Flow Photo Explorer (FPE) Project in New Jersey
The Flow Photo Explorer (FPE) is a database, machine learning, and data visualization platform using timelapse images to develop new approaches for collecting hydrologic data in streams and other waterbodies where traditional monitoring methods are not feasible or cost-prohibitive. The Musconetcong River was chosen to participate in the early study, and we are expanding the work to other rivers.
In the spring of 2023, in conjunction with the ongoing FPE project, ten trail cameras were deployed within the Musconetcong basin. The FPE project is a collaboration between U.S. Geological Survey, U.S. Environmental Protection Agency, Walker Environmental Research, Microsoft Research, and other contributing partners, which aims to develop new and innovative approaches to hydrologic monitoring in streams and rivers where flow data are historically sparse or non-existent. This project seeks to use time-lapse images captured by inexpensive trail cameras to train the FPE models to better identify flow conditions at low flow or dry conditions.
Flow is a critical variable in streams since it affects aquatic and riparian biological communities and human uses of water (i.e., recreation, public water supply, etc.). Flow regimes are changing due to anthropogenic (e.g., water withdrawals) and natural impacts (e.g., extreme weather events). Many states consider issues with altered hydrology as a growing priority, and some have developed flow criteria for streams. However, most USGS gages are not located in small to mid-size streams and deployment of hydrological equipment is costly and requires specialized expertise. This project was created to help states, tribes, and other entities by promoting a low-cost and user-friendly alternative with the use of continuous imagery for flow monitoring.
This project has two main objectives:
- Create a web-based data portal for uploading, storing, and exploring streamflow photos and data.
- Develop machine learning models to estimate flow (or stage) directly from timelapse imagery using the photos and data collected through this portal.
Together the web-based data portal and machine learning models will provide an exciting new platform for monitoring flow in rivers and streams based on low-cost time-lapse imagery.
Due to the preexisting intensive monitoring in the Musconetcong basin with air and water temperature sensors, continuous streamgages, discrete discharge measurements, as well as other networks monitored by USGS cooperators, the Musconetcong basin was chosen to participate in this ongoing project. The basin will benefit greatly from more information provided by the trail cameras such as relative flow and to better define groundwater and surface-water interactions. Funding for the trail cameras was provided by NGWOS.
To learn more about FPE or use the Photo Explorer, visit Flow Photo Explorer | USGS
New Jersey Subbasin Paired Air & Stream Water Temperature Networks
The Flow Photo Explorer (FPE) is a database, machine learning, and data visualization platform using timelapse images to develop new approaches for collecting hydrologic data in streams and other waterbodies where traditional monitoring methods are not feasible or cost-prohibitive. The Musconetcong River was chosen to participate in the early study, and we are expanding the work to other rivers.
In the spring of 2023, in conjunction with the ongoing FPE project, ten trail cameras were deployed within the Musconetcong basin. The FPE project is a collaboration between U.S. Geological Survey, U.S. Environmental Protection Agency, Walker Environmental Research, Microsoft Research, and other contributing partners, which aims to develop new and innovative approaches to hydrologic monitoring in streams and rivers where flow data are historically sparse or non-existent. This project seeks to use time-lapse images captured by inexpensive trail cameras to train the FPE models to better identify flow conditions at low flow or dry conditions.
Flow is a critical variable in streams since it affects aquatic and riparian biological communities and human uses of water (i.e., recreation, public water supply, etc.). Flow regimes are changing due to anthropogenic (e.g., water withdrawals) and natural impacts (e.g., extreme weather events). Many states consider issues with altered hydrology as a growing priority, and some have developed flow criteria for streams. However, most USGS gages are not located in small to mid-size streams and deployment of hydrological equipment is costly and requires specialized expertise. This project was created to help states, tribes, and other entities by promoting a low-cost and user-friendly alternative with the use of continuous imagery for flow monitoring.
This project has two main objectives:
- Create a web-based data portal for uploading, storing, and exploring streamflow photos and data.
- Develop machine learning models to estimate flow (or stage) directly from timelapse imagery using the photos and data collected through this portal.
Together the web-based data portal and machine learning models will provide an exciting new platform for monitoring flow in rivers and streams based on low-cost time-lapse imagery.
Due to the preexisting intensive monitoring in the Musconetcong basin with air and water temperature sensors, continuous streamgages, discrete discharge measurements, as well as other networks monitored by USGS cooperators, the Musconetcong basin was chosen to participate in this ongoing project. The basin will benefit greatly from more information provided by the trail cameras such as relative flow and to better define groundwater and surface-water interactions. Funding for the trail cameras was provided by NGWOS.
To learn more about FPE or use the Photo Explorer, visit Flow Photo Explorer | USGS