Click Here for the Full Description of the Challenge, Submitted Forecasts, and Resources
The Virginia Tech Center for Ecosystem Forecasting (ecoforecast.centers.vt.edu) is excited to announce a new forecasting challenge: Virginia Ecoforecast Reservoir Analysis (vera4cast). The VERA forecasting challenge is an open forecasting challenge designed to compare the predictability of different physical, chemical, and biological variables across two paired reservoirs to identify the fundamental predictability of freshwater ecosystems. The Challenge includes 12 focal freshwater variables to date with over 11 years of historical data. Data are available in near-real time to train and evaluate models, with forecast evaluation and scores updated daily.
We invite you to submit forecasts! Learn more about the challenge and access tutorial videos at https://www.ltreb-reservoirs.org/vera4cast/learn-more.html.
Physical target variables: daily mean water temperature, secchi depth, daily discharge from the inflow stream
Chemical target variables: daily mean fluorescent dissolved organic matter (fDOM), daily mean oxygen, mean daily flux of carbon dioxide and methane from the reservoir surface, daily sampled dissolved methane and carbon dioxide
Biological target variables: daily mean chlorophyll a, binary indicator of bloom occurrence
Additional target variables: Many additional datasets in near-real time are available that can be used for forecasting. See details at: https://www.ltreb-reservoirs.org/vera4cast/targets.html#additional-target-variables
The VERA Forecasting Challenge is one component of the NSF-funded Virginia Reservoirs LTREB program (https://www.ltreb-reservoirs.org/), which is monitoring two connected drinking water supply reservoirs with contrasting dissolved oxygen conditions to broadly advance our understanding of freshwater ecosystem processes.
Please reach out to Cayelan Carey (cayelan@vt.edu) or Quinn Thomas (rqthomas@vt.edu)with questions!