Atmospheric Environment (v.41, #16)

Flow field and concentration measurements have been performed in an idealized model of an urban street canyon with one row of trees arranged along the center axis. The model was set up in an atmospheric boundary layer wind tunnel and the approach flow was directed perpendicular to the street axis. A line source embedded in the bottom of the street was used to release tracer gas for the simulation of traffic exhaust emissions. Trees with spherical crowns were modeled and positioned inside the street canyon, varying crown diameter, crown permeability, trunk height and tree spacing. Traffic-induced turbulence was simulated by rotating belts with thin plates. Concentrations were measured at the facades of the street canyon. For small tree crowns, only little changes in concentration were measured, however, increasing crown diameters led to increasing concentrations at the leeward street canyon wall associated with a reduction of local concentrations at the windward wall. For some cases, a variation of trunk height led to a modification of the concentration pattern on the walls. Increasing the tree spacing resulted in a noticeable concentration decrease. When compared to the situation with standing (but emitting) traffic, the traffic-induced turbulence by two-way car movements always contributed to a more homogenous concentration field inside the street canyon yielding to reduced mean concentration levels.
Keywords: Street canyon; Pollutant dispersion; Trees; Canyon vortex; Traffic-induced turbulence;

Modelling photochemical reactions in atmospheric water droplets: An assessment of the importance of surface processes by Davide Vione; Claudio Minero; Ahmed Hamraoui; Mireille Privat (3303-3314).
The relative importance of surface versus bulk processes was computed in spherical water drops, based on the model photochemical reaction yielding phenol from benzene via the OH radicals, formed on UV photolysis of nitrate. The coadsorption of benzene and nitrate at the droplet surface was taken into account by means of the Wagner–Onsager–Samaras interaction model. The results indicate that the surface accumulation of benzene and the coadsorption of benzene and nitrate, which substantially increase the value of the concentration product of the model species at the droplet surface compared to the bulk, would cause surface processes to play a very significant role, in particular in small drops (1 μm radius, as can be found in haze and mist). Under such circumstances, over 15% of phenol photoformation would take place in just 0.1% of the drop volume. Similar results are expected in the case of benzene hydroxylation and nitration on nitrite photolysis.
Keywords: Atmospheric photochemistry; Aerosols; Surface layer;

Regional scale modelling of particulate matter in the UK, source attribution and an assessment of uncertainties by J.D. Whyatt; S.E. Metcalfe; J. Nicholson; R.G. Derwent; T. Page; J.R. Stedman (3315-3327).
The models HARM and ELMO are used to investigate the importance of different source categories contributing to total PM10 (SIA, SOA and primary particulate matter) across the UK and the impact of uncertainties on both present day and future concentration estimates. Modelled concentrations of SIA (secondary inorganic aerosol) are compared against data from the UK's Nitric Acid and Aerosol Network and SOA (secondary organic aerosol) against measurements made at the Bush Estate, Edinburgh. These data indicate that the HARM/ELMO modelling approach comes close to achieving mass closure. Comparison with national maps of total PM10 indicate that the models underestimate particulate matter concentrations around large conurbations, probably due to the localised nature of emissions of primary particulates in these areas and model scale. The models are used to attribute particulate matter to different source and size categories, assessing the relative importance of primaries, SIA and SOA; the contributions of anthropogenic and biogenic precursors of SOA; the relative importance of PMcoarse (PM10–PM2.5) and PMfine (PM2.5) and UK vs. other EMEP area sources. The implications of these attributions for emissions control policies are discussed. The impact of uncertainties in emissions of the sources of primaries, SIA and SOA are explored. For primary PM10 and SOA this has been achieved through emissions scaling and for SIA using the GLUE (Generalised Likelihood Uncertainty Estimation) approach. The selection of acceptable model parameter sets has been based on the need to retain the capability to model deposition of S and N species. The impact of uncertainty on estimates of present day SIA concentrations is illustrated for sites in the Nitric Acid and Aerosol Network. A more limited assessment for 2010 has been carried out at the national scale, illustrating that inclusion of uncertainty can change modelled concentrations from no exceedance of current air quality objectives, to one of exceedance over large areas of south and east England.
Keywords: PM10; PM2.5; Secondary inorganic aerosol; Secondary organic aerosol;

Size distributions and condensation growth of submicron particles in on-road vehicle plumes in Hong Kong by Xiaohong Yao; Ngai Ting Lau; Chak K. Chan; Ming Fang (3328-3338).
Volume concentration data of submicron particles ( 0.1 – 1.0 μ m ) in on-road vehicle plumes ( NO x > 400 ppb ) gathered by the Mobile Real-time Air Monitoring Platform (MAP) on city streets, highways and in tunnels in Hong Kong are used to study the size distributions and growth of vehicular submicron particles due to gas condensation and, particularly, its dependency on ambient temperature. Three particle volume size distributions are observed: a uni-modal distribution with an accumulation mode at 0.2 ± 0.1 μ m , and two bi-modal distributions with a minor mode at 0.2 ± 0.1 μ m and the dominant mode at either 0.5 ± 0.1 or 0.7 ± 0.1 μ m . In tunnels, the ratio of particle volume concentration to black carbon (BC) mass concentration correlates negatively with ambient temperature ( r 2 = 0.57 ) ; the dominant mode tends to be at the large particle size when the temperature is low, and when the temperature is high, the uni-mode appears at the small particle size. Thus temperature seems to exert a significant influence on the condensation growth of fresh vehicular particles. On the other hand, this ratio correlates positively with ambient particle concentrations ( r 2 = 0.35 ) . Ambient particles measured in this study are mostly > 0.3 μ m while BC in vehicle plumes is generally believed to be mainly in the < 0.3 μ m size range in the literature. Temperature-dependent gas-condensation competition between fresh BC and ambient particles is suggested to play a role in the bi-modal formation.
Keywords: ELPI; Condensation mode; On-road vehicles; Submicron particles; Soot particles; Hong Kong;

Ammonia and ammonium measurements from the southeastern United States by Eric S. Edgerton; Rick D. Saylor; Benjamin E. Hartsell; John J. Jansen; D. Alan Hansen (3339-3351).
Twenty-four-hour integrated gaseous NH 3 and fine particulate ( PM 2.5 ) NH 4 + were measured during 2004 at eight sites in the southeastern U.S. Mean NH 3 concentrations for 2004 ranged from 2.44 ppbv at an urban-industrial site in North Birmingham, AL, to 0.23 ppbv at a rural-forested site near Centreville, AL. NH 3 mixing ratios were found to be higher at urban sites than at nearby rural (or suburban sites) except for sites which were directly influenced by local sources. Only weak correlations with temperature were observed for NH 3 at the sites; slightly greater correlations were observed for total ammonia ( NH x = NH 3 + NH 4 + ) vs. temperature. A weak seasonal variation was observed for NH 3 mixing ratios at all sites, with all but one site exhibiting biannual maxima in spring and late summer/fall. Mean PM 2.5 NH 4 + concentrations ranged from 1.78 μ g m - 3 in Atlanta, GA, to 1.02 μ g m - 3 at Oak Grove, MS, and were more uniform across the network than NH 3 mixing ratios, with only slightly larger values at urban sites as compared to nearby rural (or suburban) sites. All sites exhibited highest NH x between July and September and lowest NH x in November and December. The gaseous NH 3 fraction ( NH 3 / ( NH 3 + NH 4 + ) ) was observed to decrease with increasing values of PM 2.5 SO 4 2 - at all sites. At two rural-forested sites and two sites near the Gulf of Mexico, the NH 3 gaseous fraction approached a relatively constant value of 5–10% as PM 2.5 SO 4 2 - increased beyond 5– 7 μ g m - 3 , suggesting that NH 3 availability at these locations limits aerosol neutralization.
Keywords: Ammonia; Ammonium; Fine particulates; Reduced nitrogen; Aerosol partitioning;

During a coastal environment study an artifact associated with a widely used particulate mercury (Hg) sampling method was discovered that produces inconsistent results for Hg in PM10 samples of different duration (12 versus 4-h). The artifact is attributed to loss of Hg from filters over the course of the longer sampling periods, perhaps due to chemical reaction, enhanced by high humidities common to the marine boundary layer. On the seven days on which three successive 4- and 12-h duration samples were obtained, the mean Hg concentration of PM10 was significantly higher for the 4-h samples (22.5 pg m−3) than for the 12-h samples (8.4 pg m−3) ( p = 0.012 ). The day-to-day differences between 4- and 12-h samples ranged from −1 to 29 pg m−3. No consistent difference was observed in the Hg concentration of PM2.5 of different durations. Adsorption of reactive gaseous mercury (RGM) onto PM2.5 was also evaluated by collection of samples with and without an upstream KCl denuder for RGM removal. Concentrations were not significantly different, indicating that RGM adsorption was not evident in this low aerosol, low RGM environment. The discovery of an artifact related to sample duration demonstrates the need for continued evaluation and validation of the currently accepted methods for atmospheric Hg measurements.
Keywords: Mercury; Sea salt; Aerosol; Annular denuder; Florida;

Time-series observations of the atmospheric concentrations of the halocarbons, trichlorofluorocarbon (CFC-11), dichlorofluorocarbon (CFC-12), 1,2-trichlorofluoroethane (CFC-113), methyl chloroform (CH3CCl3) and carbon tetrachloride (CCl4) were conducted at a site in Lukang, in Central Taiwan between April and August 2004. Fluctuations in atmospheric concentrations of CFC-11, CFC-12 and CH3CCl3 were generally driven by diurnal land–sea breeze and anthropogenic activity in this area. Elevated levels of CFC-11, CFC-12, and CH3CCl3 frequently occurred when the air was stagnant and the prevailing seaward land breeze was dominant. Atmospheric concentrations of CFC-113 and CCl4 were much less variable relative to CFC-11, CFC-12 and CH3CCl3 during the same period, indicating that emissions of these two species from anthropogenic activities were small. The time-series distributions of atmospheric levels of CFC-12, CFC-11, CH3CCl3 and CO were characterized as a diurnal cycle with an elevated level at night and a low level during the daytime for most of the observed periods. As CFC-12, CFC-11 and CH3CCl3 behave as traffic- and industry-derived airborne pollutants in the urban atmosphere, they provide as a useful tracer in the application for the study of terrestrial airborne pollutants transport across the coastal area driven by land–sea breezes in this area.
Keywords: Time-series; Land–sea breezes; Diurnal cycle;

Characteristics of trans,trans-2,4-decadienal and polycyclic aromatic hydrocarbons in exhaust of diesel engine fueled with biodiesel by Hsi-Hsien Yang; Mei-Yu Lo; John Chi-Wei Lan; Jenn-Shye Wang; Dennis P.H. Hsieh (3373-3380).
The use of biodiesel fuel as a substitute for fossil fuel in diesel engines has received increasing attention in recent years. This study is the first to investigate and compare the characteristics of mutagenic species, trans,trans-2,4-decadienal (tt-DDE), and polycyclic aromatic hydrocarbons (PAHs) in the diluted exhaust of diesel engines operated with diesel and biodiesel blend fuels. An engine of current design was operated on a dynamometer consistent with the US federal test procedure transient-cycle specifications. Petroleum diesel and a blend of petroleum diesel and biodiesel (B20) were tested. Exhaust sampling was carried out on diluted exhaust in a dilution tunnel with a constant-volume sampling system. Concentrations of tt-DDE and PAHs were analyzed by GC/MS. Although average PAH emission factors decreased from 1403 to 1051 μg bhp-h−1, the results show that tt-DDE is evidently generated (1.28 μg bhp-h−1) in the exhaust of diesel engine using B20 as fuel. This finding suggests that tt-DDE emission from the use of biodiesel should be taken into account in characterization and health-risk assessment. The results also show that tt-DDE is depleted in the diesel engine combustion process and the existence of tt-DDE in biodiesel is the major source of tt-DDE emission. The distribution of tt-DDE in the particulate phase is 55.3% under this study's sampling conditions. For diesel and B20, PAH phase distributions have similar trends. Lower molecular weight PAHs predominate in gaseous phase for both diesel and B20. Cold-start driving has higher tt-DDE and PAH emission factors, as well as a higher percentage of tt-DDE in particulate phase, than for warm-start driving.
Keywords: Biodiesel; Mutagenic; trans,trans-2,4-decadienal (tt-DDE); Polycyclic aromatic hydrocarbons; Phase distribution;

Long-term personal exposure to PM2.5, soot and NO x in children attending schools located near busy roads, a validation study by Sofie Van Roosbroeck; José Jacobs; Nicole A.H. Janssen; Marieke Oldenwening; Gerard Hoek; Bert Brunekreef (3381-3394).
Several studies have investigated the health of children attending schools located near busy roads. In this study, we have measured personal exposure to traffic-related pollutants in children to validate exposure classification based on school location. Personal exposure to PM2.5, soot, NO x and NO2 was measured during four 48-h periods. The study involved 54 children attending four different schools, two of which were located within 100 m of a major road (one ring road and one freeway) and the other two were located at a background location in the city of Utrecht, The Netherlands. Outdoor monitoring was conducted at all school sites, during the personal measurements. A questionnaire was administered on time activity patterns and indoor sources at home. The outdoor concentration of soot was 74% higher at the freeway school compared to its matched background school. Personal exposure to soot was 30% higher. For NO x the outdoor concentration was 52% higher at the freeway school compared to its background school. The personal concentration of NO x was 37% higher for children attending the freeway school. Differences were smaller and insignificant for PM2.5 and NO2. No elevated personal exposure to air pollutants was found for the children attending the school near the ring road. We conclude that the school's proximity to a freeway can be used as a valid estimate of exposure in epidemiological studies on the effects of the traffic-related air pollutants soot and NO x in children.
Keywords: Traffic-pollutants; Personal exposure; Validation; Children; Epidemiology;

Modelling extreme concentrations from a source in a turbulent flow over a rough wall by Zheng-Tong Xie; Paul Hayden; Alan G. Robins; Peter R. Voke (3395-3406).
The concentration fluctuations in passive plumes from an elevated and a ground-level source in a turbulent boundary layer over a rough wall were studied using large eddy simulation and wind tunnel experiment. The predictions of statistics up to second order moments were thereby validated. In addition, the trend of relative fluctuations far downstream for a ground-level source was estimated using dimensional analysis. The techniques of extreme value theory were then applied to predict extreme concentrations by modelling the upper tail of the probability density function of the concentration time series by the generalised Pareto distribution. Data obtained from both the simulations and experiments were analysed in this manner. The predicted maximum concentration ( Γ 0 ) normalized by the local mean concentration ( C m ) or by the local r.m.s. of concentration fluctuation ( c rms ), was extensively investigated. Values for Γ 0 / C m and Γ 0 / c rms as large as 50 and 20, respectively, were found for the elevated source and 10 and 15, respectively, for the ground-level source.
Keywords: Large-eddy simulation; Atmospheric dispersion; Extreme value theory;

The Interagency Monitoring of Protected Visual Environments (IMPROVE) protocols for reconstructing the ambient light extinction coefficient (b ext) from measured aerosol species are the basis for evaluating compliance under the Regional Haze Rule. Aerosol mass composition and optical properties have been measured as part of the IMPROVE program since 1988, providing a long-term data set of aerosol properties at 38 sites around the US. This data set is used to evaluate assumptions made in calculating reconstructed mass and b ext by applying statistical analysis techniques. In particular, the molecular weight to carbon weight ratio used to compute particulate organic matter is investigated. An annual average value of 1.7±0.2 for the IMPROVE sites, compared to the value of 1.4 currently assumed in the IMPROVE algorithm, is derived. Regression analysis also indicates that fine soil mass concentrations are underestimated by roughly 20% on average. Finally, aerosol mass scattering and extinction efficiencies assumed in the IMPROVE reconstructed b ext protocol are examined. Fine mode (D p<2.5 μm) mass scattering efficiencies have a functional dependence on mass concentrations at many sites, and use of a mass-concentration-dependent adjustment factor to refine the assumed efficiencies provides for closer agreement between measured and reconstructed b ext.
Keywords: Mass scattering efficiencies; Aerosol mass; Organic aerosols; Aerosol measurements;

Development of multiple-species 1 km × 1 km resolution hourly basis emissions inventory for Japan by Akiyoshi Kannari; Yutaka Tonooka; Tsuyoshi Baba; Kentaro Murano (3428-3439).
A new atmospheric emissions inventory, called “EAGrid2000-Japan”, was developed to provide reliable fundamental data for studies of the long-range transport of air pollution in the East Asia region, specifically targeting Japan, in the year 2000. In the development of this inventory, emissions from the Japanese region were estimated in detail with high temporal and spatial resolution: temporally, hourly by month, and spatially, 30 ″ in latitude and 45 ″ in longitude, equivalent to approximately 1 km × 1 km resolution. Anthropogenic annual emissions in Japan were estimated to be as follows: SO 2 , 0.87 Tg; NO x , 2.4 Tg; NMVOC, 2.0 Tg; NH 3 , 0.41 Tg; CO, 5.4 Tg; PM 10 , 0.19 Tg; PM 2.5 , 0.15 Tg; and CO 2 , 1.2 Pg. Using this detailed inventory, multiple nesting techniques can be applied to long-range transport model simulations to distinguish the contribution originating on the Asian continent from the local sources in the Japanese region. Also, the importance of emission variations, specifically diurnal, weekly and seasonal variations, was highlighted based on the estimations.
Keywords: East Asia; Japan; Emission inventory; Multiple species;

Characteristics of puff dispersion in an urban environment by J.C. Doran; K.J. Allwine; J.E. Flaherty; K.L. Clawson; R.G. Carter (3440-3452).
Instantaneous releases of sulfur hexafluoride tracer were carried out as part of the Joint Urban 2003 field campaign in Oklahoma City. Data from 10 fast-response tracer samplers were used to examine the crosswind and along-wind spread of the tracer, the decay of tracer concentrations, and the retention of tracer within approximately 1 km of the release locations. The time variation of the median values of the tracer concentrations, normalized by the peak value observed at a given sampler, could be approximately described by an exponential decay with characteristic decay times on the order of 1–2 min. The longer times were found for early morning releases and the shorter times were associated with later morning or afternoon releases, suggesting that atmospheric stability or the depth of the mixed layer may affect puff dispersion even in urban environments. The median retention times required for 99% of the exposure to be realized at a given location were found to be correlated reasonably well with the median decay times. These characteristic time scales should be regarded as lower limits for concentration decay because the analysis excluded a number of anomalous cases in which the decaying concentrations exhibited an extended tail that indicated a very slow ventilation rate. The median values of the along-wind dispersion parameter σx grouped into downwind distance ranges can be described by a linear variation with distance with an initial “hold up” contribution due to building effects of about 30–45 m, but there are considerable variations about this relationship. Downwind 0.5–1 km from the release point the lateral puff dispersion (σy ) was roughly 70% of the along-wind dispersion.
Keywords: Urban dispersion; Puff dispersion; Decay time; Retention time; Tracer;

The use of wind fields in a land use regression model to predict air pollution concentrations for health exposure studies by M.A. Arain; R. Blair; N. Finkelstein; J.R. Brook; T. Sahsuvaroglu; B. Beckerman; L. Zhang; M. Jerrett (3453-3464).
A methodology is developed to include wind flow effects in land use regression (LUR) models for predicting nitrogen dioxide (NO2) concentrations for health exposure studies. NO2 is widely used in health studies as an indicator of traffic-generated air pollution in urban areas. Incorporation of high-resolution interpolated observed wind direction from a network of 38 weather stations in a LUR model improved NO2 concentration estimates in densely populated, high traffic and industrial/business areas in Toronto-Hamilton urban airshed (THUA) of Ontario, Canada. These small-area variations in air pollution concentrations that are probably more important for health exposure studies may not be detected by sparse continuous air pollution monitoring network or conventional interpolation methods. Observed wind fields were also compared with wind fields generated by Global Environmental Multiscale-High resolution Model Application Project (GEM-HiMAP) to explore the feasibility of using regional weather forecasting model simulated wind fields in LUR models when observed data are either sparse or not available. While GEM-HiMAP predicted wind fields well at large scales, it was unable to resolve wind flow patterns at smaller scales. These results suggest caution and careful evaluation of regional weather forecasting model simulated wind fields before incorporating into human exposure models for health studies. This study has demonstrated that wind fields may be integrated into the land use regression framework. Such integration has a discernable influence on both the overall model prediction and perhaps more importantly for health effects assessment on the relative spatial distribution of traffic pollution throughout the THUA. Methodology developed in this study may be applied in other large urban areas across the world.
Keywords: Nitrogen dioxide; Wind; Air pollution; Urban air quality; Human health; Land use regression models; Toronto; Hamilton;

Evidence for large average concentrations of the nitrate radical (NO3) in Western Europe from the HANSA hydrocarbon database by S.A. Penkett; R.A. Burgess; H. Coe; I. Coll; Ø. Hov; A. Lindskog; N. Schmidbauer; S. Solberg; M. Roemer; T. Thijsse; J. Beck; C.E. Reeves (3465-3478).
The nitrate radical (NO3) was first measured in the atmosphere in the 1970s and suggestions were made that it could play a major role in oxidising many unsaturated hydrocarbons, such as those emitted from the biosphere. Analysis of the hydrocarbon mix over the North Atlantic Ocean suggested subsequently that the influence of NO3 radical chemistry at night was even more extensive, being on a par with hydroxyl radical chemistry at some times of the year.The paper presents a detailed analysis of an extensive database of many nonmethane hydrocarbons collected at various sites around the North Sea in the mid 1990s during the HANSA project. By comparing the relative rates of oxidation of iso and normal pentane with that of toluene and benzene it clearly shows that the efficiency of NO3 radical chemistry and hydroxyl radical chemistry over northwest Europe are similar in springtime and predicts an average nighttime NO3 concentration of the order of 350 pptv, assuming an annual average OH concentration of 0.6×106  cm−3. This value is very dependant on the average emission ratios of the different hydrocarbons and values between 200 and 600 pptv are possible. It is much larger than direct measurements made in Europe at the surface, but is of the same magnitude as concentrations measured recently from aircraft in the boundary layer over the northeast USA, and previously in vertical profiles by remote sounding over Europe.A simple analytical expression can be derived to calculate the NO3 concentration at night with the only variables being ozone and the loss rate of N2O5, either to the ground or to aerosol surfaces. The concentrations of NO3 calculated in this manner are similar to those derived from the analysis of the HANSA hydrocarbon database for typical conditions expected over Europe, but they are very dependant on the efficiency of the aerosol sink for N2O5.It is shown that NO3 oxidation of many unsaturated hydrocarbons can indeed be more efficient than OH oxidation, especially at times of the year outside the summer season. Direct evidence for hydrocarbon oxidation by NO3 radicals is shown by a series of peroxy radical measurements where the nighttime concentrations can be significantly higher than daytime concentrations in polluted air on occasion. Also the winter/summer (W/S) ratios of many unsaturated hydrocarbons are much lower than those expected from their removal purely by hydroxyl radical chemistry.The consequences of these findings are profound especially as satellite measurements of NO2, a major precursor to NO3, suggest that these high average concentrations of several hundred pptv could be widespread over most of the continents. This needs to be confirmed by direct in-situ measurement of nitrate radicals but it suggests a much larger role for NO3 chemistry in the oxidation capacity of the atmosphere than realised hitherto.
Keywords: Nitrate radical; NO3; Hydrocarbons; HANSA database; Peroxy radicals; Nighttime chemistry;

Particle volatility in the vicinity of a freeway with heavy-duty diesel traffic by Subhasis Biswas; Leonidas Ntziachristos; Katharine F. Moore; Constantinos Sioutas (3479-3493).
During February–March 2006, a major field sampling campaign was conducted adjacent to the Interstate 710 (I-710) freeway in Los Angeles, CA. I-710 has high traffic volumes (ca. 11,000 vehicles h−1) and a high percentage (17–18%) of heavy-duty diesel vehicle (HDDV) traffic. The volatility of ambient particles of 20, 40, 80 and 120 nm in diameter was investigated using a Tandem Differential Mobility Analyzer (TDMA) at two locations—close to the freeway (10 m) and approximately 150 m downwind. The smallest particles (20 nm) are largely volatile at both locations. Larger particles, e.g., ⩾40 nm) showed evidence of external mixing, with the non-volatile fraction increasing with particle size. Particle volatility increased with decreasing ambient temperature. The HDDVs contribute to relatively larger non-volatile particle number and volume fractions and greater external mixing than earlier observations at a pure light-duty gasoline vehicle freeway [Kuhn et al., 2005c. Atmospheric Environment 39, 7154–7166]. Finally, the fraction of externally mixed soot particles decreased as the downwind distance increased from the I-710, due to atmospheric processes such as vapor adsorption and condensation as well as particle coagulation.
Keywords: Ultrafine particles; Volatility; Diesel; Elemental carbon; Organic carbon;

Subsides to the creation of a regional model of forest fire hazard: Taquari River Springs Park, MS—A case study by L.M. Mercê de Albuquerque; A.C. Paranhos Filho; T.G. Torres; E. Kassar; H.J.S. de Matos Filho; M.G.G. Carrijo; H.G. Pavão; A. de Souza (3494-3501).
Using map algebra, in the GIS (geographic information system) environment, this study integrates the B-RAMS, Brazilian Regional Atmospheric Models Software (CPTEC, 2005) climate model data with remote sensing data, intending to obtain a wildfire hazard map. The Taquari River Springs Park (TRSP) was chosen as a case study, due to the presence of springs which are considered important contributors to the Upper Paraguay River Basin, and it also contains essential remnants of the Cerrado Biome. The B-RAMS model has provided relative humidity, components of the horizontal wind and temperature. The TRSP land cover was identified by object oriented classification of a LANDSAT ETM+ image, supported by field observations. From the land cover phytophysionomic type characterization, a forest wildfire fuel map has been elaborated. The integration of the different maps has been made using a GIS, and a new map with its associated GIS database was generated showing the most vulnerable zones to wildfire hazard.
Keywords: Forest fuel; Atmospheric modeling; Remote sensing; Map algebra;

This study focuses on applying a Takagi–Sugeno fuzzy system and a nonlinear regression (NLR) model for ozone predictions in six Kentucky metropolitan areas. The fuzzy “c-means” clustering technique coupled with an optimal output predefuzzification approach (least square method) was used to train the Takagi–Sugeno fuzzy system. The fuzzy system was tuned by specifying the number of rules and the fuzziness factor. The NLR models were based in part on a previously reported, trajectory-based hybrid NLR model that has been used for years for forecasting ground-level ozone in Louisville, KY. The NLR models were each composed of an interactive nonlinear term and several linear terms. Using a common meteorological parameter set as input variables, the NLR models and the Takagi–Sugeno fuzzy systems model exhibited equivalent forecasting performance on test data from 2004. For all 2004 ozone season forecasts for the six metropolitan areas, the mean absolute error was 8.1 ppb for the NLR model and 8.0 ppb for the Takagi–Sugeno fuzzy model. When a nonlinear term (which was part of the NLR model) was included in the fuzzy model, the combined NLR–fuzzy model had slightly better performance than the original NLR model. For all 2004 metropolitan area forecasts, the mean absolute error of the NLR–fuzzy model forecasts was 7.7 ppb. These small differences may be statistically significant, but for practical purposes the performance of the fuzzy models was equivalent to that of the NLR models.
Keywords: Ground-level ozone predictions; Fuzzy system models; C-means clustering technique; Nonlinear regression models;