Spatial-based assessment at continental to global scale: case studies in petroleum exploration and ecosystem services

Abstract

Access to natural resources is a problem current and future generations must overcome. To better understand potential changes in access for society due to environmental impacts, global spatio-temporal modeling is required. Spatial data for global assessments have uncertainty, and decision making can be problematic in regards to combining data. The basic problem associated with combining data for global assessments is the data’s heterogeneous nature in terms of source, type, quality, uncertainty, and accuracy. The data uncertainty must be communicated. The overarching question is: how can uncertain information be defined and modelled for global scale studies? This thesis applies geographic information systems for solving spatial questions requiring the combination of several data sets that cover large geographical areas, come from many different data providers, and have varying quality with known and unknown uncertainty. The thesis uses combined spatial data for decision-making related to petroleum exploration and climate change. The first section elucidates the spatial understanding of climate class change on ecosystem services. An ecosystem service assessment valuation composed of 19 criteria is proposed. A global, spatial valuation is calculated by combining empirical data from 23 spatial and nonspatial databases. The global value presented is less than previous studies; however, this study excludes criteria from previous publications and was unable to reliably quantify all criteria. Uncertainty in the input data are described but not quantified. Using five general circulation models and four representative concentration pathways, the Köppen-Geiger climate classification transitions were modeled from 2005 to 2099. The ecosystem service value was evaluated to quantify how much would be in a new climate class. Depending on the representative concentration pathway, between 25-66% of the ecosystem service value will be affected by a climate class change. Uncertainty in the results are presented as a range of values. The second section focuses on using spatial multi-criteria evaluation and error propagation for petroleum exploration. The first part proposes a fuzzy logic multi-criteria evaluation by combining 16 criteria comprised of 26 data sets. The data sets are empirically defined and expert-based. The use of fuzzy logic is meant to mimic the traditional approach by a geologist evaluating the data but provides a framework for consistency, repeatability, and uncertainty evaluation. A case study of northern South America predicts new areas for exploration. Uncertainty is quantified in the second half of the study. After the multi-criteria evaluation is applied, the attribute classification, fuzzy logic memberships, and multi-criteria evaluation calculations are subjected to an error propagation analysis in order to quantify the uncertainty. The error propagation analysis showed that attribute classification using expert-defined uncertainty was the preferred approach over a constant defined uncertainty. Further, attribute classification uncertainty affected the results more than the fuzzy membership uncertainty, which could therefore be omitted; uncertainty related to the multi-criteria evaluation calculations heavily influenced the results and thus require care when selecting. In the concluding remarks, a data and model transparency checklist is provided based on what may be considered appropriate points in a study to quantify or describe the uncertainty in a model.