The way in which Strategic Environmental Assessment (SEA) succeeds in its key aim – to integrate the environment into strategic decision-making – is affected by the choice of both data and scale. The data and scale used within SEA fundamentally shape the process. However, in the past, these issues were often not discussed in an explicit or in-depth way. This article proposes a research agenda, and recommendations for future practice, on data and scale issues in SEA. Future research on data issues, spatial and temporal scales (both in terms of detail and extent), tiering, data quality and links to decision-making are recommended. The article concludes that questions of data and scale in SEA are not just technical, they are essential to identifying and understanding the issues that SEA should be addressing, and therefore are a core element of SEA.
This article is about data and scale issues in Strategic Environmental Assessment (SEA). This topic goes to the heart of the SEA process in more ways than one. From a pragmatic point of view, data is essential as it can be argued that data is needed in order to carry out an SEA. From a quality perspective, the choice of ‘appropriate’ data and respective scale (both temporal and spatial) are also key, as different data (or scale) might produce different outcomes to the SEA process. In more philosophical terms, data issues are also fundamental in terms of what the SEA process is about. Namely, should issues identified by the SEA process work as a ‘cookie-cutter’ to select the data needed or should data collection precede the identification of key issues? One overall-arching theme discussed in this article is what should come first — data or issues.
The article starts by discussing the links between SEA and decision-making (Section 2). Ultimately, which data (and which scale) is chosen, depends not only on the needs of the SEA process but also on the requirements of the decision-making that the SEA is informing. SEA data issues are discussed next (Section 3), taking into account that SEA is composed of different topics (e.g. biodiversity, health, water) and can be applied to different sectors (e.g. transport, forestry) — all with many different data and scale needs. Section 4 proposes a research agenda related to scale issues in terms of scale effects, spatial and temporal scales, and choice of both the amount of detail and the extent of the study. Tiering versus multi-scale analysis is discussed in Section 5, in terms of scale choice for each tier and also in terms of the need to investigate the interaction or dialogue between those tiers. The study of data quality, metadata and uncertainty issues is fundamental considering the importance of generating results that decision-makers and the public can trust. This is covered in Section 6. Section 7 discusses the importance of researching the possibility of data and/or scale abuse. This is the rare occasion when data and/or scale might be chosen by corrupt practitioners or politicians as a way to get a preferred result. Finally, Section 8 concludes the article.
This article outlined a research agenda for data and scale issues in SEA. Data and scale in SEA are not just technical matters but are fundamental for identifying and understanding the issues that SEA should be addressing (e.g. health, accessibility, biodiversity). In other words, data and scale help define what SEA is about (and for what purpose) and set it apart from project EIA. As an overall conclusion, and as a recommendation for future practice, this article proposes that issues (defined for example by SEA objectives) should normally come before data collection. This way it can be ensured that issues considered important (e.g. reduce air pollution, improve people's health, maintain biodiversity) are selected for further assessment (irrespective of the existence or not of data) while at the same time avoiding unnecessary data collection. That said, the selection of issues can be affected by iterative elements, such as data collected (see Fig. 1) and what is found at other stages or other parts of the assessment.
However, this might not be what is happening in practice. Organisations attempting SEA for the first time might be spending a lot of time and money collecting data before they start to do their SEA. The risk is that data is collected that is not necessary or that data needed for certain issues is neglected. It is also more difficult to identify data gaps if issues are not identified first. In order to help SEA practitioners shy away from collecting too much data in SEA (and remaining ‘strategic’ in its true sense), Partidário (2007-this issue) argued that the terminology of ‘baseline data’ was not very useful and should be replaced by a new term (e.g. context).
The reality is that not all issues need be dealt at the same time or with the same level of detail. Therefore, data collection could probably be incremental and spread through the SEA process (linked for example to some form of adaptive environmental management — see Morrison-Saunders et al., 2004). It is also crucial that SEA keeps up with the timing and the needs of the decision-making process. This might be incompatible with large and time-consuming data collections. Perhaps a ‘just in time’ SEA is needed that can work more effectively within political arenas.
Data (and scale choice) are of course not politically neutral, with different stakeholders and different sectors having preferences for different scales (see Karstens et al., 2007-this issue). This links up to the concept of political ecologies of scale – how different hierarchical levels of vested interests (e.g. local, regional, national, global) determine the management of key resources – see Natter and Zierhofer (2002). Scale choice affects the problem addressed, the options found, and the impacts evaluated. The study of scale effects in SEA – how scale choice might change the identification of the type and importance of SEA issues, and how it might affect the results of SEA and, ultimately, the outcomes of the decision-making process – is critical. It is therefore always necessary to clearly specify what scale and what data are used and why. Fundamentally, what is needed in the future, is a careful, transparent and accountable choice of scale.
