APA. Samuel Pottinger

A. Samuel Pottinger

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


A. Samuel Pottinger has a total of 30 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Removing barriers to science-informed decision-making through data science and human-centered design

    by Lucia Sophie Layritz, Maya Zomer, Magali le Bruyn, Sam Pottinger, Nick Gondek, McKalee Steen, Maya Weltman-Fahs, Ciera Martinez, Kevin Koy, Fernando Pérez, et al.

    Having policy and regulation be based on the best available scientific evidence is a widely accepted goal, yet relevant knowledge often fails to reach those most in need due to gaps in data accessibility and technological barriers. We share our experience developing tangible, scalable tools that support policymakers, indigenous groups, and land managers in bringing science to the table when decisions are being made. These projects are united by common principles of participatory user-centered design, digital sovereignty, open-source software development, modern data science, and scientific integrity.Specifically, we present three case studies across different governance scales: 1) At the international level, we discuss interactive decision support tools to facilitate science-based policymaking in the United Nations Montreal Protocol and Global Plastics Treaty. 2) At the national level, we present a stochastic, open-source simulation platform built in collaboration with the U.S. National Park Service. It enables land managers to model vegetation resilience and evaluate post-fire management scenarios under diverse future climate projections. 3) At the local level, we highlight custom-built, co-developed software to monitor cases of Indigenous land return alongside a biodiversity monitoring application for improved land management decision-making by Indigenous communities.Across these projects, we will discuss lessons learned regarding the challenges of working with partners in highly interdisciplinary environments, how open science principles can be used to support community sovereignty instead of clashing with it, and the creation of resilient services that survive long-term regardless of infrastructure constraints or organizational change.

  • Open Access

    Integrating future climate projections into post-fire management: A stochastic decision-support toolbox for adaptation in arid ecosystems

    by Lucia Sophie Layritz, Maya Zomer, Nick Graver, Nick Gondek, Amanda Anderson-You, Sam Pottinger, Maya Weltman-Fahs, Carl Boettiger

    Wildfire is a multi-dimensional hazard, impacting both human livelihoods and ecosystem function. Beyond wildfire prediction and containment, post-fire reconstruction is a major management challenge. With shifting and novel fire regimes, post-fire recovery represents a complex risk-management challenge where decisions made under high uncertainty have long-term implications for systemic resilience.There is an urgent need for tools which allow land managers to explore their options in an accessible, systematic and transparent way.Here, we present a joint effort between the Schmidt Center for Data Science and Environment and the U.S. National Parks Service to design a decision-support platform, enabling park managers to create future management scenarios based on current understanding of climate futures to guide their decision making. Using the Mojave Desert ecosystem in Southern California as a case study, we discuss our collaborative co-design process, technical infrastructure and scientific reasoning in translating high-performance vegetation modeling into actionable policy insightsMore specifically, we present josh, an open-source, domain-specific scripting language linked to a high-performance simulation engine. We illustrate how josh can be used to design vegetation models and management intervention for a range of ecosystems, integrate different high-resolution future climate projections and quantify risk and uncertainties through running large, stochastic ensemble simulations. The platform is freely available, open-source, and runs in any web browser, as well as on distributed computing systems; providing a transparent and accountable tool for evidence-based adaptation planning.

  • Open Access

    Climate-driven doubling of U.S. maize loss probability: Interactive simulation with neural network Monte Carlo

    by A. Pottinger, Lawson Connor, Brookie Guzder-Williams, Maya Weltman-Fahs, Nick Gondek, Timothy Bowles

    Climate change not only threatens agricultural producers but also strains related public agencies and financial institutions. These important food system actors include government entities tasked with insuring grower livelihoods and supporting response to continued global warming. We examine future risk within the U.S. Corn Belt geographic region for one such crucial institution: the U.S. Federal Crop Insurance Program. Specifically, we predict the impacts of climate-driven crop loss at a policy-salient "risk unit" scale. Built through our presented neural network Monte Carlo method, simulations anticipate both more frequent and more severe losses that would result in a costly doubling in the annual probability of maize Yield Protection insurance claims at mid-century. We also provide a configurable open source pipeline and interactive visualization tools to further explore these results. Altogether, we fill an important gap in current understanding for climate adaptation by bridging existing historic yield estimation and climate projection to predict crop loss metrics at policy-relevant granularity.

  • Top Cited Papers

    Pathways to reduce global plastic waste mismanagement and greenhouse gas emissions by 2050

    by A. Samuel Pottinger, Roland Geyer, Nivedita Biyani, Ciera C. Martinez, Neil Nathan, Molly R. Morse, Chao Liu, Shanying Hu, Magali de Bruyn, Carl Boettiger, et al.

    Plastic production and plastic pollution have a negative effect on our environment, environmental justice, and climate change. Using detailed global and regional plastics datasets coupled with socioeconomic data, we employ machine learning to predict that, without intervention, annual mismanaged plastic waste will nearly double to 121 million metric tonnes (Mt) [100 to 139 Mt 95% confidence interval] by 2050. Annual greenhouse gas emissions from the plastic system are projected to grow by 37% to 3.35 billion tonnes CO 2 equivalent (3.09 to 3.54) over the same period. The United Nations plastic pollution treaty presents an opportunity to reshape these outcomes. We simulate eight candidate treaty policies and find that just four could together reduce mismanaged plastic waste by 91% (86 to 98%) and gross plastic–related greenhouse gas emissions by one-third.


  • Open Access

    Pyafscgap.org: Open source multi-modal Python-based tools for NOAA AFSC RACE GAP

    by A Samuel Pottinger, Giulia Zarpellon

    Program produces longitudinal catch data which support ocean health


  • Open Access

    Machine Learning Techniques for Detecting Identifying Linguistic Patterns in News Media

    by A Samuel Pottinger

    An article's tone and framing not only influence an audience's perception of a story but may also reveal attributes of author identity and bias. Building upon prior media, psychological, and machine learning research, this neural network-based system detects those writing characteristics in ten news agencies' reporting, discovering patterns that, intentional or not, may reveal an agency's topical perspectives or common contextualization patterns. Specifically, learning linguistic markers of different organizations through a newly released open database, this probabilistic classifier predicts an article's publishing agency with 74% hidden test set accuracy given only a short snippet of text. The resulting model demonstrates how unintentional 'filter bubbles' can emerge in machine learning systems and, by comparing agencies' patterns and highlighting outlets' prototypical articles through an open source exemplar search engine, this paper offers new insight into news media bias.

  • Open Access

    Machine Learning Techniques for Detecting Identifying Linguistic Patterns in the News Media v3

    by A S Pottinger

    Proposing and comparing multiple neural network-based techniques in order to choose an optimized configuration, the described method can train a classifier to predict the publisher of news articles given a short snippet of text. This classifier's scores can then help in creation of a exemplar-based news search engine. Specifically, this model can be used to compare linguistic patterns between different news organizations and to identify prototypical articles for a news source that most "exemplify" a publisher’s coverage. Finally, these "exemplars" are made searchable through an open source browser-based application.

  • Open Access

    Machine Learning Techniques for Detecting Identifying Linguistic Patterns in the News Media v1

    by A S Pottinger

    Proposing and comparing multiple neural network-based techniques in order to choose an optimized configuration, the described method can train a classifier to predict the publisher of news articles given a short snippet of text. This classifier's scores can then help in creation of a exemplar-based news search engine. Specifically, this model can be used to compare linguistic patterns between different news organizations and to identify prototypical articles for a news source that most "exemplify" a publisher’s coverage. Finally, these "exemplars" are made searchable through an open source browser-based application.

  • Open Access

    Machine Learning Techniques for Detecting Identifying Linguistic Patterns in the News Media v2

    by A S Pottinger

    Proposing and comparing multiple neural network-based techniques in order to choose an optimized configuration, the described method can train a classifier to predict the publisher of news articles given a short snippet of text. This classifier's scores can then help in creation of a exemplar-based news search engine. Specifically, this model can be used to compare linguistic patterns between different news organizations and to identify prototypical articles for a news source that most "exemplify" a publisher’s coverage. Finally, these "exemplars" are made searchable through an open source browser-based application.