Social Simulation of Fisheries and Coastal Management workshop
Attila Lazar (University of Southampton, UK) was invited as a keynote speaker for the Social Simulation of Fisheries and Coastal Management workshop (Manchester Metropolitan University, 6-7 June 2016) organised by the Centre for Policy Modelling, the Stockholm Environment Institute (Oxford Centre) and the University of Tromsø – the Arctic University of Norway. The aim of the workshop was (i) to further establish the academic-developer-practitioner network, (ii) to improve the use of data and methods, and (iii) to work towards the integration of tools and methods and creation of decision support frameworks.
Attila presented the philosophy, methods and results of the ESPA Deltas project with a special focus on the ΔDIEM assessment framework and the plausible futures of coastal Bangladesh. He also introduced the DECCMA project and its planned integration activities.
There were many interesting presentations and discussions. For example, Anthony Charles (St. Mary's University, Canada) talked about the importance of multi-level considerations in systems analysis: different scales, objectives and believes exist for different stakeholders that makes the modelling and engagement processes even more complex. John Theodoru (TEI, Greece) discussed the observed human response to large scale fish-kills in Greece. He stressed the importance of the communication between officials/expert and the general public to avoid large-scale and long-term impact on livelihoods and protests. Valker Grimm (UFZ, Germany), the 'father' of agent-based modelling, showed examples for theory development and the usefulness of ABMs to study ecological and human systems, and stressed the importance of proper documentation and multi-objective fitting and 'validation' of all models. Mike Bithel (Cambridge University, UK) outlined the NERESUS program, a 10-year research program to further develop a global individual-based ecosystem model called Mandingley. Steve Shaul (ASU, Florida) presented his (multi-objective) efforts to quantify the uncertainties and estimate uncertain human behaviour-related model parameters for his Bay of Mexico fisheries ABM, and thus learn about the fisher-fish behaviour and their consequence on stock assessment. Richard Taylor (SEI, Oxford, UK) the usefulness of 'local' fisheries toy-models in stakeholder engagement and demonstrated the 'Shallow Seas' participatory game, a coastal fisheries model, that allows 4-6 players to interact in an online game environment, and learn about the importance of social-interaction on livelihood success and ecosystem exploitation. The workshop was closed with a generic discussion. Three key topics were raised by the participants:
(i) It is not straightforward to couple qualitative survey data with quantitative models, due to contextual bias (i.e. answers relate to the context set by the interviewer). This makes its generic model-use difficult. Furthermore, adaptation to the new circumstances are continuous, and this makes future predictions uncertain if only the survey data is considered in the model.
(ii) It often thought that if the model works reliably (i.e. reproducing observed patterns) that is the end of the model development and testing. However, models should be pushed to make wrong predictions in order to learn about tipping points that would result in a shift in dominant processed, and thus, learn about the model performance under extreme conditions and more about the system behaviour.
(iii) Under ideal situation, large, complex, expert simulation models, such as ΔDIEM (to explore realistic, plausible futures) should be used in conjunction with toy models, with which experts, stakeholders and the general public can interact and learn.
To find out more about the workshop, click here
Posted 08/06/2016 13:48 by Attila Lazar« Back