The dynamics of flowing water moving over the landscape is an immensely complex process that plays a crucial role in shaping our environment, both natural and built. Understanding how these processes affect water quality or pose risks to infrastructure are essential for effectively managing our water resources for recreation, drinking water, and ecosystem health. Given projections for population growth and climate change, our environmental systems are expected to experience greater natural hazard disturbances and urbanization. As a result, there is a need to harness the data available from new sensor technologies to the greatest utility. To address this need we develop tools to characterize the current state of environmental systems and monitor how they change over time. In particular, our research has focused on developing and applying computational tools (broadly now considered machine learning methods) to water quality sensor data and the use of advanced geomatics (sensing) technologies for studying landscape erosion processes.
harnessing sensor data for environmental monitoring
Vermont EPSCoR Basin Resilience to Extreme Events (BREE) Project
While characterizing individual hydrological events in watersheds has important applications, our watershed systems are complex socio-ecological systems. This is exhibited clearly through pressing challenges associated with implementing EPA total maximum daily loads (TMDLs) to address critical water quality issues (e.g., Lake Champlain, Chesapeake Bay) and the desire to better manage human practices on the landscape. As part of the Vermont ESPCoR Basin Resilience to Extreme Events (BREE) project, we are working to integrate learning from high-frequency water quality sensor data into an integrated socioecological model (i.e., an integrated assessment model [IAM]). The effort is focused on developing tools to automate event analysis to better handle the expected deluge of sensor data expected in the near future. Furthermore, we seek to find opportunities to inform watershed management by identifying linkages between hydrological event types and watershed characteristics (e.g., land use, geology, extent of best management practice adoption, etc.).
An Ecologically Inspired Human-Machine Intelligence Approach to Recognizing Similitude in Multi-Scale Watershed Research
Identifying similarities in watershed characteristics and water quality conditions across the continental U.S. can help identify the causes of environmental change and enables the translation of research from individual studies to larger regions. Because the sheer size and diversity of these long-term monitoring data present challenges for traditional scientific and statistical methods, we will employ artificial intelligence methods along with domain experts in hydrology and ecology in an integrated human-machine learning framework. In so doing, we aim to identify the common environmental variables and parameters that are linked to similarity across scales and investigate trends in short-term and long-term water quality and streamflow data. Organizational frameworks from ecological and biological disciplines will be applied to create a new approach to organizing and summarizing similarities in watershed signals, where watersheds are grouped into guilds based on their similar functional traits. By identifying patterns of similitude in these large data sets and extracting watershed attributes with linkages to these patterns, findings of research on ecosystem processes conducted in individual watershed studies can be more readily translated to larger regions.
Using Big Data approaches to assess ecohydrological resilience across scales
Land-cover transformation, amplification of biogeochemical flows, and climate disruption are triggering transitions in the Earth system that are unprecedented on human timescales. To ensure biosphere integrity and continued human flourishing, we need to understand the factors that determine ecosystem resilience to these diverse disturbances. This project brings together researchers from across the country in a Critical Zone Network, combining data science, ecology, hydrology, and biogeochemistry. Students, researchers, and outreach partners will work in dynamic teams to create new knowledge through field and lab work, and improve education, policy, and participation in STEM fields.
This project is part of the new NSF Critical Zone Collaborative Network
Exploring linkages between suspended sediment‐concentration data and geomorphic process in the Upper Esopus Watershed
Given the significant stream monitoring efforts of the Ashokan Watershed Stream
Management Program (AWSMP), NYC DEP and USGS to characterize erosion sources in the upper Esopus Creek watershed, a key next step is to advance research on data analysis methods useful for informing watershed and stream management efforts. The upper Esopus Creek, and the Stony Clove Creek in particular, experience excessive levels of turbidity and suspended sediment concentration (SSC) necessitating the implementation of sediment and turbidity reduction projects (STRPs) aimed at reducing sediment loads to the Ashokan Reservoir. Effective prioritization of STRPs within the upper Esopus Creek watershed relies on characterizing the spatial variability of SSC dynamics relative to known or suspected source locations and time periods of turbidity. This pilot research project aims to identify linkages between event-scale suspended sediment dynamics and geomorphic process.
advanced geomatics and uas-based sensing for landscape characterization
Testing a responsible innovation approach for integrating precision agriculture (PA) technologies with future farm workers and work
Precision agriculture (PA) includes data-based agricultural technologies and practices that use localized farm data, at the appropriate time and location to increase farm profitability and reduce negative impacts on the environment. The goal of this project is to advance our knowledge of if and how PA tools will disrupt existing social, technological and economic relationships in the agriculture sector. We seek to address the question, how can designing, building, and deploying aerial and ground-based nutrient sensors and AI algorithms help augment farm-level workforce? We propose to use farms as living laboratories for developing PA tools, policies, and workforce training that can enhance farmer trust, farm productivity and environmental sustainability. This NSF funded project includes an interdisciplinary team from UVM and South Dakota State University.
UAS & TLS Bank and Gully Erosion Monitoring
Streambank erosion is estimated to account for 30-80% of sediment loading into waterways. In many cases, this sediment is carrying important pollutants, such as phosphorus. Recent developments in UAS-based photogrammetry provide opportunities for rapidly and economically determining streambank erosion and deposition at variable scales.
This study aims to evaluate the effectiveness and capabilities of UAS to measure streambank erosion. An UAS and terrestrial laser scanner are used to capture fine-scale topographic data of various types of streambanks and qatify the change over time. UVM Spatial Analysis lab contributes to the fieldwork effort of mapping with UAS in the Mad River and Winooski River watersheds in Vermont.
Over-Summer Snow Storage Climate Adaptation
Can snow be stored over the summer at low elevation, mid-latitude sites such as Vermont? We are working with project lead Dr. Paul Bierman to apply terrestrial laser scanning and surveying to monitor the rate at which snow storage piles settle and melt over the summer months. Results of this project aim to inform strategies for nordic ski areas to adapt to a warming climate.
For more information on the project, please visit the project website: https://www.uvm.edu/~snowstor/
past research projects
Turbidity-based monitoring of Suspended Sediment Loads in the Mad River Watershed
When Tropical Storm Irene impacted the state of Vermont in 2011, sediment (and associated nutrient) loading into Lake Champlain was highly visible and stakeholders were uncertain as to the legacy impact of Irene on watershed sediment dynamics. To address this need, we deployed the first multiple station network of high-frequency in-stream turbidity sensors in Vermont. Six locations along the Mad River and its tributaries were monitored between 2012 and 2016 and provided data for 600+ individual storm (hydrological) events. This research was conducted in coordination with the Vermont EPSCoR Research on Adaptation to Climate Change (RACC) project
Building resilience through community-based action research: Identifying vulnerabilities and facilitating change in rural mobile home parks
Project funded by USDA Rural Development Grant that studied Vermont mobile home parks and their vulnerabilities to natural hazards led by Dan Baker, Kelly Hamshaw, and CVOEO. Tropical Storm Irene highlighted the vulnerabilities of mobile home parks when 154 homes were destroyed. It also highlighted the disproportional impact that mobile home parks can bear in disasters. Building a comprehensive understanding of the state of mobile home parks in Vermont and their vulnerabilities was needed.
As the project GIS consultant and spatial analyst, developed spatial data on the location of mobile home parks in Vermont and analyzed the vulnerability of parks to natural hazards such as flooding and fluvial erosion.
For more information on the project please visit the research project website: