In de interimrapportage wordt de huidige stand van zaken beschreven van, hoofdzakelijk, (inter)nationaal epidemiologisch onderzoek naar de relatie tussen PM10 en het optreden van gezondheidseffecten. PM10 ("particulate matter") is de afkorting van de massa van de in de lucht zwevende deeltjes (aerosol) met een diameter kleiner dan ongeveer 10 micrometer (10 mum = 0,01 mm). PM10 wordt in Nederland ook wel "fijn stof" genoemd en geldt als indicator voor de ernst van het wintersmogmengsel. Daarna wordt verslag gedaan van het onderzoek dat tot nu toe is afgerond. Uit de literatuur blijkt dat blootstelling aan wintersmog geassocieerd is met een toename van de dagelijkse sterfte, met ziekenhuisopnamen voor ademhalingsklachten, met een toename van medicijngebruik bij astmatici en met longfunctieverslechtering. Er lijkt geen drempelwaarde voor dergelijke associaties te bestaan waar beneden effecten op mortaliteit en morbiditeit niet meer voorkomen. Gelet op de moeilijkheid een drempelwaarde vast te stellen, wordt een schatting van de gezondheidseffecten uitgedrukt als procentuele toename per 100 mug/m3 verandering in de daggemiddelde concentratie van PM10. Deze toename is voor dagelijkse sterfte 10-15%, voor toename van ziekenhuisopnamen voor respiratoire aandoeningen 20-40% en voor toename van medicijngebruik bij astmatici ongeveer 30%. De longfunctieverslechtering wordt geschat op 2-4% voor 100 mug/m3 toename van de PM10 concentratie. De daggemiddelde PM10-concentraties in Nederland kunnen tijdens wintersmogepisoden oplopen tot 140 mug/m3 en hoger. Er zijn weinig toxicologische gegevens beschikbaar omtrent de gezondheidseffecten en het werkingsmechanisme van deeltjesvormige luchtverontreiniging. Enkele proefdierstudies met gecontroleerde inhalatie van verschillende (redelijk oplosbare) deeltjes suggereren dat de grootte van het oppervlak en de reactiviteit van het oppervlak van ultrafijne deeltjes (< 0,1 mum) een rol spelen in het veroorzaken van pulmonale effecten. Het grootste deel van de massa van alle deeltjes wordt veroorzaakt door deeltjes met een diameter van enige micrometers (mum) tot enige tientallen micrometers. Omdat de hoogte van de PM10-niveaus afhangt van de massa van de deeltjes worden deze PM10-niveaus vooral bepaald door deeltjes groter dan enkele micrometers in diameter. De ultrafijne deeltjes met diameters van enige honderdsten tot tienden van micrometers vormen wel de grootste aantallen deeltjes, maar dragen vrijwel niet bij aan de PM10-niveaus. Deeltjes van tussenliggende diameters bepalen voor het grootste deel het oppervlak. Zij hebben een verwaarloosbaar effect op de PM10-niveaus. Het is momenteel nog onbekend of het toxische werkingsmechanisme van de deeltjes afhangt van de massa dan wel de chemische samenstelling van de deeltjes, of afhangt van het oppervlak van de deeltjes of juist van de aantallen deeltjes. Daarom is de keuze van het juiste bestrijdingsdoel op dit moment niet eenvoudig.

This report presents a halfway score of the research project "Winter smog and Traffic", one of the themes of the research programme "Air Pollution and Health". A state of the art is presented of the health effects associated with exposure to winter smog and of the toxicological effects caused by the inhalation of particles. A summary of the assessment of air quality and the results of epidemiological research is presented. Some policy questions are answered as far as possible at this stage of the project. Finally, an outline of a plan of activities is presented which is based on the policy needs and the most serious gaps in knowledge.

Current concentrations of air pollution during episodes of winter-type smog are now lower than in the late eighties. Traffic is a major contributor to the particulate and gaseous air pollution. There are indications that hospital admissions, mortality and other less severe health effects are associated with particulate and gaseous air pollution during episodes of winter-type smog. As an example an estimate has been made of the expected effects on daily health after exposure to extreme 24 h concentration of 230 microg/m3 PM10. The health impact varied for different endpoints, of which high and low values will be presented. During such an extreme episode acute mortality is expected to rise by approximately 20%, while use of medication by asthmatic children is expected to rise threefold. Other health endpoints showed increases between these two. Current knowledge indicates that (even a modest) decrease in yearly average concentrations has a greater positive effect on health effects than stopping all traffic emissions in the Netherlands during such an extreme episode. Therefore a more permanent policy of curbing (traffic) air pollution is probably more effective in reducing public health risks than a one-off termination of all urban traffic in the Netherlands during a winter-type smog episode.

This report discusses the reliability of environmental indicators obtained from calculations, used by RIVM in the composition of reports on the state of the environment. The considered calculation chain (for the Netherlands) starts at the emission of nitrogen compounds in 1992 and ends with forecasts of the nitrate concentration in untreated water in 2100. To this end the notion of reliability is specified in terms of a reliability factor and a confidence interval. The structure of the chain is depicted with flow diagrams. It is composed of measurement data, registration data, model operations and GIS operations. The primary data with respect to the nitrogen load are obtained from sources outside RIVM. They are based on registrated data, for example agricultural surveys. For the nitrogen emissions a reliability was postulated. The applied RIVM models are respectively OPS (atmospheric transport), DEADM (atmospheric deposition), NLOAD (nitrate leaching from agricultural soils), BOSMODEL (nitrate leaching from forest soils and heathland) and FLOPZ1 (groundwater streamlines and residence times). The measurement data are obtained from the National Monitoring Networks for Air Quality (LML, wind velocities and NOy concentrations) and Groundwater Quality (LMG, nitrate concentration in deep groundwater). Applying uncertainty analyses on models and using statistical techniques, uncertainties in the various indicators are determined, where uncertainties are propagated through the chain. The reliability of the indicators, as reported for 1992, is calculated on the basis of the resulting uncertainties. Most of the calculations are performed on three different scales, 5 x 5 km blocks, Acidification Areas (20 areas in the Netherlands) and on national scale. The reliability of nitrate concentration in untreated water is assessed for two different drinking water stations. Results of the project are presented in the form of tables and graphics, such as histograms and geographical plots of indicators and reliability factors. Table 7.1 on page 73 summarizes the reliability of the investigated indicators. A subsequent goal of this report is to contribute to a (discussion about a) generally applicable method to determine the reliability in the calculation of environmental indicators.

This report presents tabulated National Air Quality Monitoring Network results for the city and street stations in all regions in the Netherlands for the calendar year 1994, summer 1994, winter 1993-1994 and 1 April 1993 - 31 March 1994 (tropical year: EC reference period). The components monitored were: fine dust (PM10), CO, Ox (=NO2+O3), O3, NO2, NO, NOx (=NO2+NO), black smoke (= suspended matter measured by the black-smoke method) and SO2. The fine dust (PM10) measurements were corrected for a systematic underestimation caused by the sampler. In a check made at all stations in the network for exceedance of Dutch limit values in 1994, exceedances for black smoke, PM10 and O3 were observed. The limit values for SO2, NO2 and CO were not exceeded.

Measurements to determine the chemical composition of precipitation in the Netherlands were carried out in 1994. Measurements were taken from 4-weekly samples obtained from the National Precipitation Chemistry Monitoring Network (LMRe). Samples from 15 stations were analysed for main components and inorganic microcomponents (heavy metals). Additionally, different samples were taken at two stations for analysis of the component mercury and the pesticide lindane (gamma-HCH) and at one station for analysis of a supplementary set of pesticides. This set contains 2,4-D, 2,4,5-T, atrazine, bentazone, cyanazine, desethylatrazine, desisopropylatrazine, dichlorprop, MCPA, mecoprop, metolachlor and simazine. Analysis of the main component samples determined the concentrations of hydrogen ions (free acid), sodium, potassium, calcium, magnesium, fluoride, chloride, nitrate, sulphate, bicarbonate and phosphate, along with conductivity and pH value. The samples for heavy metals were analysed for cadmium, copper, iron, manganese, lead and zinc. Arsenic, chrome, nickel and vanadium were also determined in samples from two stations. Information on the LMRe and the calculation procedures precede the results presented.

Volatile Organic Compounds (VOC) have been measured in the Dutch National Air Quality Monitoring Network since 1993. An evaluation of the measurement strategy has been executed, which showed that the network configuration and the time resolution of measurements was found no longer fullfilling the information requirements with respect to policy support and development. The basic principles for the redefinition of the measurement strategy for VOC were: a) meeting the measurement obligation for benzene and corresponding requirements (the OPS and the CAR model being the most important) and b) supporting the 'Carbohydrates-2000' project. For the measurement configuration , three regional sites have been chosen according to the principle 'north-middle-south'. The definition allows for influences of sources in Germany and Belgium, and in the Rijnmond area, and influences from large-scale background levels. Four street stations along with one city background station, will be used to describe air quality in cities. One supplementary site has been defined within the Rijnmond area. A spatially coherent set of stations (i.e. 'regional-city background-street') has been selected within this configuration. The time resolution of the sampling will be one day (for the spatially coherent set of stations and the site in the Rijnmond area) or one week (other stations). The OPS model which calculates monthly to yearly average VOC concentrations on a national grid with a spatial resolution of 5x5 km, will require for validation purposes, results from regional, and to a limited degree, city stations. The CAR model, which calculates yearly average concentrations of traffic-related components, will use results from city background and street stations. The 'Carbodydartes-2000' project will be supported by the measurement site in the Rijnmond area and, to a lesser degree, by the regional sites in the middle and northern part of the country.

Bottom-up data for carbon dioxide, methane and nitrous oxide from the official national inventories (National Communications) were compared with data from EDGAR (Emission Database for Global Atmospheric Research) and top-down emission estimates, based on the results of dispersion and climate models using measured concentrations of greenhouse gases in the atmosphere. The aims of this preliminary study were to investigate the possibilities of comparing different types of emission inventories, to develop a methodology for this comparison and to use the results in an analysis to identify areas for improving the IPCC methodology. The main conclusion points to the value of an international program to review and evaluate national inventories of greenhouse gases. Both bottom-up and top-down emission data were found to improve the scientific understanding of the global and regional budgets. Data exchange promotes consensus on the data among scientists and policy-makers. For CO2 , it is not possible at the moment to estimate fossil fuel emissions derived from atmospheric measurements and global carbon dioxide budget calculations with more accuracy than for emissions based on bottom-up emission inventories. For CH4, a global or even a zonal comparison of bottom-up emission inventories with top-down results of transport models is possible; furthermore, uncertainty for specific sources may have been reduced. Uncertainties for N2O emissions are so large that we may expect both top-down and bottom-up emission estimates to benefit from results of a careful comparison of these emission estimates. Comparing national inventories with EDGAR data has identified areas for future improvement in the IPCC Guidelines.

This manual describes the use of the CAR-AMvB program (version 2.0). This program can be used for the calculations of air quality alongside roads in cities in the Netherlands, for the execution of the Air Pollution Act for carbon monoxide and lead, nitrogen dioxide and benzene. The RIVM provide this program to municipals. These municipals have to report their exceedance of concentrations to the provincial authorities. This manual of the CAR-AMvB program could be used for the obligation reports of 1997 and the probably reports in 1998, 1999 and 2000. This version could be used for future calculations in 2010 and 2020. The emission factors and background concentrations are described in the Global Competition scenario of the fourth national environmental survey.

The CAR-Parking computer program, for which this manual has been written, was developed for calculating the air quality in the vicinity of parking garages and for testing the concentration of benzene against the standards for this compound. The model calculates the emission of benzene from cars inside the parking garage and the contribution to the concentration of benzene in the vicinity of the parking garage. The input data apply to traffic (e.g. number of cars per day, average velocity, length of time car is parked), parking garages (dimensions, number of levels, distance between levels), ventilation (open parts of the outer wall, capacity of mechanized ventilation) and configuration (presence of other buildings). Outer-wall concentrations are calculated for user-defined distances. The CAR-Parking program is a Windows application. The manual comes complete with two disks containing installation software.