An official website of the United States government
A goal common to several disciplines within earth and life sciences is to understand how earth processes and abiotic conditions shape the diversification and distribution of species on our planet. To develop a mechanistic and detailed understanding of these relationships across taxonomic-geographic settings should inform a set of boundary conditions that describe the geologic and climatic conditions under which new biodiversity is generated along with the organismal traits (e.g., generation time, dispersal ability) that govern why species vary in their evolutionary responses to the same external influences. However, earth and life sciences each encompass a set of highly complex and sometimes nested relationships. This presents a need for new ways to guide the integration of domain knowledge across these complex systems in a way that can generate new hypotheses, facilitate interdisciplinary collaboration, and shape earth-life theory moving forward. Here, I outline the use of causal structures, which are a set of tools to diagram cause-effect relationships at different levels of detail (specification) that include structural equation meta models (SEMMs), causal diagrams (CDs), and structural equation models (SEMs). I will give examples of how to use SEMMs and CDs to detail earth-life relationships, what we can learn from doing so, and pose a way for how we might quantify these relationships. I hope to demonstrate the usefulness and applicability of thinking about earth-life systems within a causal framework, and speculate about temporal dynamics and the potential for abiotic-to-biotic causal thresholds that may occur over time in different earth-life systems.
more »
« less
Visualization and ecohydrologic models: Opening the box
Abstract Earth system models synthesize the science of interactions amongst multiple biophysical and, increasingly, human processes across a wide range of scales. Ecohydrologic models are a subset of earth system models that focus particularly on the complex interactions between ecosystem processes and the storage and flux of water. Ecohydrologic models often focus at scales where direct observations occur: plots, hillslopes, streams, and watersheds, as well as where land and resource management decisions are implemented. These models complement field‐based and data‐driven science by combining theory, empirical relationships derived from observation and new data to create virtual laboratories. Ecohydrologic models are tools that managers can use to ask “what if” questions and domain scientists can use to explore the implications of new theory or measurements. Recent decades have seen substantial advances in ecohydrologic models, building on both new domain science and advances in software engineering and data availability. The increasing sophistication of ecohydrologic models however, presents a barrier to their widespread use and credibility. Their complexity, often encoding 100s of relationships, means that they are effectively “black boxes,” at least for most users, sometimes even to the teams of researchers that contribute to their design. This opacity complicates the interpretation of model results. For models to effectively advance our understanding of how plants and water interact, we must improve how we visualize not only model outputs, but also the underlying theories that are encoded within the models. In this paper, we outline a framework for increasing the usefulness of ecohydrologic models through better visualization. We outline four complementary approaches, ranging from simple best practices that leverage existing technologies, to ideas that would engage novel software engineering and cutting edge human–computer interface design. Our goal is to open the ecohydrologic model black box in ways that will engage multiple audiences, from novices to model developers, and support learning, new discovery, and environmental problem solving.
more »
« less
- Award ID(s):
- 2012821
- PAR ID:
- 10452353
- Publisher / Repository:
- Wiley Blackwell (John Wiley & Sons)
- Date Published:
- Journal Name:
- Hydrological Processes
- Volume:
- 35
- Issue:
- 1
- ISSN:
- 0885-6087
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
There is increasing awareness of the imperative to accelerate materials discovery, design, development, and deployment. Materials design is essentially a goal-oriented activity that views the material as a complex system of interacting subsystems with models and experiments at multiple scales of materials structure hierarchy. The goal of materials design is effectively to invert quantitative relationships between process path, structure, and materials properties or responses to identify feasible materials. We first briefly discuss challenges in framing process-structure-property relationships for materials and the critical role of quantifying uncertainty and tracking its propagation through analysis and design. A case study exploiting inductive design of ultrahigh-performance concrete is briefly presented. We focus on important recent directions and key scientific challenges regarding the highly collaborative intersections of materials design with systems engineering, uncertainty quantification and management, optimization, and materials data science and informatics, which are essential to fueling continued progress in systems-based materials design.more » « less
-
Engineering projects, such as designing a solar farm that converts solar radiation shined on the Earth into electricity, engage students in addressing real-world challenges by learning and applying geoscience knowledge. To improve their designs, students benefit from frequent and informative feedback as they iterate. However, teacher attention may be limited or inadequate, both during COVID-19 and beyond. We present Aladdin, a web-based computer-aided design (CAD) platform for engineering design with a built-in artificial intelligence teaching assistant (AITA). We also present two curriculum units (Solar Energy Science and Solar Farm Design), where students explore the Sun-Earth relationship and optimize the energy output and yearly profit of a solar farm with the help of the AITA. We tested the software and curriculum units with over 100 students in two Midwestern high schools. Pre- and post-survey data showed improvements in understanding of science concepts and self-efficacy in engineering design. Pre-post analysis of design performance gains reveals that AI helped lower achievers more than higher achievers. Interviews revealed students’ values and preferences when receiving feedback. Our findings suggest that AITAs may be helpful as an additional feedback mechanism for geoscience and engineering education. Future efforts should focus on improving the usability of the software and providing multiple types of feedback to promote inclusive and equitable use of AI in education.more » « less
-
We present our new ideas on taking the first steps towards cultivating synergy between model-driven engineering (MDE), machine learning, and software clones. Specifically, we describe our vision in realizing a cognizant virtual software modeling assistant that uses the latter two to improve software design and MDE. Software engineering has benefited greatly from knowledge-based cognizant source code completion and assistance, but MDE has few and limited analogous capabilities. We outline our research directions by describing our vision for a prototype assistant that provides suggestions to modelers performing model creation or extension in the form of 1) complete models for insertion or guidance, and 2) granular single-step operations. These suggestions are derived by detecting clones of the in-progress model and existing domain, organizational, and exemplar models. We overview our envisioned workflow between modeler and assistant, and, using Simulink as an example, illustrate different manifestations including multiple overlays with percentages and employing variant elements.more » « less
-
Abstract We develop and demonstrate a new interpretable deep learning model specifically designed for image analysis in Earth system science applications. The neural network is designed to be inherently interpretable, rather than explained via post hoc methods. This is achieved by training the network to identify parts of training images that act as prototypes for correctly classifying unseen images. The new network architecture extends the interpretable prototype architecture of a previous study in computer science to incorporate absolute location. This is useful for Earth system science where images are typically the result of physics-based processes, and the information is often geolocated. Although the network is constrained to only learn via similarities to a small number of learned prototypes, it can be trained to exhibit only a minimal reduction in accuracy relative to noninterpretable architectures. We apply the new model to two Earth science use cases: a synthetic dataset that loosely represents atmospheric high and low pressure systems, and atmospheric reanalysis fields to identify the state of tropical convective activity associated with the Madden–Julian oscillation. In both cases, we demonstrate that considering absolute location greatly improves testing accuracies when compared with a location-agnostic method. Furthermore, the network architecture identifies specific historical dates that capture multivariate, prototypical behavior of tropical climate variability. Significance StatementMachine learning models are incredibly powerful predictors but are often opaque “black boxes.” The how-and-why the model makes its predictions is inscrutable—the model is not interpretable. We introduce a new machine learning model specifically designed for image analysis in Earth system science applications. The model is designed to be inherently interpretable and extends previous work in computer science to incorporate location information. This is important because images in Earth system science are typically the result of physics-based processes, and the information is often map based. We demonstrate its use for two Earth science use cases and show that the interpretable network exhibits only a small reduction in accuracy relative to black-box models.more » « less
