Portfolio
Stephen John Huysman
Wildfire Danger Forecasting System
This project developed a robust, data-intensive pipeline for forecasting wildfire danger. It integrates large-scale meteorological data (gridMET), historical fire records (MTBS), and land cover information (LANDFIRE) to accurately model and map fire ignition risk. Specifically, the system answers the question: Are conditions dry enough to burn? The system employs R for advanced statistical and geospatial analysis, alongside shell scripts for automated data processing and forecasting. Automated forecasts, including lightning strike warnings based on ignition danger, are implemented for several ecoregions across the Western United States. The entire system is containerized with Docker and deployed using cloud-native architecture on AWS.
High-Resolution Water Balance Model
Successful planting of long-lived tree species such as whitebark pine benefits from assessing future climate conditions at the individual tree scale. While coarse gridded climate data are useful for regional screening, they lack the resolution to identify microclimates critical for seedling establishment. To address this, a high resolution 1 m water balance model was developed to downscale regional climate data using local slope, aspect, and soil properties as well as empirically determined local elevational lapse rates to estimate Actual Evapotranspiration and Climatic Water Deficit: measures of growing conditions favorable to plants and drought stress, respectively. This approach can identify microrefugia, which are small-scale topoedaphic features such as northern aspects or high water-capacity soils that can shelter seedlings from drought stress. Although macroclimatic projections indicate that whitebark pine planting sites in the Greater Yellowstone Ecosystem may transition to climate conditions unfavorable for whitebark pine, this fine-scale modeling reveals microsites that maintain favorable water balances which could be overlooked by coarser methods.
White Pine Blister Rust Model
White pine blister rust (WPBR) is the primary driver of the range-wide decline of whitebark pine (WBP), a federally threatened species. As a keystone and foundational species of high-elevation ecosystems, WBP's decline has potentially widespread consequences for forest composition and ecological functions. An understanding of the climatic drivers of WPBR infection is necessary to manage impacts of the pathogen during ongoing climate change. We assembled a long-term dataset indicating WPBR presence or absence in WBP from monitoring programs across WBP's range in the contiguous United States. We identified a spatially explicit model that included August and September temperature and precipitation as the best climatic predictors of WPBR infection in WBP during the basidiospore transmission season, with larger trees more likely to be infected than smaller trees. At high levels of precipitation (around and above 100 mm total August and September precipitation), the relationship between mean August and September temperature and probability of WPBR infection is parabolic, with highest infection rates around 11 °C. This parabolic relationship inverts at lower totals of August and September precipitation (0 to around 100 mm) and minimum infection rates occur around 11 °C and maximum infection rates at low (around 7 °C) or high (around 13 °C) temperatures. Projections of WPBR disease hazard (defined as probability of WPBR infection) through the end of the century show wide variability in geography and magnitude of disease impacts depending on plausible changes in temperature and precipitation across WBP's range.
Landcover Change Model
Climate projections almost universally show increase in temperature and changes in precipitation due to increased radiative forcing from anthropogenic emissions. Climate is a force that fundamentally shapes the distributions of plants. Understanding how plant distributions may shift in response to climate change is an important question with implications across human society. Accurate predictions of changes in plant cover due to climate change are relevant to policy makers, land managers, recreationists, and the general public as shifts in plant distributions could affect food security, industry, and land availability. This analysis identified the climatic drivers of the current vegetation cover types across the Contiguous United States (CONUS). Then, using projections of climate to the end of the century, projections of cover change across CONUS were developed.
