Search

Abstract When people use shared resources, overextraction can occur. While deliberation tends to mitigate shared resource exploitation problems, the question remains: under what conditions does group chat improve cooperation in shared resource dilemmas? This study analyzes chat and game data from about 1500 rounds of gameplay involving 143 groups across 4 resource types using Sentiment Analysis and Structural Topic Model. We find that, despite their fundamental differences, the 4 games tend to have similar discussions, including strategizing actions, coordinating choices, and socialization, but they differ in which topics explain cooperation within each game. Discussion topics promoting cooperation include coordination in the foraging game (FOR) and long-term goals in the groundwater game (GG). However, discussion topics negatively associated with cooperation include off-topic/socialization in FOR and the irrigation game (IRR) and crop choice affirmation in GG. We suggest that the context in which communication occurs matters, and biophysical characteristics, rules of the game, and levels of uncertainty explain some variations of our findings.

Successful collaborative governance (CG) of social-ecological systems (SES) involves multiple stakeholders convening iteratively over the long term to reach a commonly held vision. This often involves building knowledge for social learning processes induced to come to collective decisions about managing complex systems in flux. Because of the complexity of any SES in the Anthropocene, this coproduced knowledge is frequently transdisciplinary, using a convergence of applied and scientific knowledge from a variety of disciplines and stakeholders outside academia. We find evidence that these cases of effective SES CG involve both knowledge coproduction and convergence research. We evaluated seven case studies of CG across four continents using criteria (principles and methods) developed to facilitate and describe convergence research on SES and found them to be largely present. We also assess these CG cases using indicators of knowledge coproduction, and show that they all involved transdisciplinary knowledge coproduction, which can provide an informative lens for deepening our shared understanding of convergence and its application to complex adaptive systems. All the cases selected for this paper are examples of CG of SES in which research was conducted as part of a collaborative effort to improve the social-ecological conditions in a particular place, and several incorporate various forms of knowledge and ways of knowing. We suggest that these cases demonstrate both convergence research and knowledge coproduction because of the overlap and similarity of these concepts, providing a brief comparison and contrasting of these approaches to addressing sustainability problems collaboratively.

Understanding and managing uncertainty is critical for robust governance. In groundwater management, where collaborative, community-based governance is increasingly common, scientific uncertainty about hydrological conditions could pose challenges to effective and equitable resource management. This study bridges two literatures – collaborative governance and collective action – to examine whether scientific uncertainty about hydrologic conditions undermines the performance of groups that engage in collaborative governance of shared groundwater resources. We conducted a modified groundwater game experiment, based on Meinzen-Dick et al. (2016), where participants engage as resource users in a crop choice game over multiple rounds. But unlike the original game, where participants had full information about recharge rate, two treatments introduced scientific uncertainty in water recharge: uncertainty framed as a range of estimates about groundwater recharge, and uncertainty framed as competing hydrological models predicting different groundwater recharge rates. We also expand on the original game by exploring a wider range of outcomes that include not only sustainable resource use but also group earning and equitable distribution of earnings across players. Analyzing data from 30 group games, our findings suggest that scientific uncertainty can help safeguard shared groundwater resources by prompting users to exercise caution in the face of uncertain recharge rates. This effect was more consistent for the range of estimates treatment than for the competing hydrological models treatment. To unpack the mechanisms behind the experimental result, we also analyzed participants’ communications during the game to understand the strategies that collaborative groups use to cope with uncertainty. In the presence of scientific uncertainty, collaborative processes foster cautious behavior and protect shared resources.

Linux is becoming the de-facto standard operating system for today’s high-performance computing (HPC) systems because it can satisfy the demands of many HPC systems for rich operating system (OS) features. However, owing to features intended for the general-purpose OS, Linux has many OS noise sources such as page faults or thread migrations that can result in the unstable performance of HPC application. Furthermore, in the case of the non-uniform memory access (NUMA) architecture, which has different memory access latencies to local and remote memory nodes, the performance stability of the application can be more exacerbated by the OS noise. In this paper, we address the OS noise caused by Linux in the NUMA architecture and propose a novel performance-stable NUMA management scheme called Stable-NUMA. Stable-NUMA comprises three techniques for improving performance stability: two-level thread clustering, state-based page placement, and selective page profiling. Our proposed Stable-NUMA scheme significantly alleviates OS noise and enhances the local memory access ratio of the NUMA system as compared to Linux. We implemented Stable-NUMA in Linux and experimented with various HPC workloads. The evaluation results demonstrated that Stable-NUMA outperforms Linux with and without its NUMA-aware feature by up to 25% in terms of average performance and 73% in terms of performance stability.