INTEGRATING HIGH-FREQUENCY MONITORING AND FORECASTING TO QUANTIFY THE PREDICTABILITY OF RESERVOIRS

How predictable are reservoir ecosystem dynamics? Reservoirs exhibit high levels of ecosystem variability, which affects the provisioning of water for drinking, hydropower, and irrigation. However, despite the ubiquity of reservoirs in the landscape and their critical role for society, the limits of ecological predictability in reservoirs are unknown. To address this gap, our team has launched VERA, the Virginia Ecoforecast Reservoir Analysis. VERA is an open forecasting challenge to predict physical, chemical, and biological variables at two water supply reservoirs before the data are collected, thereby enabling us to identify ecosystem predictability. VERA is integrated with state-of-the-art ecosystem monitoring led by the Virginia Reservoirs LTREB (Long-Term Research in Environmental Biology) program, which serves as a testbed for developing new approaches for data collection, access, and publishing. To date, >6200 forecasts have been submitted to VERA for 32 reservoir variables, with more forecasts submitted every day. Preliminary analyses demonstrate that physical ecosystem variables (e.g., water temperature) have substantially higher predictability than chemical or biological variables (e.g., dissolved oxygen, chlorophyll). Interestingly, predictability may be highest during periods of hypolimnetic anoxia. At www.ltreb-reservoirs.org, we provide forecast submission templates, a dashboard of updated forecast performance, and additional resources. We invite you to submit forecasts to the VERA Challenge to expand our understanding of reservoir ecosystem predictability!  

Presentation Preference: Oral

Primary Presenter: Cayelan Carey, Virginia Tech (cayelan@vt.edu)

Authors:

Robert Thomas, Virginia Tech  (rqthomas@vt.edu)

Adrienne Breef-Pilz, Virginia Tech  (abreefpilz@vt.edu)

Austin Delany, Virginia Tech  (addelany@vt.edu)

Freya Olsson, Virginia Tech  (freyao@vt.edu)

R. Quinn Thomas, Virginia Tech  (rqthomas@vt.edu)