2024 USGS Risk Research Projects

Molly Blakowski, Utah Water Science Center 

Evaluating potential health impacts of Great Salt Lake dust events to support decision-making about water conservation and dust mitigation 

Dust produced from the dry lakebeds of desiccating saline lakes is a documented human health concern. Now that Great Salt Lake (GSL) has reached a new historic low, water managers, public health experts, and policymakers in Utah need to understand the specific health risks that low GSL levels pose to nearby communities. The USGS Utah Water Science Center (UTWSC) dust team, in cooperation with the Utah Department of Health and Human Services (UDHHS), Utah State University (USU), and the University of Utah (U of U), propose to co-produce a retrospective dust health study examining temporal associations between GSL lakebed exposure, dust event occurrence, and emergency department (ED) visits for respiratory and cardiovascular conditions. Although GSL water-surface-elevation data and regional ED visit data are readily available, particle pollution (PP) monitors in this area have limited records and none are close enough to the lakebed to isolate impacts from GSL dust events. To address this data gap, we will reconstruct a spatially explicit timeseries of GSL dust event occurrence in northern UT using wind speed, weather codes, and visibility data from regional airports. We will relate the timeseries to nearby ED visit data using time-lagged regression models to evaluate health risks. We will explore connections between lake levels and dust event occurrence by regressing the number of GSL dust events per year against average annual lakebed area. Results from this work will help decision-makers evaluate health costs of changing lake levels and support data-driven decision-making about GSL water management and dust mitigation. 

Evan Grant, Eastern Ecological Science Center 

Optimizing management actions to mitigate drought risk 

Optimizing the number and location for management throughout a species’ range is challenging, given differences in species abundances, environmental characteristics, and spatially varying sensitivity to climate change and other drivers. This combination of spatial variation results in non-linear benefit accumulation from management actions. Thus, a simple ranking of relative expected improvement of each site under management results in a sub-optimal allocation of resources. To evaluate the optimal allocation of drought mitigation management resources across a species’ range, we will use the Cascades frog (Rana cascadae) as a case study. We will use an integrated species distribution model, which incorporates multiple species datasets and environmental covariates, to estimate the current intensity of R. cascadae. We can then intersect the species intensity with a drought map and evaluate the change in species by varying the number and position of management actions in the region. We will use return on investment curves to evaluate the optimal set of actions to maximize the net benefit to R. cascadae. Our results will be presented in a report and associated interactive application so that combinations of management action locations can be evaluated by the decision makers. Our approach can then be scaled to other species or other geographic regions. 

Riley Mummah, Eastern Ecological Science Center 

Incorporating nested uncertainty into management decision making 

Models are commonly used to evaluate the current or future state of a system and guide management decision making. However, there is often uncertainty about which model is correct or which management action to take. In these instances, value of information (VoI) analysis can be used to evaluate model outputs across management options to identify which source of uncertainty is the most valuable to resolve. Uncertainty in decision making can arise from discrete hypotheses expressed as alternative models, but parametric uncertainty (about the system dynamics as well as the effects of management actions) is nested within the models and is typically ignored. Our aim is to develop a tool to evaluate nested sources of uncertainty on the decision making process, using Shenandoah salamander (Plethodon shenandoah) as a case study. Shenandoah National Park is currently weighing eleven management actions for the conservation of the Shenandoah salamander. To date, we have developed six alternative models to explain P. shenandoah occupancy, which each contain 12-19 parameters. We will develop a procedure to simultaneously evaluate parametric uncertainty, model uncertainty, and management uncertainty in an Expected Value of Partial Information (EVXI) approach. We will thus be able to identify the relative contributions of each uncertainty source, which will provide a quantification of the importance of all the uncertainties to the management decisions of the National Park Service within Shenandoah National Park. 

Michala Phillips, Pacific Island Ecosystems Research Center 

Preparing for Ecological Emergencies: A Framework for Pre-Crisis Data Collection in a Changing World 

The rapid pace of global change necessitates that federal scientists work to address ecological emergencies (i.e., unpredictable natural hazards, ecological invasions, disease, and extreme climate events). In the past year, crises like Typhoon Mawar, wildfires in Hawaiʻi, and quagga mussels in the Snake River prompted urgent responses from USGS scientists. Effective responses often depend on the quality and availability of data collected before the emergency occurs. Data collection before an ecological emergency, aids in both learning and action for swift decision-making in crises. Post-disaster success often hinges on access to pre-disaster data, stressing the pivotal role of strategic pre-collection. The time scale of an emergency (i.e., immediate vs. long-term) influences data needs – USGS scientists had just days to gather baseline data for the Snake River mussel crisis, while monitoring for wildfire preparedness involves years of data collection and analysis. Well-informed prioritization is key: resources for pre-emergency data collection are often limited and must be strategically deployed, supporting societal needs collaboratively with partners and communities. To address these challenges, we propose to convene a working group to create a framework for co-development of pre-crisis data collection plans in the Ecosystems and Water Mission Areas. This collaborative effort aims to unify diverse perspectives; craft a typology of ecological emergencies; and delineate data needs from long-term preparedness to rapid response. We will create a flexible decision support tool that USGS scientists and partners can use to co-create proactive pre-emergency plans for data collection, contributing to effective monitoring and preparedness across disciplines.