Back

Optimizing Wastewater Surveillance Sites for COVID-19 Hospitalization Forecasting Across U.S. States

Kaur, G.; Chen, J.; Adiga, A.; Adiga, A.; Espinoza, B.; Lewis, B.; Marathe, M.; Venkatramanan, S.; Warren, A.; Vullikanti, A.

2026-07-22 infectious diseases
10.64898/2026.07.20.26358167 medRxiv
Show abstract

In this study, we analyze viral load data from wastewater treatment plants (WWTPs) across multiple U.S. states and address the challenge of selecting an optimal subset of sites to improve COVID-19 hospitalization forecasts. Using forward (greedy) regression with a baseline ARIMA model, we identify the most informative WWTPs that enhance forecast accuracy while reducing the number of sampling locations. Our analysis, based on NWSS data, shows that the optimal number of sites typically ranges from 2-8, though some states, including NY, IL, and WI, benefit from a larger set (10-20). We also leverage a Virginia-level digital twin model specifically designed for wastewater data modeling and analysis and our results show that for different parameter settings, that forecast accuracy can be achieved with strategically chosen small number of sites, providing

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.