Note by the secretariat
1. The meeting of the Task Force on Physical Environmental Accounting was convened in Geneva on 24 - 25 November 1994. It was a concluding meeting that brought to the end the work that was commenced in 1991.
2. The meeting was held in two phases. Separate meetings by the two Pilot Groups - one on land use/land cover, the other on nutrients and the environment - were held on the first day. The purpose of these meetings was to discuss substantive and organisational matters with regard to contributions by member countries to the overall Report of the Task Force, also available for the 1995 CES plenary session.
3. The second day was devoted to the Plenary meeting of the Task Force. The main preoccupation of this meeting was overall conclusions that could be drawn from the exercise, in particular in view of possible continuation of the project. The results of this consideration are summarized in the annex to this note (English only), which also provides background information on objectives and history of the project as well as on methodologies used in the pilot studies.
4. The Task Force, in the light of the results of the work undertaken by the two Pilot Groups, and in particular as a result of the overall conclusions drawn from this work, recommends to the Conference of European Statisticians the inclusion of the development of a comprehensive conceptual framework for physical environmental accounting in its programme of work.
5. The Task Force adds the following remarks of a systematic or organizational nature to the above recommendation:
(a) The work to be undertaken should be carried out as a collective effort of experts from ECE member countries, to be coordinated and serviced by the ECE secretariat. The selection of experts should include staff of national statistical offices, as well as of subject-matter administrations and institutions. In preparation of this selection, it is suggested to members of the Conference that they circulate the report of the Task Force to relevant administrations and institutions in their respective countries.
(b) As a first step, the scope of physical environmental accounting should be determined for the purposes of this project. The most adequate method of work to be applied to this task would be a consultation by correspondence. The results should be submitted for approval by the Conference of European Statisticians in 1996.
(c) In order to accelerate the conceptual work, the consultation by correspondence should seek commitments by ECE member countries to participate actively in the future development of physical environmental accounting. These commitments should permit the constitution of one or two expert groups during the year 1995/96. These groups should start the substantive development of the accounts in their fields of competence.
(d) It could be envisaged to include broad areas of environmental concerns in the scope of the accounting system, for which the chances of applying accounting techniques are relatively large. Examples for such sub-areas of a future system are: natural/environmental resources (water, soils, forest, underground reserves, land use/land cover); materials flows and balances (materials to be selected from the points of view of relevant policy concerns); impacts of sectoral economic activities; environmental risks and damages; etc. These sub-areas could be periodically revised.
(e) The organization of substantive work should, to the extent possible, make use of interest and willingness of all international institutions ready to participate in the task.
(f) The ECE secretariat is accountable for the adequate coordination of the work with related projects pursued at the international level. In particular, the work done by UNSTAT on the SEEA and the work envisaged here are complementary. The coordination of the two activities would therefore benefit both the work of the Conference and the further development of the SEEA.
1. The subject of environmental accounting had been among the first developmental issues, when international work on environment statistics began to preoccupy statistical offices. During the mid-seventies, the UN Statistical Commission had mandated pioneering work on materials and energy balances, and the results of this work are available in UN document E/CN.3/492, submitted to the 1976 session of the UN Statistical Commission.
2. However, the outcome of this work did not find generalized application in the field of environment statistics. Its main impacts were in the area of waste statistics and of materials balances for selected chemical substances. In the first of these two domains, a 'materials approach' to the description of waste flows has gained acceptance over the years. In the second, a limited number of countries has produced materials balances for some materials of particular environmental concern to them.
3. The recent resumption of work on environmental accounting was not so much triggered by a wish to explore systematically a new field for statistics, but more by the need to find tools that would facilitate the integration of environmental with other - mainly economic - concerns in social analysis and decision-making. As (national) accounting has become the central piece of statistical information for economic policy, accounting techniques developed quite naturally into the centre of this renewed interest.
4. The Conference of European Statisticians (CES) reviewed the issue as a whole as well as its possible contribution to it at its 1991 plenary session. The discussion was based on document CES/700. The results of the discussion are summarized in paras. 66-70 of the report of the plenary session, document ECE/CES/38. The discussion prompted the Conference to create the Task Force on Environmental Accounting. The intentions that had governed this decision can be summarized as follows. While accounting techniques have clearly stood their test as a powerful instrument in the statistical production technique, their usefulness in relation to environmental issues remained hypothetical. The main explanatory circumstance for this appreciation of the state of the art was the relatively undefined work area of environment statistics, together with the bewildering multitude of variables as well as units of measure that characterize it. In the light of the corresponding doubts, the Conference hesitated to plunge into a major developmental effort. However, it was clear that some exploratory further study was required.
5. In order not to overload the authors of such exploratory work, it was felt that accounting techniques should not be tested for the totality of environmental issues, but only for a selection of them. The objective for the exploratory work then was to test the feasibility of environmental accounting in relation to a small number of environmental phenomena. A report on the experience gained from this exploratory work was intended to guide the Conference at its 1995 plenary session, at which it was envisaged to take a decision on whether or not a project should be launched, aiming at a comprehensive conceptual framework for environmental accounting. Already during the exploratory phase, coordination was to be maintained with accounting work done in other fora of international statistical cooperation. This was notably the case with regard to the work done at the UN Statistical Division.
6. 1A meeting among countries interested in participating in the Task Force was convened in Geneva in December 1991. The conclusions reached at this consultation are the subject of document CES/717. The meeting proposed to conduct two pilot studies by groups of volunteering countries, each coordinated by a lead country. This proposal ensured minimum secretariat resources to be allocated to the project.
7. The pilot issues selected were "land use/land cover" and "nutrients and the environment". In order to avoid unnecessary duplication of international work, the Task Force was expected to focus on physical accounting, i.e. accounting in terms of physical as opposed to monetary units, this latter work being entrusted to the UN Statistical Division. After approval of these plans by the CES Bureau, a letter was sent in March 1992 by the ECE secretariat to member countries, asking for volunteers to participate in the Task Force.
8. A first meeting of the volunteering countries was held in Geneva on 19-20 October 1992. The meeting agreed on the basic approach to be followed in work on the two pilot issues, as well as on the basic organizational arrangements. The conclusions reached in these respects are set out in document CES/754. They include in particular the statement that the report from the Task Force "should permit the Conference of European Statisticians, at its 1995 plenary session, to come to a conclusion as to the continuation of work on this form of environmental accounting" (para. 7 of the Annex to document CES/754). The Conference approved these conclusions and included a corresponding project in its programme of work (see para. 36 of the report from the 1992 plenary session, document ECE/CES/40, and project 10.5.2(viii) in Annex II to the same document).
9. The following countries participated in the two pilot groups:
Pilot Group on changes in land use/land cover:
Austria, France (lead country), Germany, Poland, United Kingdom
Pilot Group on nutrients and the environment:
Finland, Hungary, Italy (lead country), Netherlands, Norway, Sweden
10. The participants agreed to pursue a joint effort in the work envisaged. However, it was clear that, owing to the nature of this developmental effort, resource and data availability would determine the extent to which any given country can at this stage compile data accounts in accordance with the accounting structure agreed in each group.
11. The Pilot Groups met each on several occasions. The annual 1993 meeting was held on 67 December. The conclusions from this meeting were brought to the attention of the Conference in document CES/803. The final meeting of the Task Force was held on 24 and 25 November 1994. It agreed on the remaining issues to be solved for the purpose of the final report from the Task Force. In particular, it drafted the overall conclusions as well as the recommendations to the Conference as they are included in this report.
12. Land is a key factor in environmental assessment, both in terms of human pressure and state. The potentials of land relate to the richness of the natural habitats in terms of extent and biodiversity, to the characteristics of the soil, to the availability of water (quantity and quality of surface and underground waters, regularity of flows...), to economic activities of which it is the support (agriculture, forestry, but also tourism, transport, industry, housing...). These potentials can be assessed from several points of view, one of them being the capacity of the landscape to sustain natural life under the pressure of human activities.
13. The changes in natural wealth of land can be described according to major sources of information including inventories of land cover, land use, fauna and flora.
14. The land cover results from both the use of land by activities and the natural process, whether modified by human activities or not. Land use may result in changes in land cover (deforestation, building of a road, urbanisation...) or in changes of the conditions of the natural or modified biotopes (due to the use of fertilizers and pesticides or to leaving it fallow, due to the intensity of traffic on a road, due to the density of population in a town...).
15. Fauna and flora inventories provide information on the number of species in biotopes or on the population of selected species (endangered, of international status like some migratory birds or considered as indicators of the state of the biotopes).
16. The understanding of the relations between socio-economic activities and their consequences on environmental conditions often implies taking into account the functioning of elementary units described in geographical terms, at local level. In this respect, environmental accounting provides a structured framework essential for integrating data from diverse origins, some derived from cartographic analysis, others from sampling inventories or from general surveys of economic actors or activities.
17. Thus, as a preliminary condition for further integration of the data, the approach of the pilot study on land cover/land use accounting led to the design of a general model which makes a distinction between core accounts and target-oriented accounts. This solution provides a minimum of consistency with the common set of core accounts and some principles for geo-referencing and classification. In addition, the flexibility of the general model made it possible to experiment with more detailed accounting by using specific national sources, each target-oriented account being linked to the core.
18. Then, the practical exercises have been worked out starting, on the one hand from the current and expected availability of data and, on the other hand from national concerns with major environmental issues.
19. The general structure is based on a distinction between core accounts, concentrating on the changes in land cover/land use and supplementary accounts, which are issue-oriented.
20. The two main issues considered are the changes in artificiality due to the intensity of use of land and the biodiversity depending on the natural potentials of land. A final linkage can be made between the changes in biodiversity and the changes in artificiality, both in terms of land cover changes (i. e. deforestation, drainage of wetlands, building up on arable land...) and of changes of the intensity of the management of land (without change of the land cover type). The system provides key information for the formulation of policies towards sustainable development.
21. The core accounts provide a basic reference in terms of structure and changes of land cover and land use. They are elaborated from land observation techniques, in which remote sensing by satellite and by aerial photographs have an important role. However, other sources are necessary to detail actual land uses (over and above of what is reflected in the cover of land).
22. The core accounts include matrixes describing state and changes, together with an analysis of the changes in terms of types of flows. Basically, the accounting unit is the surface, either directly observed (remote sensing) or calculated from data collected on plots (some aspects of land use, fauna and flora...) and extrapolated to larger areas.
23. The supplementary accounts rely both on land observation techniques and on other sources of data. In addition, the definition of accounting units may require the reference to physical landscape types (combining relief, climate and geology) and/or natural regions. They provide specific information and linkages with statistics on economic activities or/and on hydrology, fauna, flora and biotopes.
24. The supplementary accounts aim to provide information on specific issues. The initial choice of possible supplementary accounts was pragmatic. Participants in the pilot group chose supplementary accounts according to national concerns and available data.
1. Sealing of soils/ water resources 2. Partitioning of land (by transport networks) 3. Artificiality/naturalness 4. Biodiversity/vegetation 5. Landscape stability 6. Industry (including mining industry) impacts 7. Agriculture impacts 8. Tourism impacts 9. Transports and human settlements impacts
25. Later on, it became possible to develop from the results of national pilot studies a more systematic framework for future land cover/ land use accounting (see the overall conclusions).
26. The core accounts are detailed on Figure 2. They are enhanced by two sets of data:
a. landscape physical types, which, combined with land cover provide landscape types useful for extrapolating data collected on a sampling basis;
b. additional information, necessary in particular for taking account of linear features like streams, hedgerows, roads, railways..., when the scale of the basic survey does not allow their identification.
27. The two sets of supplementary accounts deal with artificiality of land and biodiversity.
28. First, the accounts describing changes in artificiality have to be linked with the origin of the process. Artificialisation can be due to economic activities such as manufacturing and mining industries, agriculture, construction and public works, tourism, transport... An opposite process can take place, for example in the case of aban1doning of previously agricultural land. Economic activities are listed in the ISIC nomenclature. Consumption patterns have to be considered as well, both individual and collective (municipal waste landfills, nature conservation...). For a given production, the intensity of land use depends on the technologies implemented. Artificialisation can be described in terms of changes of land cover (i.e. consumption of natural land by artificial features) or changes in the intensity of use of land for which cover remains the same during the accounting period (i.e. intensification of agriculture).
29. Second, accounts for potentials of land and biodiversity need additional data, mainly on the water resources and on fauna and flora. The natural potentials of land depend on its capacity to support the habitats of fauna and flora. An assessment can be made by comparing actual presence of species with theoretical lists. Changes in biodiversity in most cases are a consequence of human activities.
30. The data required for land cover/land use accounting can be classified in three types: background data, data from general land surveys and specific data. For the three types of data, the geo-referencing (strictly speaking), or the referencing to a grid or to geographical units (administrative units, natural regions...) is an essential characteristic, the type of assessment depending on the precision of the referencing.
31. The background data relate to the permanent characteristics of landscape: relief, geology, soil types, average climate. They are necessary to assess the state of the biotopes as well as the intensity of use of the land. These data are generally available on maps. Administrative boundaries (useful for referencing socio-economic data) can be included in this group.
32. Combined with data on the major types of land cover, the physical background maps permit the breakdown of the territory in landscape types and ecological regions.
33. Land cover data can be obtained from remote sensing or from surveys on sample plots.
34. Remote sensing provides exhaustive assessments of land cover. The so-called CORINE land cover inventory (CLC) implemented by the European Union in its member states as well as in central and eastern European countries will provide, when completed and updated, a consistent basis for future land cover accounting. Derived from satellite imagery, CLC describes 44 classes of land cover units at the scale of 1/100 000, the smallest mapping unit being of 25 hectares.
35. The methodology of CLC, however, does not allow one to describe most of the linear features of landscape, natural (i.e. streams), semi-natural (hedgerows) or artificial (i.e. roads). Taking into account these linear features requires working at a larger scale (i.e. 1/25 000), which implies presently the use of remote sensing from airborne systems (aerial photographs). This information can be obtained from the digitised maps produced by national geographic institutes or through specific surveys.
36. In general, the quantity of information, including the detail of the classification, obtained from remote sensing is linked to the scale at which the survey is worked out. At a small or average scale, the land cover units are largely composite and can be classified according to their main characteristic. At large scales, the land cover units are the parcels or the biotopes and can be used for organising the information according to land use, fauna and flora.
37. However, remote sensing techniques, including sophisticated image processing, cannot provide sufficient information on land use and biotopes. Much can be done to identify crops or types of natural vegetation but some limits result from the observation tool. First, the cost of detailed surveys by remote sensing may be very high and does not allow the production of exhaustive national inventories. Second, some information can only be collected in situ. It is for example the case of the use of a forest (timber production or recreation or protection of soils), the low vegetation in a forest...
38. For these reasons, additional, more detailed information on land use and fauna and flora has to be collected by (sample-based) field surveys (see the British and the French modules, below).
39. Additional specific data are necessary in reference to economic activities or to sectorial accounting. They relate to production, intermediate consumption (i.e. of pesticides), road traffic, emissions of pollutants, disposal of wastes. As mentioned, the geographical referencing of such data determines the detail of the possible accounting.
40. Other data relate to rare, threatened or important biotopes. Their identification does not necessarily depend on geographical or statistical surveys. However, they need to be taken into account.
41. Austria is able to present results of changes in land use and land cover caused by 5 socio-economic sectors : agriculture, forestry, tourism, increase of buildings and road traffic.
42. The contribution of Austria is confined to statistical data, because a geographical information system is not available at the moment. For this reason, it is only very partially possible to combine these data with natural regions.
43. Two sources of data are to be used : Corine Land Cover and TERUTI.
44. Although the theoretical work is well advanced with CLC, the exercise has to be postponed due to difficulties in recovering an old Landsat image on the test area. The work relating to changes in the structure of landscape due to human activities, including the partitioning of land, may not be carried out before 1996.
45. Both to overcome the mentioned difficulty and to test the interest of another data source for the problem at hand, a national exercise is presently done with TERUTI data.
46. This survey is carried out yearly by the Ministry of Agriculture and has two parts. A physical classification (reasonably close to the CLC classification) provides data on land cover. A functional classification provides statistics on land use, with a focus on crops but also a description of built-up areas. TERUTI consists in detecting land use on the ground around points represented by crosses on aerial photographs. The sampling is systematic and aligned. The sample consists of 15 449 photos with a square grid of 36 points on each of them. The results can be mapped at the level of the Region or the Department.
47. Accounts of stocks and total changes for the period 1982-86 are currently compiled and will be completed with results for 1986-90.
48. Presently, a plan is developing to test a detailed classification of flows (or changes). In many cases, nonambiguous flows can be identified through changes in land cover or/and land use. In other cases, difficulties arise from insufficient details of the classification of land use/land cover or because of multiple causes of changes or multiple uses of land. Solutions have then to be found in combining information on changes of both land use and land cover. Some difficulties are not yet solved. An example for the case of forests is provided in annex.
49. The linkage between land use and economic activities will not be carried out before the end of 1995.
50 The data sources are STABIS (1952, 1972, 1992) and Corine land cover (1992, possibility to go back to 1984 with Landsat TM images available). Due to the fact that there are linear features in STABIS but not in CLC, some accounts are only made from STABIS. Additional data on biotopes are expected from local authorities.
51. Another difficulty is that land use and land cover are not independent and that there is presently no independent source for land use data. 52. The core set of accounts is linked with economic activities. Rough estimations have been computed. Best results could be obtained on a sectoral basis (ie, transport or waste disposal by economic sector), were data do exist in official statistics.
52. Other additional accounts are prepared for topics as partitioning of land (from STABIS only), sealing of soils (through a vegetation index for urban areas, using Landsat TM images) and artificiality/naturalness. At this stage, the hydrological aspects of soil sealing are not covered.
53. Artificiality accounts need information on biotopes. However, a classification of 7 degrees of artificiality has been established, taking into account the ability of self-regulation of the system, from natural ecosystems only influenced by air pollution on one side of the spectrum, to artificial ecosystems mostly formed by artificial structures.
54. Each account is composed of three levels : the stocks, the total flow or change in stocks between two periods and the desegregation of flows. The last level is the most difficult, due to the difficulty to separate the causes for changes and, above all, to make a clear distinction between human impacts and natural processes.
55. In the UK, the work is seen in the context of the formulation of a sustainable development strategy and forms part of a long-term monitoring programme. The accounting exercise will be based mainly on the Countryside Survey (CS 1990) and Corine Land Cover (CLC). The latter involved a feasibility study to convert the Land Cover Map of Great Britain to the Corine standard, using semi-automatic processing techniques, which has been completed with satisfactory results.
56. The work is described in the complete Task Force report. The core accounts are enhanced by supplementary accounts aimed at describing biotope (classified in 29 Main Plot Classes) changes according to the intensity of use of the land. Botanical diversity is assessed from the changes in the mean species number in each of the major biotopes. Intensity of use (five classes) is derived from the nutrient status of vegetation which is obtained in turn from the application of a multivariate statistical technique.
57. The accounts will be based either on georeferenced exhaustive data (ie land cover derived from the analysis of remotely sensed data) or on samples of point data (ie, land cover and vegetation plots in CS1990). A background breakdown of the territory in a few landscape types based on physical parameters (climate, relief, geology) combined with land use has been made. These landscape types are closely correlated with patterns of agricultural activity and ecological characteristics. They provide a good basis for sampling and for making extrapolations to regions.
58. The working group agrees with the necessity of such a mapping of landscape types and two countries intend to follow the UK approach.
59. The work carried out by the Pilot Group was coordinated on two occasional meetings and one final meeting (apart from the plenary meetings of the Task Force), held respectively in Rome (21 and 22 May 1993), Voorburg (11 and 12 April 1994) and Geneva (24 November 1994). In addition, discussion was continued by correspondence.
60. Besides the overall effort made by each country, written contributions were provided on the occasion of the first informal meeting: background information on nutrients, by Finland (Kauppi, 1993); a discussion on environmental problems related to nutrients and information on a project on environmental accounting of the Nordic Council, by Sweden (Fangstrom, 1993a; 1993b); information on the Norwegian resource accounting system, by Norway (Saebo, 1993); documentation on Dutch balance sheets for nitrogen and phosphorus (Olsthoorn, 1993), the description of material balances for phosphorus and nitrogen provided in a proposed contribution to a UN-ECE manual (UN-ECE, 1993a) as well as a set of provisional proposals in terms of tables, classifications and matrix presentation of the tables (Bosch, 1993a), by the Netherlands.
61. Following the first informal meeting, each country participating in the pilot study concentrated its efforts on testing a number of tables (except Hungary, which was not in a position to participate actively, due to resource constraints). The said tables were, initially, those proposed by the Netherlands. In a second step (following the second informal meeting), a new set of tables was considered for testing. They represented the common understanding of a possible accounting framework, mainly based on the Nordic countries' experience; they were expected to provide greater feasibility. Then complete case studies, based on tests of the proposed accounts were developed (one for each country), following a model provided by Italy (Bonanni-Costantino, 1994).
62. Subsequent methodological thinking took ongoing work in progress concerning the case studies into account. Besides the methodological work developed by the individual countries through the case studies and in addition to the overall methodological contribution provided by each country through discussion by correspondence, some written methodological contributions were made dealing with the following subjects: the use of (parts of) an environmental accounting framework as database for input-output analyses (Bosch, 1993b); physical environmental accounting in the context of integrated environmental and economic accounting as developed by the United Nations (Costantino, 1994a); proposals on how to deal with emissions of nitrogen and phosphorus contained in wastes (Costantino, 1994b).
63. According to the conclusions reached at the CES Task Force meeting held in Geneva at the end of 1991 (Conference of European Statisticians, 1991), the pilot study on nutrients and the environment was expected, to the extent possible, to specify the following: the actor accounts needed; the stocks and flows subject to accounting; their corresponding accounting units, including the necessary classifications; the movements between different environmental media/uses of the environment; changes in relevant environmental conditions and their effects; possible aggregate environmental indicators to be derived from the envisaged accounting system. The accounting methodology was expected to be developed for countries and major natural geographical units as applicable.
64. In accordance with a basic rationale adopted by the Task Force (UN-ECE, 1992b), human activities, material flows and environmental effects were considered as the main building blocks for physical accounting on nutrients and the environment. In other words, the focus was put on the human activities that are responsible for certain environmental effects of use and emission of nutrients, on these effects and on the relevant material flows that may explain how the said activities produce such effects. Describing environmental developments in terms of key ecological indicators or variables both within and between the building blocks was then considered to be the primary aim for the accounting scheme. Possible linkage of the envisaged accounts with economic accounts and compatibility with relevant international standard classifications, such as ISIC or NACE for the accounting of human activities, were regarded as essential requirements to be met.
65. Particular attention was paid, in general, to the capacity of physical environmental accounting to meet, once fully developed, the full range of requirements which statistical offices usually specify with regard to accounting techniques of data organization (e.g. consistency checks of statistics involved, identification of data gaps, use of data in modelling, etc.). Furthermore, the objective was also considered of specifying and arranging the accounting framework in such a way that meaningful correlations can be investigated by the users of accounting data.
66. Following a broad scheme developed by the Task Force (UN-ECE, 1993b), in an attempt to develop a number of possible accounts a distinction was made between "core accounts" and "supplementary accounts".
67. An appropriate consideration of nutrients and the environment from the viewpoint of natural sciences was deemed preliminary to any discussion on the relevant statistical issues. With regard to this, different aspects related to functioning of ecosystems, as well as to pressure on, state and use of the environment were taken into consideration. The following was noted in particular.
68. The concept of nutrients, in an environmental context, is generally applied to certain elements which are necessary for mineral nutrition of plants and, among these, especially the elements nitrogen and phosphorus (Fangstrom, 1993).
69. Phosphorus and nitrogen are relatively scarce on earth in available forms and ecosystems have developed mechanisms for conserving nitrogen and phosphorus compounds in relatively closed cycles. This means that the said elements are normally incorporated into the plant biomass and maintained there to a certain extent; some loss from the land surface occurs, nevertheless, particularly in periods during which precipitation is greatest and plant biomass is minimal. That is considered to be natural. Man-made impacts, on the other hand, may increase the input of nutrients to ecosystems beyond natural levels. The direct emissions of nutrients in biologically available forms (phosphate, nitrates, ammonium) from point sources (e.g. untreated domestic sewage, effluent from municipal and industrial waste-water treatment plants), as well as run-off from agricultural and urban areas, cause severe eutrophication problems in aquatic ecosystems (UN-ECE, 1992a). As eutrophication proceeds, in fact, the species composition of phytoplankton as well as fish changes to less desirable (Kauppi, 1993). The effects of eutrophication on an aquatic ecosystem, furthermore, can render the water unsuitable for many traditional uses.
70. From this point of view, eutrophication may in particular adversely affect (UN-ECE, 1992a): a) human health through contamination of drinking water, shellfish and marine bivalves with toxic substances excreted by algae; b) raw-water treatment, leading to higher treatment costs; c) irrigational water use, due to excessive growth of aquatic macrophytes in irrigation channels; d) fisheries, because of loss of economically valuable fish species; e) recreational use of surface waters; f) navigation, because of loss of bank protection by plants in navigable waters, increased rate of sedimentation; etc.
71. Apart from eutrophication, however, environmental and resource problems related to the use and emissions of nitrogen and phosphorus compounds are also caused by other characteristics of these compounds, not only by their being necessary elements for growth and development of plants (Fangstrom, 1993). To summarize, among the various presently known environmental and resource problems related to the use and emissions of phosphorus and nitrogen compounds, the following problems were pointed out (Fangstrom, 1993): a) phosphate and nitrogen concentrations in soil as eutrophicating agents with influence on natural vegetation and fauna; b) phosphate and nitrogen emissions and leaching as eutrophicating agents in inland waters; c) phosphate and nitrogen emissions and leaching as eutrophicating agents in sea waters. In addition to that, as far as phosphorus is concerned, the following problems were also pointed out: d) cadmium emissions due to the use of phosphate fertilizers; e) scarcity of phosphate mineral; and last, but not least, the following problems were pointed out with reference to nitrogen: f) acidification; g) high ozone concentrations at ground levels, causing injuries to vegetation and risks to human health, due to nitrogen oxides; h) destruction of ozone in the stratosphere due to nitrogen oxides; i) possible change of the global climate due to nitrogen oxides.
72. A general approach was envisaged according to which the following aspects should be highlighted in particular in the physical accounting framework to be developed: extraction of raw materials, imports and exports of goods, use and production of goods, emission of pollutants, transboundary flow of pollutants, redistribution of pollutants between environmental media and state of the environment (Bosch, 1993a); such phenomena should be focused as far as nitrogen and phosphorus are concerned. The diagrammatic presentation reported hereafter (Hellsten, 1994) provides a synoptic view of the general approach followed.
73. As it can be seen, the proposed framework is basically in harmony with the SEEA proposals concerning integrated environmental and economic accounting in physical terms (United Nations, 1993), one main difference being the scope of the accounting system. On one hand, in fact, while the SEEA proposals provide a comprehensive framework including both physical and monetary accounts linked together, the UN-ECE Pilot study is limited to physical accounting; the scope of this latter, of course, is narrower also because only nutrients are covered. On the other hand, while the physical part of SEEA is limited to recording physical flows from natural assets to the economy (use of natural assets) and flows back to the natural environment (residual flows), flows and transformations within the natural environment (e.g. redistribution of pollutants between environmental media and transboundary pollution) are not neglected, in principle, in the UN-ECE pilot study. Keeping in mind all the above, the core part of the proposed accounting scheme can also be regarded as further work in the context of the SEEA proposals concerning the "residual flow accounts" expressed in physical units, since the said core part is devoted to emissions and activities generating them.
74. On the basis of the envisaged general approach, a number of specific accounts were developed in detail. The focus was put on nitrogen and phosphorus emissions, with a breakdown of the by economic sector of origin, type of material flow involved and receiving environmental media. In addition, transboundary flows of nitrogen and phosphorus in air and water were considered as part of the "core accounts", while "supplementary accounts" were devoted to the description of exchanges of nitrogen and phosphorus between different environmental media. A supplementary accounting scheme was specifically developed with reference to treatment of waste-water and waste containing nitrogen or phosphorus, carried out by specialised activities (ISIC O90). Moreover, the use of appropriate balance sheets related to agricultural land was deemed necessary for calculating the net emission of nitrogen and phosphorus in manure, fertilizer and sewage sludge from the agricultural sector to the soil/groundwater.
75. The following graphical frameworks concerning nitrogen and phosphorus (Bonanni-Costantino, 1994) correspond strictly to the accounts.
76. In the graphical frameworks, the material flows taken into account are indicated in the rectangles. The economic sectors (economic activities and the household sector) are shown in the upper part of the picture, while the environmental media are at the bottom. The arrows shown above the economic sectors represent the material flows delivered to treatment of residuals (specialised activities, ISIC O90). The arrows shown below the economic sectors represent the material flows that go into the natural environment. Each arrow represents the material flow indicated in the rectangular box that it enters; these material flows are also given numbers (1 to 5). The capital letters put at the end of the arrows indicate the economic sectors that generated the material flows involved (e.g. A stands for Agriculture, T for Treatment).
77. In the case of agriculture, the boxes showing "Manure, fertiliz." and "Sewage sludge" represent amounts of these materials used as inputs in agricultural production; the same could apply, in principle, to other economic sectors, even if this is not considered in the graphical frameworks, for the sake of simplicity.
78. As far as treatment of residuals is concerned, it is to be kept in mind that the material flow going to water is represented by treated ("purified") water coming from waste-water and still containing nitrogen or phosphorus. On the other hand, the material flow going to the soil/groundwater does not consist of waste and sewage sludge containing nitrogen or phosphorus; it is material flow that is generated instead by the stock of waste and sewage sludge accumulated in disposal plants. Detailed proposals regarding the accounting treatment of material flows linked to the generation and accumulation of waste are reported under item 6 below (Costantino, 1994).
Proposals on how to deal with "Nitrogen/phosphorus in wastes" as a separate category under the column "Soil/groundwater" in accounts on emissions
79. Wastes are given special attention in the accounting framework envisaged by the Pilot Group, in as much as they contain nitrogen/phosphorus (N/P). According to the proposed general framework, each of the relevant flows of "N/P in wastes" should find its place in some of the envisaged accounts, as appropriate.
80. As far as "soil/groundwater" is concerned, the main aspects to be highlighted seemed to be the following: a) generation of wastes containing N/P delivered to treatment/disposal (specialised activities, ISIC O90); b) accumulation in ISIC O90 of N/P contained in wastes after treatment/disposal; c) emission of N/P to soil/groundwater by ISIC O90, due to the stocks of wastes accumulated in disposal plants (leakage); d) direct discharge of wastes containing N/P to soil, outside the waste treatment/disposal system.
81. With regard to the most important phenomena under consideration as mentioned above, three specific accounts, among those selected for testing, are to be focused: 1) the account on material flows containing N/P delivered to ISIC O90; 2) the N/P balance sheet of ISIC O90; 3) the Emission of N/P account. Having selected the mentioned tables for testing and presentation, their consistency within the proposed accounting system would have to be ensured, while the links between the physical aggregates shown in the accounts would have to be made evident to the extent possible.
82. Three aspects among the starting points of the rationale followed in the pilot study should not be neglected in particular: 1) emissions were intended as having their origin in the technosphere and their destination in the environmental media; 2) in the N/P balance sheet of ISIC O90 the breakdown of the "Outflows" would include both "Economy" (for recycling of materials containing N/P) and "soil/groundwater" (for emission of materials containing N/P); 3) accumulation of wastes containing N/P in ISIC O90 would not be treated as emission to soil/groundwater, while being regarded, in harmony with the SEEA proposals, as accumulation of wastes (and of N/P thereafter) within the technosphere.
83. On the basis of what is considered above, it is suggested to include explicitly in the emission of N/P account, under the column "soil/groundwater", the following emissions: a) the corresponding emissions taken into account in the N/P balance sheet of ISIC O90; b) the emissions of N/P contained in wastes discharged directly to soil by the remaining economic sectors. The said emissions would be shown as a separate category under the column "Soil/groundwater", in addition to the category proposed earlier, dealing with manure, fertilizer and sewage sludge.
84. The proposed additional category would also include N/P in sewage sludge as a type of waste, insofar as the said sewage sludge is stored in landfills by ISIC O90 or is discharged directly to soil by economic sectors.
85. Sewage sludge shown in the N/P balance sheet of ISIC O90 among the "outflows" as a material flow (containing N/P) going to "Economy" would enter, on the other hand, the emission of N/P account under the category dealing with manure, fertilizer and sewage sludge.
86. In conclusion, two categories are proposed for inclusion under the column "Soil/groundwater" in emission of N/P account: a) "Tot-N/P in manure, fertilizer and sewage sludge used as inputs by economic sectors"; b) "Tot-N/P in wastes".
87. The last category would cover all the relevant material flows linked to the generation and treatment/disposal of unwanted residuals containing N/P (not flows of excessive use of N/P inputs by economic sectors, as it is the case with "Tot-N/P in manure, fertilizer, etc."). The flows thus to be taken into account would represent either leakage from wastes stored in landfills (as far as ISIC O90 is concerned) or N/P accumulated in wastes discharged directly to soil (as far as the remaining economic sectors are concerned, since accumulation of N/P in ISIC O90 would be shown in the related balance sheet as affecting the technosphere and not the soil/groundwater).
88. In the case of N/P accumulated in wastes discharged directly to soil, the amount of N/P to be calculated would be the N/P content of the wastes; this would comprise the amount of N/P resulting as a flow of N/P to groundwater due to leakage. Possible assessment of this latter flow would then be left to the users of the proposed accounting system, it being clear that the corresponding amounts of N/P would actually be included (as part of) in the figures shown in the proposed emission of N/P account. With reference to economic sectors other than ISIC O90 it is suggested, thus, to drop the possible solution of entering under "Tot-N/P in wastes" figures representing leakage, because such a solution would prevent data on N/P accumulation on soil from being shown in the emission of N/P account. The final result thus would be that accumulation of N/P in wastes would be described partly as affecting directly the soil and partly as confined within the landfills, while leakage to groundwater would be shown in the emission of N/P account only with reference to ISIC O90.
89. Turning to "Tot-N/P in manure, fertilizer and sewage sludge used as inputs by economic sectors", the physical aggregates suggested for calculation would represent: a) "net emission" of N/P, as resulting from an appropriate balance sheet, as far as agricultural activities are concerned (ISIC A01 and perhaps cultivation carried out by "Households"); b) accumulation of N/P in the relevant materials (sewage sludge) used by economic sectors as inputs to their activities other than cultivation.
90. A general consequence of the suggested accounting treatment would be that certain figures on N/P emissions to the "Soil/groundwater" would regard soil and groundwater considered jointly (e.g. data on "net emission" due to agricultural activities and data on leakage due to ISIC O90), while other figures, such as those relating to direct discharge to soil of wastes containing N/P, would describe emission in terms of accumulation of N/P on the soil as such (as distinct from groundwater).
91. The overall conclusions were drawn in relation to the question:
92. Should the Conference of European Statisticians embark on the development of a comprehensive conceptual framework for physical environmental accounting?
93. While the Task Force had worked only on identified particular environmental concerns, an attempt was made to give a general answer to this question.
94. The Task Force reviewed its work with regard to two special concerns. Firstly, the question of possible generalization of the lessons learnt from the exercise was asked. Secondly, the possible benefits and insufficiencies of physical environmental accounts in general were assessed from the point of view of the more limited investigation undertaken, as well as from that of their relations with current priorities of environmental policies and existing neighbouring statistical information systems.
95. The national case studies done within the Pilot Group on Land Use/Land Cover provided a first, tentative approach to environmental accounting of land use/land cover. Similarly, the case studies done by the participants in the Pilot Group on Nutrients and the Environment also resulted in a general blueprint for physical environmental accounts in this area of concerns. The following sections provide a general overview of the results achieved in each of the pilot studies.
96. Land cover/land use accounting, with its links to human activities on the one hand and to biodiversity on the other, can be considered a tool for sustainable development policies. It can provide information at national and regional scales and can be used both for national accounting of natural assets (as defined in the SNA and the SEEA) and in land planning. Comprehensive indicators can be computed from these accounts.
97. Starting empirically, the conclusions obtained in the pilot studies lead to the design of a framework for land cover/land use accounting.
98. The changes in land cover resulting from changes in land use reflect the richness of nature and its renewal. They provide useful information on the possible sustainable use of ecosystems, either natural, modified or artificial. These potentials can be assessed (positively) according to the biodiversity and the availability of natural components, and (in general negatively), to the amount of human input necessary to ensure the functioning of these ecosystems.
99. Thus, the framework identifies linkages between land cover and land use with both economic activities as described in national accounts and natural components, with special attention to fauna, flora and biotopes. On the economic side, a clear linkage can be made to the classifications of the so-called SEEA satellite account of the revised SNA. Both classifications of non-financial assets (CNFA 2.1.3. Land incl. ecosystems and soils, except 2.1.3.1. Soils) and other volume changes of non-financial assets (COVC, several items) can be used with minor adjustments. The UNECE standard land use classification is the appropriate classification for land use, when the EU Corine land cover nomenclature is implemented in a large number of countries. In that respect, land cover/land use accounts can be considered fulfilling the needs of the relevant building block (Accounts of land use, ecosystems) of "Version III - Physical accounting" of the SEEA. Another important linkage with SNA relates to the economic flows as described by ISIC. The future development of materials and energy balances will also have to consider the land dimension.
100. On the nature side, the linkage will be made easier with the development of other natural resources accounts, in quantity and quality, for water, soils, atmosphere and climate, fauna, flora and biotopes.
101. The intensity of use and the artificialisation process can be considered at different scales. A first set of indicators can be computed from general data on land cover : surface of land cover units (so-called ecozones), structure of landscape (size of the ecozones, distance between ecozones of a same type, impacts of artificial linear structures...). Further, artificialisation of ecozones which remain of the same type need additional information on stresses due to activities or can be deduced from their consequences on biodiversity.
102. In principle, the best linkage between activities and land has to be found through land use. However, land use information requires field data collected from sources that are independent from land cover surveys. When such data are not available, land cover data at large scale (where a good correspondence between land use and land cover units can be found) can be accepted as a proxy for land use. Thus, a linkage can be made between activities and land cover.
103. Another problem arises at this stage : economic data are not georefenced but collected trough institutional or functional units which can be referred to administrative units. The integration of both sets of data has to be achieved at the scale of these units, with possible further geographical breakdown. A similar problem relates to fauna and flora statistics collected through sample-based field surveys. The data can only be extrapolated to natural regions including physical characteristics. These analyses require the use of a geographic information system.
104. Considering the data, the pilot exercise shows that some work can be done in spite of discrepancies and gaps. For the future, needs have been identified from the present difficulties. As long as it refers to a single classification, the Corine land cover programme, presently realised in 20 European countries, can possibly be the basis for future accounting. However, some improvements are necessary, mainly in terms of smaller mapped unit, whose size is presently of 25 hectares. According to the moderate rhythm of the changes in Europe and to the large number of small "ecozones", the present degree of aggregation of Corine forbids accounting for regions under a certain size and/or for periods of time shorter than five years. An improvement of quality could be obtained by mapping units as small as 15 ha.
105. Considering other uses of remote sensing by satellite, which offers economic advantages for data collection on land, it was noticed that further research is necessary, in particular for an operational use of the vegetation indexes, which could potentially provide useful assessment of the sealing of urban areas.
106. Although standardized field surveys of both land use and fauna and flora are not expected in a near future, an improvement both in quality and in international comparability could be obtained by structuring the sampling patterns in reference to landscape physical types as described previously.
107. Finally, one must insist on the necessary cooperation between statisticians, geographers and biologists. In several countries such a cooperation does exist and, for example, statistical offices developed their own capacities in GIS. However, in other countries this is not yet the case, and some involvement of national geographic institutes and research organisations could be necessary for the future development of land cover/land use accounting.
107. Nutrients are necessary elements for mineral nutrition of plants, but they may cause severe environmental problems (e.g. eutrophication), when present in excessive amounts. The circulation of nitrogen as well as of phosphorus in the technosphere and the natural environment was taken into account in the pilot study, duly considering, to the extent possible, the role of the said substances in natural processes. Key words were pressure on, state and use of the environment. This lead to the design of a general approach to physical environmental accounting for nutrients.
108. For the comprehensive scheme thus developed to be implemented, however, a limiting factor appeared to be scientific knowledge about the flows of nutrients in the natural environment and the environmental effects of their use and emission. Limited availability of data turned out to be also an obstacle for the development of a fully extended system of physical accounts on nutrients and the environment. The construction of such a system was then considered as a long run task.
109. In the light of the envisaged overall framework, a subset of accounting schemes was then developed and tested in the case studies. The proposed set of accounts appeared to be much more feasible for the time being.
110. The specific tables tested, as well as the general framework, proved to be compatible with the relevant SEEA proposals. They could be part, theoretically speaking, of "Version III" of SEEA and could be regarded, to some extent, as a further specification of residual flow accounts drafted therein, dealing in particular with nutrients.
111. The focus of the proposed accounting scheme is mainly on emission of substances containing nitrogen and phosphorus. A balance sheet related to treatment and disposal of the said substances carried out by specialised activities (ISIC code O90) was among the most meaningful building blocks of the entire framework. Another one was represented by a balance sheet in which "net emission" to soil/groundwater of nitrogen/phosphorus contained in manure, fertilizer and sewage sludge used as inputs to production by the agricultural sector was proposed for calculation.
112. A set of pressure indicators could be concluded from the proposed tables, among which: emissions of nutrients to air, water and soil/groundwater, with a breakdown by generating sector (e.g. "net emission" due to the agricultural sector) and type of material flow containing nitrogen/phosphorus; amounts of wastes containing nutrients delivered to specialised activities for treatment/disposal; accumulation of nitrogen phosphorus in disposal plants. Such indicators would enable the whole scheme to serve as a tool for sustainable development policies, both for helping definition of the said policies and for checking to what extent economic/environmental objectives are reached.
113. While the case studies provided an overall positive test for the proposed physical accounting scheme, they also highlighted a number of issues and difficulties, not only of a statistical nature, which would help to develop further work in this field. Among the main directions for future developments which emerged from the pilot study, cooperation between statisticians and natural scientists appeared to be a key factor.
114. On the basis of the concrete results of the case studies, it appeared that an immediate follow-up of the work done would enable an improvement, for the near future, in the quality of the statistical information already existing on this subject, with considerable benefits, among other things, for international comparability. More significant advancements towards the establishment of an overall system of physical environmental accounting at the international level can be expected in the medium/long term.
115. The work done in each of the Pilot Groups demonstrated compatibility between the physical accounting done and schemes of monetary accounting in this field, notably the SEEA. The Task Force identified also areas of work, which would require coordination between the work undertaken in each of the Pilot Groups. Deposition of pollutants in conjunction with critical loads and nutrient balances in agriculture were among these areas.
116. In addition, the benefits usually accruing to accounting techniques became apparent in the case studies. One major advantage of accounting relates to improved possibilities for the joint use of a) data from different data systems, in particular environmental and economic data, and b) physical and certain types of monetary data about the environment. Also, the accounts helped to present conclusions embodied in the data sets, which were not evident. Furthermore, accounting appeared to favour the consistency of data sets by enhancing their methodological standardization.
117. In addition, gaps in and between existing data sets become more readily apparent, thus enabling environmental accounts to guide future developments of environment statistics. The accounts also favour continuity of statistical practices and stability of essential time series. Finally, accounting data were seen as being particularly valuable for modelling, which, in conjunction with scenario setting, appears to be of growing importance to environmental information and studies.
118. The case studies undertaken within the Pilot Groups were placed from the outset in the context of sustainability and biodiversity considerations. The Task Force concluded that the accounts developed show a high potential for the derivation of sustainability and biodiversity indicators, although actual decisions on such indicators involve both research and political aspects. Accounts can assist in the accomplishment of this work. The circumstance that many institutions are currently developing such indicators, together with the fact that environmental accounting is very much under development make it imperative that any future work by the CES be properly coordinated with all relevant activities undertaken elsewhere.
119. Finally, the Task Force found the organizational arrangements made for its work adequate. In particular, the steering of substantive work entrusted to national experts appeared to be a suitable form of international cooperation in an area, where the volume of work is considerable.