2.1.4    The concept of Sustainable Development has been embraced by the Hong Kong Government which commissioned the Study on Sustainable Development for the 21st Century (SUSDEV21) to explore means of integrating consideration of economic, social and environmental issues into the planning and development process. To this end, SUSDEV21 has developed indicators for measurement of the development impact of projects. However, SUSDEV21 has not yet established an indicator against which the landscape baseline conditions can be monitored. The Study is intended to provide this information and thus facilitate the process of sustainable planning and development.

2.1.5    Data acquisition is to be achieved by a combination of desktop study and field survey. All collated data shall be transcribed into a coherent and consistent set of digital data, including a landscape character map, in an interactive Geographic Information Systems (GIS) format. The specific objectives in relation to GIS include:

• Establishing a comprehensive, flexible and open GIS and textual database which will act as a repository for the final Landscape Character Map and all background information used in the study;
• Providing the GIS in a format compatible with the Computer Aided Sustainability Evaluation Tool (CASET) system developed under the SUDEV21 Study;
• Recommending an indicator which will measure the impact of various proposals and policies on landscape for use within the CASET system;
• Generating CD-ROM and hardcopy versions of the Landscape Character Map.

2.1.6    It is intended that the indicator derived from the study will be used to measure and evaluate changes to Hong Kong's landscape resources over time. The data will thus be compiled in such a way as to facilitate the updating of the landscape resource database on a regular basis.

2.1.7    It is important that the views of the public and interested bodies are solicited at key stages in the project and that the views of all stakeholders are incorporated into the landscape character mapping process.

2.2      KEY ISSUES

2.2.1    Key issues can be identified under four main headings relating to Management Issues, Programme Issues, Methodology Issues and GIS Related Issues.

2.2.2    Emphasis will be placed on value management to ensure that the optimum character assessment and evaluation methodology is attained and that appropriate methods of characterising and evaluating the landscape are undertaken. A Value Management Workshop is proposed as outlined in section 3.4.

2.2.3    A Project Quality Plan has been created specifically for the Study. Rigorous Quality Assurance procedures will be applied throughout the Study.

2.2.4    The Study is necessarily a large and complex one requiring the synthesis of a large amount of data from a wide variety of sources. Consequently, it is essential that a constant management overview is maintained by a team of senior staff with experience in their field and of similar work in Hong Kong. To this end, a Technical Review Panel has been formed to ensure consistency and validity of approach throughout the Study (see section 3).

2.2.5    With a wide range of inputs from a diverse range of staff, both foreign and local, from different fields, a Team Structure has been established that minimises problems of communication and information synthesis. Urbis has developed an integrated team structure that will deal with these such problems.

2.2.6    Protocols will be established for the transfer of electronic information between the consultant team and Planning Department to ensure a fast and smooth flow of data input and data output.

2.2.7    The team will work closely with Planning Department to establish and maintain a realistic programme to ensure timely completion of the study.

2.2.8    It will be critical to get fast comment and approval of submissions to ensure that the programme can be adhered to. PlanD may need to seek timely co-operation from other Government departments in this respect.

2.2.9    At all stages of the Study, the team will keep in mind the fact that the ultimate purpose of the Study is not to undertake the landscape character mapping for its own sake, but to provide a tool for the planning and development decision making process in Hong Kong. This over-riding consideration will inform all the team's decisions on the appropriateness of alternative methodologies considered, reviewed and eventually adopted within the Study. Ways in which the findings of the Study could inform decision making relate to:

• Strategic Land Use Planning;
• Development Control;
• Environmental Impact Assessments undertaken under the auspices of the EIAO and TPO (Section 16 Applications);
• Landscape Management Strategies; and
• Influencing Design Criteria for new development.

2.2.10  Involving stakeholders in the Study process will be valuable for a number of reasons:

• It can help people to understand the existing landscape and its importance to Hong Kong;
• Valuable information may be contributed by stakeholders who can thus add to the value of the Study;
• If stakeholders are involved in the Study process they are more likely to support and be committed to the Study recommendations;
• The process can help to build consensus.
• The process helps to make environmental initiatives more efficient and sustainable.

2.2.11  Stakeholders will be invited to participate in the study by a variety of methods - principally through circulation of reports and invitation to comment, discussion and dialogue at Working Group Meetings, and Public Consultation Fora. Key stakeholders will also be invited to the Value Management Workshop.

2.2.12  It is generally accepted that landscape character assessment involves both objectivity and subjectivity. The process of characterisation is largely objective, as it involves the synthesis of elements that may be capable of being quantified objectively, whereas making judgements on the value of landscape clearly involves subjectivity. However, even in surveying the elements that make up the landscape, or mapping and describing LCTs and LCAs, subjective judgements inevitably become involved. This can only be avoided if the task is reduced to a mere measurement of quantifiable properties (such as area of woodland). However, such an approach would not recognise the genius loci. It will therefore be important in the Study to clearly differentiate between those aspects which are objective and those which are subjective. Consistency in subjective judgements then becomes another key issue, which is addressed below.

2.2.13  As there will be several field surveyors working in the field simultaneously, a key issue for the Study, if meaningful results are to be achieved, is to ensure that all field surveyors approach their tasks in a consistent manner and that room for subjective judgement and preference is reduced to a minimum. Furthermore, as the field survey will be undertaken over a period of several months it is also important to ensure that each survey team is consistent throughout time. We suggest a number of methods to ensure consistency in section 3.

2.2.14  Landscape character assessment can be applied at a number of different scales from the national or international scale down to local district level. Ideally, assessments at different scales should fit together as a nested series or hierarchy of LCTs and LCAs so that assessment at each level adds more to the one above. The Brief has identified the scale for the Study as being based on 5 hectare units. However, depending on the size of LCT in different areas, it may be necessary to amalgamate two or more 5-hectare units in order to arrive at LCTs that sensibly reflect the character of the landscape. As the ultimate purpose of the Study is to aid decision making in planning and developing Hong Kong, the most appropriate scale will be the one that can be applied with the most ease to the planning and development process in Hong Kong, particularly in relation to the landscape and visual assessments that are undertaken regularly as part of the town planning and environmental impact assessment processes.

2.2.15  One aspect of the project which is likely to cause confusion amongst stakeholders who are not landscape professionals (i.e. most stakeholders) is the semantics and nomenclature that is commonly adopted in landscape character assessment - i.e. "landscape character types", "landscape character areas", "landscape value" etc.. There can even be confusion amongst landscape professionals about what these mean. It will be important therefore to convey to all concerned a clear understanding of the concept of landscape character and the differentiation between LCTs, LCAs and landscape value. This will be set out clearly in the Technical Report 1 and explained in simple terms in all public consultation documents.

2.2.16  Landscape is a continuum and it is not always easy to draw a distinct boundary between two LCTs or two LCAs. One problem with drawing boundary lines is that they may convey a much more obvious change than is really apparent on the ground. Some character assessment studies have therefore adopted broad belts of transition between LCAs rather than precise lines. It is likely that this would be difficult to do at the scale of assessment to be addressed in the Study. Nevertheless, careful consideration will be given to identifying the methods by which boundaries between LCTs and LCAs should be drawn in the Hong Kong context.

2.2.17  Whilst it will be necessary to continually review and update the assumptions with regard to LCTs and LCAs as the Study and field survey progress, it is also important that that sufficient expertise be brought to bear early in the Study to ensure that the initial classification and categorisation is as accurate as possible. Getting it almost right at the beginning will reduce abortive work and programme problems later in the Study. To this end, the Technical Review Panel, comprising senior staff at Urbis and ERM with over 50 years combined experience of working on the environment in Hong Kong, will sit early in the study to ensure that the right direction is being followed.

2.2.18  A key issue to be determined by the Study is whether to adopt a 'resource based' methodological approach (e.g. McHarg system in the USA) or a 'character based approach' (e.g. UK systems). Most probably, the methodology adopted will be a combination of the two ideas.

2.2.19  Quantifiable data is easily handled by GIS systems. Qualitative data (e.g. 'scenic beauty' or 'character') is more difficult to compare. It is important therefore that the methodology adopted pursues an 'integrated approach' (such as those used in New Zealand) where both sets of data can be synthesised equally.

2.2.20  A significant amount of work has been done world-wide on methodologies for rural or urban fringe landscape classification. However little work has been done to find approaches which cater for both urban and rural landscapes equally. Developing a methodology which achieves both of these goals will be one of the more challenging parts of the Study. Urbis will use its in-house team of urban designers and planners to assist with the development of classification and evaluation criteria for urban areas.

2.2.21  Two potential approaches to landscape evaluation are common. The first is to evaluate landscapes on an area basis, that is, evaluating an area of landscape (e.g. LCT) as a whole. An alternative approach is the 'environmental capital approach' whereby individual landscape features are evaluated. Both are valid approaches and a key issue in the Study will be to determine which is the most appropriate approach in Hong Kong.

2.2.22  Defining a single indicator that will allow change in landscape quality to be assessed over time is certainly one of the key issues in the Study and perhaps the most challenging. With their experience of SUSDEV21 and CASET, ERM is ideally placed to advise on the most appropriate way to develop a SUSDEV21 landscape indicator.

2.2.23  Experience on landscape classification studies in the UK has shown that it is preferable to give descriptive names to LCTs and LCAs rather than simply giving them code numbers. LCTs normally have names associated with topographical, vegetative and/or land use characteristics, while LCAs have names that are geographically specific. It is proposed that a combination of code numbers and descriptive names for LCTs could be used.

2.2.24  The nature of the Study as an SAR-wide one means that acquiring and preparing comprehensive data is a key issue both in securing full geographic coverage, and in ensuring that sufficient range of different expertise is available from different fields to cover all the elements which influence landscape development (e.g. ecology, culture/heritage, town planning etc). Some of the major GIS data sets initially identified for used in the Study that are currently available in digital format are:

• aerial photography (in addition to PlanD's aerial photo mosaic, a new 1:10,000 orthophoto product is available from LandsD);

• satellite imagery;

• Digital Terrain Model - DTM;

• digital geological maps;

• land cover data available from the mapping produced for the SUSDEV21 study;

• cultural and heritage data including historic buildings and monuments;

• building topography, including building polygon, podium line, and building heights; and

• hydrological data.

2.2.25  A key issue is the ability to collect the above data rapidly and cost-effectively and to meet the established data quality criteria. Such criteria require that the data are complete, consistent, up-to-date, easily updateable (with newer releases of data), efficiently stored (not taking up excessive disk storage space), well organised and structured (so that the data can be easily and quickly retrieved), and meet the required level of accuracy.

2.2.27  One of the critical parts of the project is to check and refine where necessary the LCA and LCT boundaries by carrying out field surveying. It will be crucial that the information gathered is collected in a consistent, accurate and comprehensive manner. While the survey design will determine what information is to be collected, mechanisms should be put in place to validate the data at point of collection. An inexpensive mobile hand-held PC GIS system will be used to do this (see section 2.4 and figure 5).

2.2.28  The Study Brief has rightly emphasised the importance of system and GIS requirements in the whole Study. GIS capabilities are required to take-on the very important task of analysis and clear presentation of the Landscape Character Map and related information. After completion of the project, the system will very likely be used for producing various other outputs and as such needs to be easy to use, robust and flexible.

2.2.29  Preparing data and developing indicators that are consistent with the SUSDEV21 and CASET systems will be a key task in the Study. ERM set up the CASET system as part of SUSDEV21. ERM's experience on SUSDEV21 will be helpful in determining a suitable indicator for integration with CASET (Figures 8 and 9 illustrate the CASET interface).

2.2.30  A key issue of the Study is that the product should not simply be an end-state one, but that the Landscape Character Map and GIS database should be capable of periodic updating and that indicators should continue to be able to illustrate landscape change.

2.3.1    Due to the size of the LCT's which the Study Brief requires (5 hectares, which equates to approximately 21,900 for the total area of Hong Kong), this Study will be one of the largest landscape character mapping exercise ever carried out anywhere in the world. Additionally, a large-scale landscape character map has never been created in Hong Kong. To complete the whole Study within 18 months (excluding public consultation) will only be achievable with a combination of practical innovative ideas, sound methodology and a well-designed work programme executed by a highly experienced and professional Study Team.

2.3.2    The Study Programme means that it is likely that much of the field survey work will be carried out over the summer months of 2002, the wettest period of the year. It is therefore likely that a certain amount of survey days will be lost due the inconvenience and danger of working in such conditions. Urbis has developed a field survey programme which will allow for such down-time and still meet the overall project programme as well as other Study criteria.

2.3.3    There is a limited resource of professionals familiar with current best practice in the field of landscape mapping, classification and evaluation in Hong Kong. It is therefore important that the Study Team brings in International Specialist Advisors in Landscape Value Mapping Methodologies to advise them.

2.3.4    The special requirement for public consultations as part of the Study raises two issues that must be addressed if results of consultations are to be incorporated meaningfully into the process. The first is to ensure adequate communication of potentially complex findings and data simply and clearly so that the full implications can be understood by non-experts. The second is to find mechanisms by which public responses can be organised in a structured way and incorporated meaningfully into the Study. Urbis will use a Structured Public Consultation Questionnaire to assist in ensuring that public input is received in a structured and manageable way.

2.3.5    Where official data are available, it must be recognised that some data are more up-to-date than others. For example, the Lands Department's Digital Topographic Map Database B5,000 series contains more than 192 data sheets. At any one time, the up-to-date quality of the data sheets ranges from a few months to two years. The data sheets therefore have different degree of up-to-date quality, though, as a general principle, the less frequently updated areas usually have less development activity during the period.

2.3.6    The Study requires dissemination of landscape character map information using CD-ROMs. CD-ROM is the most common mass storage medium with a maximum storage capacity of about 650 MB. Because of the expected huge volume of map information, the data must be properly organised and presented. The map data on the CD-ROMs must be easy to use, interactive, efficient (short loading time), and be prepared cost-effectively.

2.4.1    In order to ensure that there is consistency of approach throughout the Study and that fundamental decisions are reviewed by those with experience to assess their validity (e.g. development of landscape classification system and evaluative criteria) a Technical Review Panel has been set-up. The Panel is composed of senior staff from Urbis and ERM, namely Alexander Duggie, David Morkel, Matthew Pryor, Alan Macdonald and Steve Laister, who bring to the project a total of over 50 years experience working in Hong Kong's urban and rural environments. The Technical Review Panel will sit when key project deliverables are to being formulated in order to ensure that fundamental Study precepts are valid.

2.4.2    In order to draw together international experience and local experience rapidly and efficiently, and involve stakeholders directly in the Study, a one-Day Value Management Workshop will be held at the commencement of Task 3. The objective of the workshop will be to draw upon international and local expertise so as to confirm and ratify a coherent project methodology and landscape classification criteria, and to identify initial evaluation criteria for further development later in the Study. The Value Management Workshop will be attend by the Study Team, relevant Government Departments and if the DR wishes, other interested parties.

2.4.3    Past experience of public consultations suggests that retrieving and translating responses received into a meaningful form can often be difficult. For this reason, the use of a Structured Public Consultation Questionnaire issued with public consultation documents will be employed. The Questionnaire will use a pro-forma format to ensure that as far as possible, responses are received in a consistent and comparable format. This will make responses and results of public consultation much easier to input into the Study as a whole and ensure that public views influence the Study appropriately.

2.4.4    The Land Information Centre (LIC) of the Survey and Mapping Office (SMO) of the Lands Department has invested tens of years of time and effort in the preparation of high quality digital map data for use both within the Government and by the public. The more commonly known Digital Topographic Map Database, comprising topographic data series at different scales (e.g. B10,000 series at scale 1:10,000) will form part of the input data for the Landscape Character Map.

2.4.5    More recent and lesser known is the Geo-Reference Database (G1000) available from LIC at a scale of 1:1000. The G1000 series comprises three different products, Building Polygon, Site Polygon and Road Centre Line. These map products will greatly facilitate data preparation which otherwise may become very cumbersome and inefficient in terms of resources and time.

• Building Polygon Series - The Building Polygon series has a number of useful building attributes which are not available in the Digital Topographic Map Database. First and foremost is the building top level (measured from Hong Kong Principal Datum) which is now available for the whole SAR except the North and the Tai Po Districts. It is expected that the building top levels for these remaining areas will also be available within the Study period. As building top levels will need to be considered when examining visibility of various landscapes, data from LIC will obviate the need to obtain building top levels by surveying or by other means. The use of the Building Polygon series will therefore save much time and effort in preparing input data for the desktop study.

2.4.6    As briefly described above, the use of the Building Polygon Database on top of the Digital Topographic Map Database from LIC will save significant time and effort in preparing the input data for the Desktop Study. Such data will also be consistent and reliable, benefiting from quality assurance from LIC. Moreover, the data will be updated on a regular basis by LIC so that the latest data will always be available when it is required to update the landscape maps at a later time.

2.4.7    A DEM is a grid of data points whose attribute values represent the elevation at that point. DEMs are normally made up of regularly spaced grid points resulting in efficient data storage. DEMs can be generated from the topographic data (using contour lines, spot heights, hydrologic data, etc) available from LIC. The DEM will be useful for:

2.4.8    Aerial photographs provide a quick means of identifying features of a particular area, for example, in areas where recent major development is known to have taken place. Aerial photographs are taken regularly (approximately once a year) at different altitudes (from 3,000 to 20,000 feet) and are available from LIC. The use of aerial photographs can be further enhanced by digital photo-grammetric methods resulting in a digital photographic layer, geo-referenced and rectified for geometric and parallax distortion, and super-imposable on other GIS data layers. Use of the aerial photographs in both the desktop study and as a layer within the final GIS database will be of great benefit. Compression techniques now available make it feasible to store all the image data on one CD-ROM.

2.4.9    As mentioned, it will be important during the field survey phase of the project that information be collected in a consistent, comprehensive and accurate fashion. Paper based surveys are often inefficient due to a variety of reasons including incomplete data capture, inconsistent nomenclature when many survey teams are involved and mistakes made while transcribing data during entry at base. These errors could be eliminated by inputting the field data directly into a computer. Current technology would allow the loading of a digital pro-forma, base maps of the proposed Landscape Units and a 'light' version of GIS software (such as ESRI's ArcPAD) onto a hand held Windows CE PC's. This would ensure validation of the data at point of collection, eliminate double entry and ensure completeness. The other advantage of these units is that they can be linked to GPS units quite easily which would greatly assist surveyors in locating where the LCA's/LCTS they are surveying. Figure 5 shows an example of how such a system might look. (Where tall buildings make GPS use impossible, further means of positioning will have to be found, as appropriate).

2.4.10  There are licence-free and low cost software which could be used to view the Landscape Character Map information on CD-ROMs. Most of these software have interactive capabilities and are customisable. For example, LIC has used ArcExplorer from ESRI in its sample data CD-ROM. Alternatively, low run-time cost software such us MapObjects from ESRI, which is highly customisable, can be used.

2.4.11  By using the "Kodak DC290 Global Positioning System Solution Kit" which integrates a handheld GPS with the professional digital camera, an image captured at a field survey point can automatically linked with data from the GPS unit, including longitude, latitude and the precise time and date of capture. Both the GPS and the image information can be directly imported to the GIS database. The GPS information together with other image capturing data is also attached to the image as Metadata, which can be either retrieved or imprinted on the hardcopy.