*Paper presented at “South-East Asian Land/Forest Fires: Science and Policy”-Workshop, organised by the Centre for Remote Imaging, Sensing and Processing (CRISP), 30 April2 May 1998, Singapore
German Technical Cooperation (GTZ)
Strengthening the Management Capacities of the Indonesian ForestMinistry (SMCP)
Integrated Forest Fire Management Project (IFFM)
In the second half of 1997, South-East Asian countries witnessed the most extensive smoke haze() disaster in ASEANs history. Exacerbated by drought related to El-Nino, forest fires in Indonesia started spreading significantly at the beginning of July 1997, affecting around 4.5 Mio. hectare (45,000 km2) and 3% of the total Indonesian forest area until the end of December (estimates CRISP). Most affected by the fires were the provinces in the southern-east part of Sumatra and the southern coastal area of Kalimantan. Satellite detection revealed that most of the fires started in areas dedicated to be converted into plantations (palm-oil, pulp-fibre, rubber and other agri-forestry products). Due to the extreme drought, fires escaped into the surrounding vegetation affecting logged forests, peat swamps and grassland [1,2].
The smoke haze emitted from the fires accumulated in the atmosphere up to an altitude of 3000 m and spread over neighbouring countries. At its climax in the third week of September, an area of more than 1 Mio. km2 was covered by thick haze.
Air pollution in Indonesia, caused by particles which is the major pollutant in the haze, reached levels that would correspond to a Pollution Standard Index (PSI)() of above 2000. A PSI of 400 is already categorised as hazardous. The highest pollution levels recorded in Malaysia and Singapore, were equivalent to a PSI of 900 and 226, respectively .
Health officials estimate that in Indonesia alone, 20 million people suffered acute health problems due to the haze . The long-term effects of an exposure to exceedingly high air pollution levels are unknown. In contrast to haze-affected neighbouring countries, few information was given on the regional pollution levels in Indonesia. Due to this fact, an on-time assessment of the ensuing health risks and the dispersion behaviour of the haze was limited. Beside the health impacts as such, the fires and haze in 1997 are estimated to have resulted in over US$ 3 billion in damages resulting from expenses for health treatment, lost tourism revenues, decreased industrial production and losses of commercial timber. Not considered are the global impacts in regard to global warming and losses in biodiversity .
This paper presents the results of a consultancy mission dedicated to post-evaluate the development of the haze in Indonesia 1997 and the ensuing health risks.
Physical and Chemical Properties of Particles
Human-started forest fires, traffic, power plants, industrial and residential combustion processes and industrial fugitive dust are anthropogenic sources of particles. In contrast to natural particle sources like sea spray and soil resuspension by wind, particles from anthropogenic sources are predominately composed of very small, fine particles .
This discrepancy in size distribution is based on the fact that condensation processes of gaseous compounds form the majority of anthropogenic particles. Such condensation processes take place either during the combustion process or subsequently in the atmosphere. The former results in the direct emission of primary particles, the latter in the subsequent formation of secondary particles, which occur either as the formation of new particles or the addition of particulate material to pre-existing particles [6,8]. Main precursors of secondary particles are sulphur dioxide and oxides of nitrogen, which are transformed by oxidation to sulphate and nitrate particles, respectively. Secondary processes are also attributed to a portion of organic aerosols ().
Due to the multiple formation processes, particulate matter is not a chemically defined homogenous substance but differs from site to site with respect to particle size distribution, components, and pattern of exposure.
The particle size is the major determining parameter for particles; most effects associated with particles strongly depend on their particle size, such as atmospheric deposition rate and residence time, light scattering properties and visibility, deposition pattern within the lungs and health impacts.
The size distribution of atmospheric aerosols is shown to have three modes with ascending size as illustrated in Figure 1. The life span of particles in the smallest mode, the nucleation mode, which corresponds to particle diameters() below 0.1 µm, is less than one hour because they readily agglomerate and coagulate and transmute into the next larger mode (accumulation mode). The accumulation mode consists predominately of agglomerated and coagulated particles as well as particles from secondary condensation processes and comprises particles from 0.1 to 1 or 2 µm. Particles in this mode have, with up to weeks, the longest residence time in the atmosphere and can travel long distances (hundreds to thousands of kilometres). Their elimination out of the atmosphere is mainly due to precipitation [6,7]. Natural sources determine the composition of coarse() particles with a diameter of more 2.5 µm. Those particles sediment within several hours up to a day out of the atmosphere .
3-modal size distribution of atmospheric aerosols and size ranges of particles from different sources [6,7,8]
Emissions from Forest Fires
Emissions from forest fires represent a complex mixture of solid, liquid and gaseous compounds. Their composition varies widely dependent on the chemical composition of the biomass burned, the combustion conditions and its efficiency.
Gaseous compounds adjacent to fires include carbon monoxide, sulphur dioxide, methane, oxides of nitrogen and a variety of organic compounds . Solid and liquid compounds, summed under the term particles, are predominantly composed of organic and elemental carbon and are characterised by their very small diameter .
The relation of particulate matter with a diameter smaller than 10 µm (PM10) to the total particulate emission in agricultural burning emissions, is approximately 90% . Ward et al.(1997) found that total particulate matter (TPM) emitted from forest fires with flaming combustion contains 80 to 95% fine particles (PM2.5) and from smouldering combustion up to 90 to nearly 100% .
Organic carbon might contribute to over 90% of the dry mass of particulate emitted from biomass burning, with a maximum in the fine particle fraction. The ratio of organic carbon to elemental carbon is highly variable ranging from 10:1 to 95:1 . The major trace elements are Potassium, Sulphur, Chloride, Aluminium, Silicium, Calcium and Ferrum. Fly ash, containing inorganic residues from the biomass, range from 0.1 to 50 µm and more .
During transport and subsequent dilution from the emission source to the surrounding environment, the smoke components are subjected to various transformation processes. Whereas reactive gaseous compounds are readily decomposed, fine particles due to their longer residence time in the atmosphere have a high potential to be transported over long distances. Consequently they are of major concern regarding air pollution from forest fires in medium and long distances. It is generally expected that the fine particle fraction increase with the distance from the emitter as the larger particles deposit to the surface with a larger velocity. However, coagulation, agglomeration and condensation processes, which take place during the atmospheric transport, lead simultaneous increase of the particle size and might even overcompensate this effect.
Health Impacts from Exposure to Particles
While breathing, particles are retained according to their size within the respiratory system. Larger particles are deposited in the upper respiratory tract, while smaller particles may penetrate deeper into the lungs, where they are retained for a longer period. The part of particles that penetrate and subsequently deposit within the respiratory system varies widely dependent on the properties of the particles, the individual breathing pattern, the structure of the respiratory system and other influencing factors.
Inhalable particles with a diameter over 10 µm are predominately deposed in the nose, the mouth-throat area and in the larynx. The residence time of deposited particles in these areas is several hours. Particles below 10 µm in diameter may advance until the thoracic respiratory system (thoracic particles) and mainly deposit in the trachea-bronchial area, where they are removed within several hours up to a day. Finer particles below 6 and 8 µm penetrate the alveolar area (respirable() particles). The deposition probability might amount up to 60%. The elimination process of particles, which have been deposited in the alveolar area, takes between days and years .
The acute health hazards caused by the interactions between deposited particles and the respiratory system range from acute respiratory symptoms and illness including bronchitis, asthma, pneumonia and upper respiratory infection, impaired lung function, hospitalisation for respiratory and cardiac disease to increases in mortality .
Little is known about the toxicological mechanism behind the health effects caused by particles and even less about the synergism with other pollutants. The organic constituents have been shown to induce some inflammations and suppress the defence capability towards bacteria. Recent findings indicated that very small, ultrafine (<0.1 µm) particles show a greater inflammation potential linked to surface-area dependent toxicity .
It is assumed that several metals and silica-derived constituents of the particles are cytotoxic to lung cells . Acidic particles of less than 0.1 µm are supposed to provoke alveolar inflammation which causes both acute changes in blood coagulability and release of mediators to provoke attacks of acute respiratory illness. The blood changes result in an increase in the exposed populations susceptibility to acute episodes of cardiovascular diseases .
Recent epidemiological studies revealed a close link between the increase of daily particulate concentration and adverse health effects, which are shown to occur even at very low particle levels , and without apparent threshold . From the findings was derived that a 50µg/m3 increase in 24-hour average PM10 causes a 2.5 to 8.5% increase of the total non-accidental mortality in regard to the background level. Fine particles (PM2.5) showed a consistent and statistically significant relationship to acute mortality, with relative risk increases of 2 to 6% per 25 µg/m3 PM2.5 (daily average) . It has to be emphasised that this exposure-effect relationship has been derived from studies at relative low concentration levels. At concentrations amounting a few hundred µg/m3, the slope of the exposure-effect curve has shown to decrease . These correlations do not take into account synergistic effects with other pollutants as well as the multivariable character of particles emanating from different sources.
Whereas acute heath effects of particulate matter is probably best related to the deposited dose, chronic and long-term effects may be related to cumulative or retained dose but may also arise from recurring cycles of pulmonary injury and repair. Retention of particles is a function of deposition site, clearance of macrophages or the mucociliary system and particle characteristics . The retention rate increases significantly at high particle pollution levels (TPM>1000µg/m3), when “overloading” of respiratory system reduces the clearance mechanisms. Lower elimination is also linked with the presence of air pollutants like sulphur dioxides and nicotine, which inhibit the mucociliary efficiency. Accumulation of particles increases the likelihood for chronic obstructive pulmonary diseases, permanent decrease of the lung function, asthmatic symptoms and cardiovascular diseases .
Susceptible groups, represented by children, persons with pre-existing diseases and elderly but also smokers are those who will suffer first from morbidity and mortality from particle exposure. The higher risk results from increased deposition rates of lower respiratory tract in children and persons with existing respiratory diseases (asthma, emphysema, chronic bronchitis) in addition to impaired clearance and general deficiency mechanisms [6,10].
Air quality standards for particulate matter
Responding to the epidemiological findings which revealed that the health risks for the public due to particle exposure, and in particular fine particles, are much more significant than assumed before, the US- Environmental Agency (EPA) amended the National Ambient Air Quality Standards (NAAQS) in 1987. The new NAAQS replaced the previous standard for particles, which was based on Total Particulate Matter (TPM) with new standards for PM10 and PM2.5. Table 1 shows the revised EPA-NAAQS for particles and current standards for other countries .
Air quality standards
(values in m g/m³)
annual average *2
*1 98-percentile value, *2 arithmetic mean
Tab 1 : Current ambient air quality standards in selected countries [6,8,11]
The revision of the air quality standards for Europe, which are still based on TPM, is under process. The proposed standards are 50 µg/m3 PM10 as daily average and 20µg/m3 as the annual . As a tool for easily explaining air quality conditions to the public, EPA developed the Pollution Standards Index (PSI) which is widely adopted by other countries (e.g. Singapore). This system comprises five pollutants including PM10, which are monitored and each translated to an index value. The highest index value is reported as the PSI for the region of measurement. While the PSI ranges from 0 to 500, only values below 100 (standard index) are considered healthful.
In the following table (Table 2), only short-term standards for particulate matter are taken into account.
TSP (m g/m³)
PM10 (m g/m³)
PM2.5 (m g/m³)
Air Quality Description
0 £ 50
0 – 75
0 – 15
51 £ 100
76 – 260
16 – 65
101 £ 200
261 – 375
66 – 150
201 £ 300
376 – 625
360 – 420
151 – 250
Table 2: EPA-PSI-system for TPM, PM10, PM2.5 (24 hour average concentration) [12,13]
Sources of data and collection
In Indonesia, air pollution measurements are conducted on behalf of different ministries in most capital cities of its 27 provinces, independently from each other.
The Meteorological and Geophysical Agency (BMG) measures, beside daily meteorological parameters, Total Particulate Matter (TPM)() every sixth day.
The provincial departments of the Ministry of Health (DepKes) monitor TPM(7) , SO2, NOx, O3 and CO in weekly intervals.
The Ministry of Social Welfare measures TPM, SO2, NOx, O3 and CO in an irregular cycle.
During the haze 1997, the measurement frequencies of above mentioned institutions were intensified. Additional measurement campaigns were initiated on the part of the Environmental Impact Management Agency (BAPEDAL), recording PM10, SO2, NOx, O3 and CO, and private institutions. However, data were not collected in a central database nor exchanged among another and consequently scarcely accessible to the public.
TPM-data of period August to November 1997 were requested for Sumatra and Kalimantan and partly received from above mentioned institutions (BMG, DepKes, BAPEDAL). No detailed background information in respect to the sampling time, the sampling locations and other influencing parameters was provided.
A spot check like survey revealed that the measurement procedures for particulate matter are accompanied by various systematic errors, which influence significantly the accuracy of the measurement results:
The weighing procedure of the unloaded and loaded filter papers from BMG-measurements takes place in the head laboratory in Jakarta. Unloaded and loaded filters are sent by post to and from the provincial BMG-stations, respectively. While transport, which might last up to 3 weeks, the filters are subjected to biological, chemical and physical transformation processes. These processes are suspected to decrease the total amount of particle measured.
The measurement timing are not practised in a regular rhythm. There are significant variations of start- and stop-time as well as the duration of the measurement from 2 to 24 hours. Considering the variability of the particle concentration during a day, the values reported might not represent the real daily average concentrations.
At high pollution levels during the haze, the filter resistance increased significantly within a few hours due to the high load of particulate matter deposited on the surface. Drops in volume flow rates up to 50% were observed at BMG-measurements. The total volume was calculated by the average of the flow rate at the beginning and the end of the measurement. As the flow rate generally decreases exponentially, the real total volume is supposed to be higher and subsequently the calculated particulate concentration lower than the real value.
The measurement locations of the provincial Ministry of Health rotate frequently with the intention of covering differently influenced air pollution levels (industrial, traffic, and residential areas). This results in a high variability of the measurement data.
The BMG-data obtained were on average 50 to 70% lower than DepKes-data of the same day and location. For most locations no daily data pair was available. For further processing, TPM-data for each location were merged by calculating their daily mean.
The above-mentioned restrictions, including unconsidered influencing factors and the fragmental character of the data obtained, limit their comparability and interpretability. However, their reflect approximately the range of concentrations occurred during the haze in Indonesia 1997.
Haze development in Indonesia 1997
Development of Forest Fires
With beginning of July 1997, the density of forest fires in Indonesia started increasing significantly but with high daily variations. In Sumatra, between 100 to 200 hotspots were detected in the provinces Riau and Jambi in July and in South-Sumatra in August, respectively. During September and October, the fires affected all 3 provinces with more than 100 hotspots per month and more than 300 hotspots in Jambi and 600 in Southern Sumatra until they were extinguished by incoming rain in the second half of November. The most fire-affected provinces in Kalimantan were West- and Central-Kalimantan where dense fire locations (up to 650 hot-spots) were observed from August to November [1,16]. Clusters of peat and peat forest fires which are supposed to have contributed a high percentage of the total forest fire emissions, were observed during September to November of 1997 in the coastal areas of South-Sumatra, Jambi, West- and Central-Kalimantan .
Development of particle pollution levels from August to November 1997
From August onwards, significant haze started accumulating in the lower atmosphere near the main fire locations, southern-east Sumatra and southern West- and Central-Kalimantan. Spreading and intensifying fire locations in the following months contributed to a further atmospheric enrichment, which was accelerated by the induced formation of inversion layers. Predominating south-east wind directions transported the haze to the north-west and caused subsequently high pollution levels in fire remote locations.
a) Development in Sumatra
In Sumatra, highest TPM-values with a maximum of 3940 µg/m3 on the 29th September were measured in Jambi, southern-easterly Sumatra. TPM-concentration in this location started increasing from below 500 µg/m3 in August via 1200 µg/m3 on the 9th to 11th September to 2600 µg/m3 on the 19th of September. After a decrease in the subsequent 5 days to 1600 µg/m3, the TPM-concentration rose continuously until its maximum with 3940 µg/m3 on the 29th September. The following day, the TPM-concentration fell down sharply to 1950 µg/m3, ranged between 1800 and 2700 µg/m3 the next 5 days and shows another maximum on 6th October with 3500 µg/m3. Until the 10th of October, particle concentration went down to 200 µg/m3, retook after a gap due to missing values the 29th October with 1600 µg/m3 (PM10) and finally went back with fluctuations to its background level from the 14th of November onwards. Padang, west north from Jambi, recorded the second highest values with a maximum peak of 2400 µg/m3 TPM on the 1st October, after being subjected to significant fluctuations the precedent 5 days. Jambi was followed by Pekanbaru, north-west from Jambi, with a maximum value of 2000 µg/m3 on the 23rd and 27th September. In both locations, the concentration was below 500 µg/m3 until the 6th of October and from this day onwards, with the exception of an intermediate rise of the concentration in Pekanbaru to 600-700 µg/m3 from the 21st to 26th October. The particle level in Medan varied between 100 and 400 µg/m3 during the time from August to November with a one-day peak of 800 µg/m3 on the 1st of October. TPM-concentration in Aceh, northern Sumatra, for which values are only available until the 27th of September, did not reveal any changes.
b) Development in Kalimantan
High particle levels in Kalimantan occurred in Palangkaraya (Central-Kalimantan) during 28th of September and 22ndof October with three major peaks from concentrations of 3200 to 4050 µg/m3 TPM. The first peak occurred from 28th September to 3rd October with a maximum of 3300 µg/m3 and fluctuations in the precedent and subsequent days from 2300 to 900 µg/m3. After an intermediate minimum of 100 µg/m3 on the 7th October, the course of the particle concentration rises sharply to its second peak the 10th of October with 4050 µg/m3 TPM and decreases again within the next two days to 400 µg/m3. After a third, highly variable peak from the 16th to 22nd October with values between 1600 and 3300 µg/m3, the TPM-concentration finally decreased to its background level below 500 µg/m3 from the 2nd November onwards. Banjarmasin (South-Kalimantan) showed a similar particle development to Palangkaraya with three- although significantly lower- major peaks. The first peak on the 2nd of October which reached 1300 µg/m3 was followed by the second peak the 10th of October with 1900 µg/m3 and a third peak of 1100µg/m3 on the 22nd of October. In the mean time between the peaks, the particle level varied around 500 µg/m3. The rise of the particle concentration in Pontianak began around 4 to 6 days earlier than in the above mentioned locations and directly lead to its maximum on the 24th of September with 1900 µg/m3. In the subsequent time it decreased gradually below 500 µg/m3, except an intermediate increase to 700 µg/m3 on the 11th October and 1000 µg/m3 the 23rd October, respectively. The particle level in Samarinda, East-Kalimantan, increased only slightly and did not exceed 500 µg/m3.
As a consequence of the increasing number of forest fires, a sudden strong rise of the particle concentration in the ambient air from below or around 500 µg/m3 to far over 1000 µg/m3 occured in Sumatra and Kalimantan from the second and third week of September 1997 onwards. The concentration finally went down to the normal background level with delayed onset of the monsoonrain in November, which not only reduced gradually the number of fire locations but also efficiently removed the particle load in the atmosphere via rain-out and wet deposition mechanism.
High pollution levels in Kalimantan with values above 1000 µg/m3 began and ended around two weeks later and earlier, respectively, than in Sumatra. Considering only the measurement data available, it can be seen that high pollution levels in Kalimantan concentrated in the period between the 24th September to 24th October, whereas in Sumatra they scattered more and distributed over a longer period but with shorter duration.
The fact that the course of the concentrations was subjected to high daily variations reflects the strong interdependency from fluctuations of fire locations, wind conditions as well as the occurrence of intermediate rain. The latter can be attributed to the strong decline of the particle concentration in the first week of October. These high daily variations make it very difficult to estimate and interpolate the development when measurement data are missing.
Jambi in southern-easterly Sumatra and Palangkaraya in southern Central-Kalimantan with particle pollution levels up to 4000µg/m3 were by far the most affected from particle pollution of the forest fires in 1997. Both locations are situated 100 to 300 kilometres north-westwards from main forest fire locations and in particular emission-intense peat forest fire clusters at the coastal area. Predominating south-east wind-directions in Sumatra and Kalimantan caused that both locations were not only affected by particle emissions from fires in the closer surrounding, but additionally experienced significant particle contributions via long range transport from more distant fire locations in eastern South-Sumatra and South-Kalimantan, respectively. The coincident increase of the particle pollution in Banjarmasin and Palangkaraya supports this assumption. Long-range transport of particles can also be attributed to a significant part of the particle pollution levels in Pontianak, West-Kalimantan, and particularly in Padang, West-Sumatra, where only few fire locations were detected in the next surrounding. Because of a gradually increasing influence of west-wind towards the northern part of Sumatra, Medan was only slightly and Aceh is not affected by wind-born haze.
Figure 5 represents the frequency of particle pollution levels at the measurement locations in Sumatra and Kalimantan for the period from August to November 1997. It has to be emphasised that for a large number of days, no measurement data are available.
From only 55 days covered by TPM-measurement data in Jambi of a period of total 91 days (August-November), 35 days (64%) were above 626 µg/m3 and consequently categorised as hazardous (PSI-threshold-system, Table 2). Among those 35 days, 14 measurement data ranged between 1000 and 2000 µg/m3 and 10 exceeded 2000 µg/m3. In Palangkaraya, 20 days out of 34 measurement data (59%) can be categorised as hazardous, of which 8 days ranged between 1000 to 2000 µg/m3 and also 8 days above 2000 µg/m3. Pekanbaru and Pontainak experienced each10 times “hazardous days” within 42 and 66 measurement data given. Taking into account the number of missing data and the course of the particle development in Figure 3 and 4, it can be assumed that the total number of days with hazardous pollution levels in Jambi and Palangkaraya was more to 20 days higher.
No consistent measurement data are available to determine the particle size distribution of the haze in Indonesia. However, it can be assumed that TPM values around 400 µg/m3 contain at least 80 % PM10 and more than 60% PM2.5(). At higher pollution levels (more than 1500 µg/m3), the PM10 and PM2.5 relation to TPM might even amount up to 95% and 90 %, respectively, because of the decreasing influence of coarser urban aerosols. These relations are much higher than those used for the PSI-threshold-system that bases mainly on the size distribution of urban aerosols (Table 2). According to this system, a PSI of 301 (lowest range of “hazardous”) equals 626 µg/m3 TSP, 430 µg/m3 PM10 and 251 µg/m3 PM2.5, respectively (relation of PM10 and PM2.5 to TPM is 69% and 40%).
The same TSP-concentration from haze corresponds to 501 µg/m3 PM10 and 376 µg/m3 PM2.5, if assuming a relation of 80% PM10 and 60% PM2.5 of TPM. A PM10-concentration of 501 µg/m3, however, is equal to PSI 396 and 376 µg/m3 PM2.5 equals PSI 418. Both represent much higher pollution levels (PSI) than the correspondent TPM-concentration. Hazardous pollution levels in the haze might already been reached at 400 µg/m3 TPM, mainly to exceeding PM2.5 concentrations. From this follows that the number of days categorised as hazardous in Figure 5 will have been even significantly higher.
These figures reveal the dimension to which the population in affected areas was exposed during the haze 1997. Health statistics registered a considerable increase of upper respiratory infection, asthma, bronchitis and pneumonia as well as eye and skin irritation. Beside the physiological effects, also depression and anxiety syndromes occurred more frequently. The inconsistency of the data of the air pollution measurements as well as the health statistics does not allow to derive a correlation between the particle level and health effects. The increase of the daily mortality during haze can only roughly be assessed by existing correlations, which give a 2.5-8.5% rise of the daily total non-accidental mortality per 50µg/m3 increase of daily PM10 and 2-6% per 25 µg/m3 PM2.5 . At PM10 levels of 501 µg/m3 (626 µg/m3 TPM) only, daily non-accidental-mortality rate will have increased 25 to 68% and even 2988% regarding the correspondent PM2.5 (336 µg/m3). Since the linearity of the exposure-response correlation has not been proven for higher pollution levels, estimates for higher particle concentrations are not possible.
The persistence of exceeding high particle levels leads to an overload of deposited particles within the respiratory system, which is most likely to induce chronic, long-term respiratory diseases.
Conclusion and Recommendations
It is evident that the only means to prevent reoccurrence of haze in future, and thus the endangerment of millions of human beings health, is to prevent a new outbreak of forest fires in this magnitude. Although the experiences of the detrimental effects of the fires and haze of 1997, a new outbreak of the fires in the first half of 1998 in Kalimantan could not be prevented, where another 3 Mio. hectare() of forest were affected and again, the population was exposed to high air pollution levels. This fact reveals that the Indonesian government still has to make huge efforts to prevent and control forest fires. Present land use and forest management policies still favour the use of fire. Immense land conversion programs will cause future large scale forest fires if no fundamental changes of the underlying political and social framework will take place. It can be hoped that these changes will be included within the recent political reformation process.
In contrast to haze affected neighbouring countries, almost no information was given to the public concerning the level of air pollution and the ensuing health effects. For medium term of view, the implementation of an air pollution-monitoring network will be needed which coordinates the considerable number of already present measurement activities in Indonesia, including standardisation and improvement of the existing measurement procedures.
Continuous distribution of update information on the air pollution and the ensuing health effects to the public will not only increase public awareness towards environmental issues and the timely initiation of necessary preventive measures. It might also contribute to a broader participation in fire prevention activities and represents a further step towards more transparency.
Due to the specific, fine particle dominated characteristics of the haze, air pollution standards exclusively based on PM10 and in particular TPM can only restrictedly represent the real air quality level and might give a false sense of security. Beside PM10 or TPM, a simultaneous determination of the PM2.5-concentration is necessary to reveal the existing air pollution level.
 Ministry of Forestry (Indonesia): Report on the Eighth Meeting of the Consultative Group on Indonesian Forestry. Special Session on Forest Fire; December 1997, Jakarta
 WHO Indonesia: Haze Disaster Assignment; Jakarta, November 1997
 Ministry of Health Indonesia: Haze Disaster and Health impact in Indonesia. Paper for the Biregional Workshop on the Health Impacts of haze-related air pollution, IMR , Kuala Lumpur, Malaysia, 1-4 June 1998
 European Commission, Technical Working Group on Particles: Ambient air pollution by particulate matter, Draft Position Paper, 1997/8
 Verein Deutscher Ingenieure (VDI): VDI-Richtlinien, Messen von Partikeln, Gravimetrische Bestimmung der Massenkonzentration von Partikeln in der Aussenluft, Grundlagen. VDI 2463 Blatt 1 Entwurf 1997, Berlin
 Israel,G.: Skript zur Vorlesung Luftreinhaltung, Fachgebiet Luftreinhaltung, Technical University Berlin, 1988
 United States Department of Agriculture (USDA): Health Hazards of Smoke. Recommendations of the Consensus Conference April 1997; November 1997
 Costa, L.C.; Amdur, M.: Air Pollution, Environmental Toxicology, 1994
 BAPEDAL/JICA: The study on the integrated air quality management for Jakarta metropolitan area, Draft Final Report, March 1997, Jakarta
 European Fire Response Group: Minutes of Meeting No.8, Nov.1997, Jakarta
 WHO: Assessment of health implications of haze in Malaysia, Mission Report Revision 1, November 1997, Malaysia
Sincere thanks are expressed for those agencies and persons, which supported the authors work in providing data and valuable inputs, in particular the Meteorological and Geophysical Agency (BMG), the Environmental Monitoring Centre (EMC), the provincial departments of the Ministry of Health in Pontianak and Samarinda and the Environmental Impact Assessment Agency (BAPEDAL) and the WHO representative in Indonesia.