Abstract
The dataset described here includes estimates of historical (1980–2020) daily surface water temperature, lake metadata, and daily weather conditions for lakes bigger than 4 ha in the conterminous United States (n = 185,549), and also in situ temperature observations for a subset of lakes (n = 12,227). Estimates were generated using a long short-term memory deep learning model and compared to existing process-based and linear regression models. Model training was optimized for prediction on unmonitored lakes through cross-validation that held out lakes to assess generalizability and estimate error. On the held-out lakes with in situ observations, median lake-specific error was 1.24°C, and the overall root mean squared error was 1.61°C. This dataset increases the number of lakes with daily temperature predictions when compared to existing datasets, as well as substantially improves predictive accuracy compared to a prior empirical model and a debiased process-based approach (2.01°C and 1.79°C median error, respectively).
| Original language | English (US) |
|---|---|
| Pages (from-to) | 287-301 |
| Number of pages | 15 |
| Journal | Limnology And Oceanography Letters |
| Volume | 7 |
| Issue number | 4 |
| DOIs | |
| State | Published - Aug 2022 |
Bibliographical note
Funding Information:Jeff Holister helped with elevation data retrieval. Samantha Oliver helped assemble midwestern temperature data. R. Iestyn Woolway helped assemble resources to consider for dataset comparison. David Blodgett helped design file formats for compliance with existing standards. Shad Mahlum assisted with error estimation. Ellen Bechtel designed Fig. 1. Kathy Webster helped us better understand lake filtering done in compiling the LAGOS-US spatial dataset, which provides access to additional data for most of the lakes in this dataset. The Minnesota Supercomputing Institute (MSI) at the University of Minnesota provided resources that contributed to the research results reported within this paper. Jared Smith and Evan Goldstein provided early review of this manuscript and we thank L&O:L journal staff and two anonymous reviewers that helped improve the paper and underlying dataset. We thank data contributors who have made future research more efficient by sharing their data through prior data releases or the Water Quality Portal. This work was sponsored by the Department of the Interior United States Geological Survey Midwest Climate Adaptation Science Center, and also NSF grant 1934721 under the Harnessing the Data Revolution (HDR) program under PI Vipin Kumar. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.
Funding Information:
Jeff Holister helped with elevation data retrieval. Samantha Oliver helped assemble midwestern temperature data. R. Iestyn Woolway helped assemble resources to consider for dataset comparison. David Blodgett helped design file formats for compliance with existing standards. Shad Mahlum assisted with error estimation. Ellen Bechtel designed Fig. 1 . Kathy Webster helped us better understand lake filtering done in compiling the LAGOS‐US spatial dataset, which provides access to additional data for most of the lakes in this dataset. The Minnesota Supercomputing Institute (MSI) at the University of Minnesota provided resources that contributed to the research results reported within this paper. Jared Smith and Evan Goldstein provided early review of this manuscript and we thank journal staff and two anonymous reviewers that helped improve the paper and underlying dataset. We thank data contributors who have made future research more efficient by sharing their data through prior data releases or the Water Quality Portal. This work was sponsored by the Department of the Interior United States Geological Survey Midwest Climate Adaptation Science Center, and also NSF grant 1934721 under the Harnessing the Data Revolution (HDR) program under PI Vipin Kumar. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. L&O:L
Publisher Copyright:
© 2022 The Authors. Limnology and Oceanography Letters published by Wiley Periodicals LLC on behalf of Association for the Sciences of Limnology and Oceanography.