called footprint) on energy balance closure. Energy balance closure was evaluated by a statistical regression of turbulent energy fluxes (sensible and latent heat) against available energy (net radiation and soil ground heat flux). The footprint was calculated using an approximate analytical model based on a combination of Lagrangian stochastic dispersion model and dimensional analysis. The data were measured by two eddy covariance towers located on maize fields in Landriano (PV) and Livraga (LO) at the Po Valley, Italy. The main results obtained using only the flux data which have a source area included into the cultivated field shows that there is a slight improvement on the energy balance closure. The stability conditions of the atmosphere plays a fundamental role on the slope of the linear regression and on footprint size, in particular way, it is shown when the energy balance closure is analysed for different sectors of the field in function of the wind directions.

Determination of field scale ammonia emissions for common slurry spreading practice with two independent methods

Abstract. At a cropland and a grassland site field scale ammonia (NH3) emissions from slurry application were determined simultaneously by two approaches based on (i) eddy covariance (EC) flux measurements using high temperature Chemical Ionisation Mass Spectrometry (HT-CIMS) and on (ii) backward Lagrangian Stochastic (bLS) dispersion modelling using concentration measurements by three optical open path Fourier Transform Infrared (FTIR) systems. Slurry was spread on the fields in sequential tracks over a period of one to two hours. In order to calculate field emissions, measured EC/HT-CIMS fluxes were combined with flux footprint analysis of individual slurry spreading tracks to parameterise the NH3 volatilisation with a bi-exponential time dependence. Accordingly, track-resolved concentration footprints for the FTIR measurements were calculated using bLS. A consistency test with concentrations measured by impingers showed very low systematic deviations for the EC/HT-CIMS results (<8%) but larger deviations for the bLS/FTIR results. For both slurry application events, the period during fertilisation and the subsequent two hours contributed by more than 80% to the total field emissions. Averaged over the two measurement methods, the cumulated emissions of the first day amounted to 17 ± 3% loss of applied total ammoniacal nitrogen over the cropland and 16 ± 3% over the grassland field.

Employing a Lagrangian stochastic dispersion model and classical diffusion experiments to evaluate two turbulence parameterization schemes

Considering a Lagrangian stochastic particle dispersion model and diffusion experiments, two turbulence parameterizations have been tested. The parameters obtained from the Taylor turbulence parameterization are derived from observed spectral properties and characteristics of energy containing eddies. Taylor turbulence scheme provides continuous values for the turbulent parameters in the planetary boundary layer. Hanna turbulence parameterization is obtained from theoretical considerations and second–order closure models and it does not provide continuous vertical profiles for the turbulent parameters. The predicted values by a Lagrangian diffusion model utilizing Taylor turbulence parameterization and the Hanna turbulence scheme are compared with observed concentration data from diffusion experiments. The use of Taylor turbulence scheme resulted in better results.

Regional Source Identification Using Lagrangian Stochastic Particle Dispersion and HYSPLIT Backward-Trajectory Models

ABSTRACT The main objective of this study was to investigate the capabilities of the receptor-oriented inverse mode Lagrangian Stochastic Particle Dispersion Model (LSPDM) with the 12-km resolution Mesoscale Model 5 (MM5) wind field input for the assessment of source identification from seven regions impacting two receptors located in the eastern United States. The LSPDM analysis was compared with a standard version of the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) single-particle backward-trajectory analysis using inputs from MM5 and the Eta Data Assimilation System (EDAS) with horizontal grid resolutions of 12 and 80 km, respectively. The analysis included four 7-day summertime events in 2002; residence times in the modeling domain were computed from the inverse LSPDM runs and HYPSLIT-simulated backward trajectories started from receptor-source heights of 100, 500, 1000, 1500, and 3000 m. Statistics were derived using normalized values of LSPDM- and HYSPLIT-predicted residence times versus Community Multiscale Air Quality model-predicted sulfate concentrations used as baseline information. From 40 cases considered, the LSPDM identified first- and second-ranked emission region influences in 37 cases, whereas HYSPLIT-MM5 (HYSPLIT-EDAS) identified the sources in 21 (16) cases. The LSPDM produced a higher overall correlation coefficient (0.89) compared with HYSPLIT (0.55–0.62). The improvement of using the LSPDM is also seen in the overall normalized root mean square error values of 0.17 for LSPDM compared with 0.30–0.32 for HYSPLIT. The HYSPLIT backward trajectories generally tend to underestimate near-receptor sources because of a lack of stochastic dispersion of the backward trajectories and to overestimate distant sources because of a lack of treatment of dispersion. Additionally, the HYSPLIT backward trajectories showed a lack of consistency in the results obtained from different single vertical levels for starting the backward trajectories. To alleviate problems due to selection of a backward-trajectory starting level within a large complex set of 3-dimensional winds, turbulence, and dispersion, results were averaged from all heights, which yielded uniform improvement against all individual cases. IMPLICATIONS Backward-trajectory analysis is one of the standard procedures for determining the spatial locations of possible emission sources affecting given receptors, and it is frequently used to enhance receptor modeling results. This analysis simplifies some of the relevant processes such as pollutant dispersion, and additional methods have been used to improve receptor-source relationships. A methodology of inverse Lagrangian stochastic particle dispersion modeling was used in this study to complement and improve standard backward-trajectory analysis. The results show that inverse dispersion modeling can identify regional sources of haze in national parks and other regions of interest.

Comparison of open-circuit respiration chambers with a micrometeorological method for determining methane emissions from beef cattle grazing a tropical pasture

Emissions of CH4 methane from beef cattle grazing pastures characteristic of northern Australia are yet to be reliably quantified. Poor quality pastures and low animal productivity are typically associated with high CH4 emissions/unit animal product. Our study reports two independent methodologies to quantify CH4 emissions from beef cattle at herd scale and from individual animals. The first method identifies each animal within a herd (n=18) as a point source of CH4. Emissions at the herd scale were estimated with a backward Lagrangian stochastic (bLs) dispersion model. This model used atmospheric turbulence data and line-averaged concentrations of CH4 measured with an open-path laser and inputs (ppm) from 5 paths across 1ha Rhodes grass (Chloris gayana) dominated pastures that were flat and free of obstructions. The second method used open circuit respiration chambers to measure 24h CH4 emissions from 18 steers fed freshly cut forage from the same pasture. The CH4 emissions determined using the dispersion method were 136.1±21.5g/d or 29.7±3.70g/kg dry matter (DM) intake and 114±5.1g/d or 30.1±2.19g/kg DM intake using open circuit respiration chambers. Estimates of CH4 emissions based on predictive algorithms for beef cattle grazing similar pastures range from 82.7±3.98g/d to 112.7±2.57g/d. Additional studies with different tropical pastures and herds are required to obtain estimated CH4 emission values for further comparisons with predictions from the Australian Greenhouse Gas Inventory and those currently reported in the literature for beef cattle.This paper is part of the special issue entitled: Greenhouse Gases in Animal Agriculture – Finding a Balance between Food and Emissions, Guest Edited by T.A. McAllister, Section Guest Editors; K.A. Beauchemin, X. Hao, S. McGinn and Editor for Animal Feed Science and Technology, P.H. Robinson.

Predicting seed admixture in maize combining flowering characteristics and a Lagrangian stochastic dispersion model

The seed purity problem has received very low attention in maize coexistence studies, and has not been considered in prediction tools that simulate pollen flow and out-crossing between maize fields. To fill this gap we developed the Seed Admixture Model (SAMETH) able to predict seed admixture dispersion combining flowering characteristics (pollen shed, silks exertion) with a Lagrangian stochastic dispersion model. The model was tested with a dataset obtained from 20 fields in 2007 and in 2008, whose seeds were mixed with 1% of a homozygous blue-kernelled hybrid. The model was first calibrated with data from 6 fields and then validated with the data from the remaining 14 fields. Moreover, a sensitivity analysis was performed to test the consequences of different pollen quantities released by the admixture and the commercial hybrids. The measured seed admixture ranged from 0.7% to 6% and the model was able to simulate the seed admixture with r 2 = 0.83. The sensitivity analysis showed that the model was sensitive to the absolute released pollen quantities but it was still able to predict seed admixture rather accurately. Because of its reliability, the model could become a useful tool for case study scenarios that involve seed admixture and for which field implementations would be too complex and time-consuming.

On concentration footprints for a tall tower in the presence of a nocturnal low-level jet

This paper reports on the location of sources contributing to a point concentration measurement in the stable boundary layer. Concentration footprints for sensors located at different heights during the night are examined using a backward-in-time Lagrangian stochastic dispersion model. Simulations of air parcel transport in a non-steady-state atmospheric boundary layer above a bare surface and a forest under different insolation suggest that sources located as far as several hundred kms away contribute to a concentration measurement made at levels as high as 500 m. The origin of the maximal contribution area shifts during the night depending on the wind direction at the sensor location, a feature most prominent in the presence of decoupling between sensor levels and surface sources. Simulations suggest that atmospheric static stability alone is not a sufficient criterion to trigger flow decoupling. The presence of the low-level jet provides vertical mixing of air parcels even for stable boundary layers with Richardson numbers higher than the critical. This finding is in agreement with earlier observations suggesting that the level of the low-level jet nose acts as a strong lid prohibiting vertical gas propagation. These findings have important implications in the interpretation of eddy-flux measurements in nocturnal conditions.