Computer Science School
Technical University of Madrid
In recent years, studies aim to understand the biosphere - atmosphere exchange of pollutants and particularly directed to parameterize the deposition processes have increased the interest in the scientific community. Removal pollution processes are an important topic for many purposes. The concept of "critical load" is directly related to the knowledge of these processes. Also, modern atmospheric photochemical simulation models -covering all possible ranges, long-range transport models, mesoscale urban models, etc.- need to know every timestep the flux to be removed from the atmosphere. An accurate parameterization of the deposition velocity is a critical factor for a correct prognostic of the transport processes.
Direct measurements of dry deposition rates using micrometeorological methods have advanced the knowledge of dry deposition processes. However, routine implementation of these methods for yearlong monitoring of dry deposition rates is difficult and no references exist in the scientific literature about it. We present a long-period of measurements of SO2 , O3 and NH3 under the basis of routine dry deposition monitoring. The methodology is already in use in different areas of Europe and USA and it is called Dry Deposition Inferential Measurement (DDIM) technique. In this approach, measured concentrations of pollutants (c) are combined with stimates of the appropriate deposition velocity, such that the flux equals to ctvd. The deposition velocities are stimated by a model using measured meteorological data and the surface characteristics of the site. Uncertainties in predicting hourly deposition velocities are assessed by comparing model stimates with measured deposition velocities from intensive field experiments.
If hourly estimates of vd are combined with hourly measurements of the pollutant concentration, then the weekly integrated flux is the following:
However, hourly measurements of c for many chemical species requires the implementation of complex chemical analyser that require frequent maintenance, calibration and climate-controlled enclosures. However, this approach is useful if the deposition velocity model has been tested by using complex micrometeorological instrumentation. Eddy-correlation techniques are able to be used on a very few pollutants. We have used this method for the ozone value and the modified Bowen-ratio method for the SO2 and NH3 . These methods allow a very accurate computation of the pollutant fluxes deposited over the soil. In the case of Ammonia, the emission/deposition simultaneous process requires the inclusion of the concept "canopy composition point" which changes slightly the value of the resistances to be applied to the "classical" resistances approach. The objective of this page is to show some results to compare the deposition velocities modelled with those measured directly by fast/slow chemical analysers. The models used take into account several characteristics of the surface and turbulent conditions of the area, such as the micrometeorological turbulent characterisation.
The field experiment was conducted at the Military Airport "Cuatro Vientos" (at the SW of Madrid, belonging to its metropolitan area). That site can be considered highly polluted because of the influence of the city pollution. Also big open areas are present on the West side of the place which allows influence of the agricultural areas in the Madrid domain. The weather conditions at these locations are typical continental, characterised by dry and hot summer periods and cold winter periods. Rain periods are mainly on October-November and April-May.
The O3 dry deposition fluxes are evaluated by using eddy-correlation method. It requires fast response and sensitive instrumentation to monitor the small rapid fluctuations of the species of interest. The components of wind velocity were measured at 5 m. using a 3-axis sonic anemometer. The O3 fluctuations were measured with fast response chemiluminiscence sensor (GFAS Ozone Sonde OS-G-2) which was installed in the same mast, sampling air through a Teflon tube as close as possible of the sonic anemometer. Its response time is 0.1 s. This sensor was used together with an absolute ozone reference monitor (Ultra Violet Detector 49 PS Thermo Environmental Ins. Inc.) in order to recalibrate the fast sensor every 15 min. Its response time is 30 s. The system is able to obtain the absolute value of O3 concentration and its turbulent fluctuation. SO2 concentrations were sampled by a chemical sensor (Pulsed Fluorescent Detector 43S Thermo Environmental Ins. Inc.) was installed near the mast to evaluate SO2 concentration at two different levels (5m. and 1m.). The air sampling were conducted through two Teflon tubes connected to a pump system placed in a metallic box near the mast which was controlled by a portable computer; the sampling levels were alternated every 2 min. The response time of the instrument was 30 s. A system calibration was performed automatically every 24 hours taken air zero from a column coated of active carbon, the value obtained in this process is used to recalibrate the absolute value of SO2 concentration during the next period of 24 by software.
The sonic anemometer was used to parameterize the turbulent surface layer. Slow meteorological sensors were used in order to obtain a complete information of surface boundary layer. A relative humidity and absolute temperature (Rhotronic MP300, Campbell Ins.) sensor was placed 3m. in height at the same mast. Absolute temperature was evaluated using a thermocoupled sensor and relative humidity as a function of the conductivity in a polymeric material. Its response time is 25 s. Moreover, a global solar radiation (Solarimeter Casella) was placed at the top of a near small mast, 2 m. in height, its response time was 15 s. Complete instrumentation was controlled by a notebook personal computer where a complete software for data acquisition was implemented. This control unit was designed and developed by ENC (Netherlands Energy Research Foundation). All instrumentation was calibrated in a previous period following the E.P.A. standard. The system has been designed for routine work and maintenance was kept as low as possible. The GFAS Sonde had to be changed every 7 days and the computer storage capacity limits data acquisition for longer periods.
Amonnia concentrations were determined by using a continuous annular denuder system. This system was located in a separated scaffold where measurements were made at 3 heights (0.45 m., 1.25 m. and 3.3 m.). A denuder tube was placed at each level, and only one detector was used to minimize bias between denuders. The denuders sample air at 30 lmin-1 . They are rotating continuously and are supplied with an acidic stripping solution (HSO4Na), which the collected NH3 was brought (as NH4+ in the stripping solution) to a common detector for analysis. Analysis of the NH4+ in this system is made by membrane diffusion of NH3 at high pH into a counter flow of deionized water, with subsequent measurement by conductivity. Its response time is 2 min. per level, which its 90 first seconds are used to stabilised the signal and only the last 30 s. are used to perform the measurement. In order to obtain an accurate determination of NH3 concentration this system needs a periodic calibration. It includes calibration of the detector (through 3 standard calibration solutions, 0 ppb, 50 ppb and 500 ppb) and a correct evaluation of the gas flow in every denuder. The system was controlled by a data logger which was able to store values of conductivity, liquids flows and temperature into the detector for a period no longer than 20 days. Values of NH3 concentration will be obtained with an external program through conductivity values as a function of the calibration parameters.
Fluxes were evaluated by eddy-correlation and modified Bowen ration techniques. The eddy-correlation technique allows us to measure directly the momentum, sensible heat and ozone fluxes. The flux, Fs, of a quantity, s, can be evaluated by measuring the covariances of the product of the vertical wind speed fluctuations and the species concentration fluctuations. The deposition velocity is directly inferred from:
The indirect method to evaluate fluxes over the canopy is the modified Bowen ratio technique which is based on the similarity between flux ratios and gradient ratios in the following way:
where Fc is the flux of the species, , is the sensible heat flux and, , is the temperature difference between the heights z2 and z1. This difference is evaluated by using
where L is the Monin-Obukhov length, , is the scale temperature of the surface boundary layer. The O3 deposition velocity was determined directly by using the eddy-correlation technique. The SO2 and ammonia deposition velocities were determined by the modified Bowen ration technique. Additionally, these deposition velocities were parameterized by using the Wesely, Erisman parameterizations and results from the CIBA experiment (taken place in Spain by Roberto San José). All these parameterizations focus on the generalisation of the deposition fluxes over large areas based on meteorological variables, canopy characteristics and land-use classification.
The Wesely model parameterizes the canopy resistance following:
the different expressions for the resistances are:
A modification of this model was made by San José and his Group. With that modification it is provided much more influence to the stomatal resistance by changing the parameters to 0.0003, 0.0006, 0.0003 respectively. This procedure allows a much more sensitivity to the stomatal gas-phase processes and the influence of the corresponding resistance.
Finally, Erisman proposed a different canopy resistance parameterization, as follows:
where rs can be parameterized following Wesely. We have used the Wesely parameterization of the stomatal resistance. rinc depends on the LAI (Leaf Area Index), the height of the canopy and the friction velocity. Finally, the model parameterizes the rsoil in function of the relativity humidity and the soil pH. This model remarks the influence of the soil resistance based on the soil pH and as a consequence based on the acidity of the pollutant.
A field study on O3, SO2, and NH3 deposition over a suburban area Madrid case study.
Roberto San José, F. Javier Moreno and M. Angeles San Feliú.
AIR POLLUTION AND VISIBILITY MEASUREMENTS (Jun 95)