Near-term forecasts of NEON lakes reveal gradients of environmental
predictability across the U.S.
Abstract
The National Ecological Observatory Network (NEON)’s standardized
monitoring program provides an unprecedented opportunity for comparing
the predictability of ecosystems. To harness the power of NEON data for
examining environmental predictability, we scaled a near-term, iterative
water temperature forecasting system to six NEON lakes. We generated 1
to 35-day ahead forecasts using a process-based hydrodynamic model that
was updated with observations as they became available. Forecasts were
more accurate than a null model up to 35-days ahead among lakes, with an
aggregated 1-day ahead RMSE (root-mean square error) of 0.60℃ and
35-days ahead RMSE of 2.17℃. Water temperature forecast accuracy was
positively associated with lake depth and water clarity, and negatively
associated with catchment size and fetch. Our results suggest that lake
characteristics interact with weather to control the predictability of
thermal structure. Our work provides some of the first probabilistic
forecasts of NEON sites and a framework for examining continental-scale
predictability.