All Funded Projects

We will develop a vulnerability assessment R Shiny web application and present to stakeholders. The stakeholder feedback will be summarized into a one page ‘lessons learned’ document that will assist researchers in designing effective climate change visualizations and an R markdown ‘quick start’ guide on R Shiny applications.

This project develops an open-source toolkit for the consistent, automated assessment of accuracy and cartographic quality of generalized geospatial data. The toolkit will aid USGS and other stakeholders with the development and use of multiscale data and with associated decision-making.

Restoring ecosystems using native plant materials is a critical pursuit of federal land management agencies following natural disasters and disturbances. The Native Plant Seed Mapping Toolkit provides practitioners with quantitative data to support successful restoration outcomes.

The value of USGS tools and products can be assessed by collecting use metrics, user feedback, and examples of practical application. We will pilot an approach to assess the utility of two Coastal Change Hazards product releases and establish a guide for tracking the use and user experience of USGS products.

Develop a process-guided deep learning modeling framework to integrate high-frequency streamflow data from gages, discrete streamflow measurements, surface water presence/absence observations, and streamflow model outputs to improve hydrological predictions on small streams.

This project will share techniques developed in two AI/ML competitions run in Fall 2022, Automated Map Georeferencing, and Automated Map Feature Extraction with USGS stakeholders. We will develop a strategy to operationalize successful approaches, benefiting any activity that uses legacy map data.

This project will combine the 117,000 orphaned oil and gas wells in the USGS Orphaned Well Dataset with groundwater quality data from the USGS National Water Information System (NWIS) to create a data product that can be used to analyze the interactions between orphaned wells, groundwater, and hazards to the environment.

We will make cutting-edge spectral analysis and machine learning algorithms available to remote sensing and chemical quantification communities, regardless of the user’s programming skills, by releasing, documenting, presenting, and developing tutorials for the Python Hyperspectral Analysis Tool.

We aim to develop a web-based game where players use human-assisted image segmentation to produce annotated “meditation drawing” images of surface water sites to accelerate the creation of machine learning imagery training datasets. The game will also public education and outreach opportunities.

We will develop and publish a stand-alone Python script to produce high-frequency and high-spatial resolution coastal vegetation maps that leverage new Planet 8-band 3m images, USGS CoNED topo-bathy DEMs, and 3DEP Height Above Ground data. These products will help improve forecasts of hurricane impacts.

In practice, e.g., in model applications, data are rarely complete without workflow code and workflows are often treated as software that include data. This project aims to understand current practice and recommend future practices that better fit the needs of reproducible workflows and models.

We will collate and publish existing datasets from collaborators and ingest them into a visualization app to help researchers with machine learning model-building and hypothesis-making. These data collation and app development methods could help other researchers increase their data accessibility.