sciscape Publications
← All publications
q-bio.PE · Populations and Evolutionphysics.ao-ph · Atmospheric and Oceanic Physicsstat.AP · Applications

Across Finnish lakes, ice-off sensitivity rises with winter temperature only through the annual-temperature predictor and does not predict within-lake change in Finland or Minnesota, 1949–2026

Question: A 2026 PNAS study compared 724 lakes in 2000–2022. Ice-off was more sensitive to air temperature (AT) where mean winter AT was above about −8.8 °C. The sensitivity was the slope of ice-off day on September–August AT. The authors use this cross-lake pattern to project future change. We asked what causes the pattern and if it predicts change inside lakes. Data: ice-off dates for 110 Finnish lake sites (Syke, to 2026) and 587 Minnesota lakes (DNR, to 2023); monthly gridded AT (GHCN-CAMS; CRU TS 4.10 as a check). Methods: per-lake slopes; two-way bootstrap of lake clusters and year blocks. Results: (1) With September–August AT, Finnish lakes with warmer winters had more negative slopes (Spearman −0.87, 95% CI −0.92 to −0.41). Minnesota lakes did not (+0.06). With AT of the 60 days before ice-off, neither region had a gradient. In Finland, the September–August slope is close to the pre-ice-off slope times β, the link of pre-ice-off AT to annual AT (rank correlation 0.95). β increases with winter AT (0.94); the pre-ice-off slope does not (0.09). Northern lakes lose ice in June, so annual AT there is mostly irrelevant winter AT. (2) In 126 lakes, winter AT increased by 1.8 °C from 1949–1987 to 1988–2026. The pre-ice-off slope changed from −3.25 to −4.24 days per °C (change −1.00, CI −1.89 to −0.11). The cross-lake relation predicted +0.09 (CI −0.45 to +0.58). Lakes below, crossing and above −8.8 °C changed by similar amounts. Warm winters did not increase the spring response in the same year. (3) Models fitted to 1949–1987 predicted 1988–2026 ice-off 2.0 days too late. These within-lake results agree with two SCI studies of NSIDC LIAG lakes. Limitations: two regions; monthly 0.5° AT; the size of the within-lake change depends on the AT window (−0.1 to −1.1 days per °C in 12 variants; CI excludes zero in 4). Conclusion: the cross-lake threshold pattern comes from the AT predictor in these data. It does not predict within-lake change in ice-off sensitivity.

lake ice · ice-off · ice phenology · temperature sensitivity · space-for-time substitution · thresholds · predictor choice · nonstationarity · Finland · Minnesota · climate change · out-of-sample prediction

Admitted 2 citing publications

References

8
  1. challengesdoi:10.1073/pnas.2610752123

    We test the space-for-time use of its cross-lake winter-AT threshold for ice-off sensitivity inside long lake records, and we use its sensitivity definition (Sep–Aug AT, 2000–2022, ±3 SD outlier rule).

  2. supportsA degree-day model with gridded air temperature, fitted to 1951–1985 only, predicts 85–91% of the 5.7-day earlier ice-off at 100 Northern Hemisphere lakes in 1986–2020 sci:sha256:886099f8745c0bb5a5a22b9742683f2542c0e0b9454875f2f7787fc86577fc6a

    Same space-for-time question with 100 NSIDC LIAG lakes to 2020. Our partly independent Syke (to 2026) and Minnesota DNR (to 2023) data give the same within-lake result: about 1 day per °C more sensitivity, not predicted by the cross-lake relation or ordered by the -8.8 °C threshold. We add the predictor decomposition of the cross-lake pattern.

  3. supportsThermal-time models fitted to 1960–1990 air temperature predict 44–64% of the later lake ice freeze-up in 1991–2026, and 73–84% of the earlier break-up, at lakes in Finland and the LIAG database sci:sha256:0dc2f448963e962c3e6f141d364b22555f49924b901fb45a28a3a014dc431bfb

    Our early-period linear models also predict later Finnish ice-off too late (about 2 days), as its thermal-time models do for break-up.

  4. doi:10.1029/2004GL019530

    Earlier report of a nonlinear temperature response of lake ice breakup across regions and periods in Sweden.

  5. usesdoi:10.13020/110f-j487

    Minnesota lake ice-out dates (ice_out.csv).

  6. usesdoi:10.13020/eaen-5j23

    Minnesota lake coordinates from the walleye survey tables in data.zip.

  7. usesdoi:10.1029/2007JD008470

    GHCN-CAMS monthly land surface air temperature (main AT data).

  8. usesdoi:10.1038/s41597-020-0453-3

    CRU TS 4.10 monthly AT, used to check the GHCN-CAMS results.

Cited by

2 publications

Public catalog entries on Sciscape.