Spring ice-off dates gave no clear forecast gain beyond local time trends at 37 lake sites in 1991–2020
We tested whether spring ice-off dates improve forecasts of the next ice-on date beyond a local time trend. We rebuilt open-water season pairs from the NSIDC Global Lake and River Ice Phenology Database. The fixed sample has 37 lake sites, including 28 in Finland. Models used only earlier years for fitting and gave 981 forecasts in 1991–2020. Adding spring ice-off increased equal-site mean squared error from 280.15 to 286.89 days squared. Error reduction was −2.40%, with a 95% range of −4.92% to +0.05% from shared three-year block resampling. The later-period mean date correlation weakened from −0.207 to −0.052 after local trends were removed. Rolling fits and prior-date controls also gave no clear gain. This result applies to the tested linear models and selected dated freeze events. It does not exclude lake heat memory or gains from other models. The package preserves unchanged source data, source-row links, all forecasts, uncertainty draws, figures, code and independent checks.
lake ice · seasonal forecasts · ice phenology · climate change · reproducibility · environmental time series
References
6- usesdoi:10.7265/N5W66HP8
Source lake ice dates and lake-site metadata.
- usesMean ice-on was 8.3 days later and mean ice-off was 5.2 days earlier at 43 lake sites in 1991–2020 than in 1961–1990 sci:sha256:1a118e94687abd047db2303e9a6f34df45ebab5e78d22a0ffc7cbc288ccdfe59
We reuse its retained unchanged NSIDC source CSV files and check them against current online bytes. We build new open-water season pairs.
- Reported freeze dates gave no clear mean forecast gain beyond time trends at 41 Finnish lake sites in 1991–2020 sci:sha256:8d546af5c153fce036e970acb0eb541c7f2c1c754207071c3602e55c9f6c8e5d
This earlier test predicts spring thaw from prior freeze. It informed the trend baseline and shared-year checks. We test the other direction across the open-water season.
- doi:10.1126/science.289.5485.1743
Prior evidence of long-term freeze and break-up date changes.
- doi:10.1038/s41467-021-22657-4
Background on the ecological role of lake stratification timing.
- doi:10.1029/2024JG008382
Prior evidence linking ice timing and winter lake temperature, with limits on event dates as measures of lake heat.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.