Building a reliable evidence base: Legal challenges in environmental decision-making call for a more rigorous adoption of best practices in environmental modelling

Environmental decision-making frequently relies on predictive mathematical modelling as an evidence base. Where legal provisions exist to challenge these decisions, models can be particularly vulnerable to such challenges. Defensible models often adhere to rigorous scientific method and the adoption of best modelling practice that ultimately facilitates a more reliable evidence base. While developing and adopting good practices and standards in mathematical modelling are crucial aspects for models, and an active area of research in many environmental disciplines, the implications of not adopting best modelling practice for environmental decision-makers remain largely implicit in the literature. The objective of this study was thus to explicitly identify and document the challenges associated with the use of predictive modelling in the environmental and resource management decision-making process through a systematic review of New Zealand legal decisions. The review produced a rich range of examples (68 legal decisions) where modelling evidence was challenged in legal proceedings and enabled the identification of modelling characteristics and practices which strengthen modelling reliability. All of the legal challenges were substantive, relating to the scientific components of the model (e.g. assumptions, input data, and parameters), model evaluation or application. None of the challenges were regulatory process challenges. There are numerous publications that describe best practice for modelling from a technical perspective, but it appears that these guidelines are not always being followed. If models are to be of substantial help in environmental decision-making then modellers and decision-makers will need to ensure that there is a clear understanding of the purpose of a model, the modelling process is transparent, limitations are acknowledged and considered, and that best practice guidelines are followed.