Project Planning

Objectives

The objectives are to:

1. translate the project Tasks into detailed summaries of the products required from each Task
2. identify the data and modelling requirements of the study
3. record a brief description of each product together with any potential issues and limitations of the output
4. use the descriptions for analysis of inter- and intra-Task products and a comparison of their compatability

Information Products

The planning process is based on description and analysis of the types of information required by individuals and groups; each description is presented as an Information Product (IP) (Appendix 8).

The primary use of an Information Product description is part of this project planning process. A set of information product descriptions collectively define the capacity, functions and data needs (hardware, software, and data) required for the project and indicate options for its establishment and management (operation and maintenance).

The information product description also has a role as a management tool. This benefits individuals and the group. The information product description, provides a documented account of the stages involved in information product generation and thereby should help with directed and efficient interactions. The information product also eases communication by providing a documented account of intended use in a common format.

Each Information Product was written with reference to the contract Tasks, the contractor and with reference to the participating group responsible for the Task. Ten Information Products were compiled describing the work on geographic, photographic and 3D modelling.

Discussion of Issues

The content of the IPs have been reviewed and are described below in terms of issues about computing, data, analysis and presentation of the results. The contents of the IPs reltes only to the materials that wre used within the study of windfarms. Other facilities may have been available within each participating organisation but were not used.

Hardware

Four hardware platforms have been used in the study: Sun workstations, Silicon Graphics workstation and PCs the details of which are noted in Appendix 8. Certain Tasks required specialized computing facilites, but the majority of the work is not hardware dependant. Specifically, the visualization and 3D modelling work carried out in Task 4 required access to a powerful graphics engine of the type provided by Silicon Graphics. The advantages relate to the speed of processing graphical images for effective visualizations and "fly-through" of the landscape.

One further (current) limitation of the hardware/software confirgurations used is that of viewing 3D models within an Internet browser such as Netscape on a Sun workstation. The limitation is a lack of suitable software and the consequence is that viweing the results of the 3D modelling, as presented in the Virtual Reality Modelling Language (VRML) may not be possible on all hardware platforms. Table 1.1 summarises the software and the hardware hosts used in the conductance of the work.

Software

Ten software packages were identified as having been required for the completion of the project: Arc/Info, ArcView, Erdas Imagine, Idrisi, Photoshop, MultiGen, Iris Performer, Iris Showcase and in-house software for topographic exposure and error modelling. In addition, the Internet browser Netscape and several word processing, graphics or statistical packages were employed. Table 1 summarises those packages which are hardware dependant.

An objective of the project was to maximize the use of widely available, usually commercial, software packages to enable the methodologies to be repeated independently. However, the number of packages involved would be an issue with any prospective user of all of the techniques described in this study. There is also an issue of efficiency if there are a large number of packages between which data has to be transferred, with additional potential problems caused by inconsistencies in the handling and referencing of data between each package.

The use of each package reflects both functionality, ease of use and individual preference as well as availability at each participating organization. Therefore, there is scope for a reduction in the number of packages that need have been used although there may be a trade-off with the quality of the output or the efficacy of the production.

Data

A principal employed throughout the project was that the digital data that was used should be from a common source or at least widely available. The data identifed from the Information Products were:

Of the datasets that were identified, the only digital map data created by the project team were the locations of the wind turbines for the Scottish study site and a second scenario of locations for the Welsh site. A model of a wind turbine was also derived for the fly-through model using published plans. Thus, almost all input data was from sources that are open to other potential users and not subject to the uncertainties of in-house derivations.

However, any digital map data is subject to uncertainty and error. Therefore, no inference can be drawn that the results obtained in Task 4 are error free. This topic is specifically tackled in Task 4 under the title of Geographic Error and an analysis of the errors in the elevation models is presented together with examples of the implications of those errors on the visibility of the wind turbines in the Scottish site.

Two scales were identified for exploration, 1:50 000 and 1:10 000. Clearly there are issues of accuracy, resolution, cost and data handling associated with each dataset that will influence the tractability of their use. However, the principal interest in this project was to assess the visual impacts of wind trubines and farms and one aspect of that was to recommend the suitability of different scales of data for the analysis.

The cost of the digital data and its collection is not addressed within this project. Much of the data was obtained under a Service Level Agreement between the Ordnance Survey and the Scottish Office or by special agreement between MLURI and Ordnance Survey. A further analysis of the costs associated with carrying out the analysis using each scale of data and the creation of a 3D model will be required.

Analysis

The nature of the analysis being carried out can be grouped into three parts: geographic analysis, photographic nodelling and 3D modelling. Geographic analysis - Geographic Information Systems (GIS) would be used to build a database comprising topography, land use, road networks, locations of public interest (such as view points) and the locations of existing or proposed wind turbines. The analysis would require: to show the number of turbines that would be visible from any location within a 50km radius; how clearly the turbine would be visible; and hwo much of the skyline they would occupy from certain view points or stretches of road.

However, nothing is ever absolutely certain (different people, of different heights, with different levels of eyesight, will be able to see different numbers of turbines). So, methods which assess the reliability of the analysis have been incorporated, and thus the results are able to be expressed in terms of their predicted reliability. This aspect of the work is primarily related to the data rather than the algorithms being used.

Photographic modelling - view points were identified as being of particular importance for public access or as locations from where the public would have their first clear sight of the wind farm when driving through the area. A range of examples at different distances from the wind farm and photographs taken from each point require to be selected. By linking the output of the GIS to a model of the view captured by the camera photographic simulations of the likely future view of the hillsides with a wind farm will be created. This provides the user with an illustration of the likely future views, to aid in the subjective assessment of the acceptability of the proposed development in a visual context.

The output is subsequently subjected to a psycho-physical model which provides a preference score of the scene from the number and type of features recorded on the photograph. This provides a quantitative estimate to be made of the visual impact of a single turbine or an entire farm with respect to observer preference.

3D modelling - The creation of a model of the terrain of each study site, with a draped cover of satellite data would provide an illustration of the texture and colour of the landscape at a particular time of the year. The user will then be able to "fly" or "drive" through the landscape and gain some impression of how visible the proposed wind farm may be from the main arterial roads, smaller side roads or nearby hill tops.

Outputs

The outputs identified from the Information Products fell into six categories:

1. maps;
2. graphs;
3. tables;
4. 3D models;
5. photographs;
6. HTML documents.

No specific cartographic scales of output are targeted, rather a range of illustrative materials to communicate the nature and extent of the visual impacts of the wind turbines and the potential cumulative impact. One implication of the range of output materials is that no single software package is likely to be the source of all of the outputs. Certain types of output are still relatively specialised (such as "vrml" or equivalent models and other WWW compatable documents) and are thus not supported by Geographic Information Systems software.

However, viewing the range of outputs in their totality certain GIS or image processing packages (such as ArcView 3 or Erdas Imagine) can produce charts, table and maps or images that can be ported directly into desktop publishing or word processing packages, such as those supported by Microsoft Windows '95, from which WWW compatable output can be obtained.

The media required for these outputs would be either paper, computer screen or video monitor. The most effective communication of the impacts will be obtained with the media for which they were originally intended. Therefore, paper copy of either 3D models from a single perspective or WWW documents with animations would severely undermine the quality of the message being communicated. However, there would be consequent cost implications, including financial, legal (copyright) and possibly infrastructural for the presentation of any report which required such a range of outputs. For example, a bound paper report is easily disseminated around a committee but a range of outputs on a CD-ROM would require either a suitable PC or video facility for inspection.