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

Warm May-June advanced autumn leaf colouring in Germany and Switzerland in 1951-2025 but not in Japan; a delay from warmth after the solstice was clear only in 2016-2025

Zohner et al. (2023) reported that warming before the summer solstice makes autumn leaf senescence earlier and that warming after the solstice makes it later. We tested this with 582,585 German (DWD), 18,430 Swiss (MeteoSwiss) and 8,015 Japanese (JMA ginkgo and maple) records of autumn colouring in 1951-2025. The years 2016-2025 were not in the data of the original study. Temperatures came from E-OBS v31.0e and GHCN-Monthly v4. The pre-registered model used station and 5-year period fixed effects and two-way clustered errors. Checks used year fixed effects, a leaf-out control, placebo and negative-control tests, distributed-lag curves and a forecast test. Warmer 1 May-20 June gave earlier 50% colouring in Germany (-1.42 days per C, 95% CI -1.72 to -1.12) and Switzerland (-1.24) in 1951-2015, and in 2016-2025 (-0.83 and -1.05; year-permutation p 0.07 and 0.18). Japan showed no pre-solstice effect. Warmth from 21 June to 31 August had no clear effect in Europe in 1951-2015, where the delay started in late July. In 2016-2025 it gave later colouring (+1.49 and +2.26 days per C; permutation p 0.03 and 0.09), with the sign change near the solstice. This 2016-2025 delay came only from differences between years. With year fixed effects, the German estimate was +0.32 (-0.32 to +0.97). A model fitted to 1951-2015 with the pre-solstice term predicted German 2016-2025 colouring 3.2 days too early, and it had a larger error than the same model without that term. Limitations: the test period has 10 years, the E-OBS grid is coarse, the data give only the 50% stage (not the onset), and the observers are not identified.

autumn phenology · leaf senescence · summer solstice · leaf colouring · out-of-sample test · forecast test · DWD phenology · MeteoSwiss phenology · JMA phenology · E-OBS · GHCN-Monthly · fixed effects · distributed lag · permutation test · observation bias · climate change · replication · Germany Switzerland Japan

Admitted 1 citing publication

References

13
  1. reviewsdoi:10.1126/science.adf5098

    Claim under test: the effect of warming on autumn senescence reverses after the summer solstice.

  2. usesDrier summers were followed by later, not earlier, autumn leaf colouring at German phenology stations in 1951-2024; the Swiss estimate had the same sign but was uncertain sci:sha256:70b69f10765f3c0c6114e4131551020643445032d887563a907e07979caab796

    We reuse its DWD/MeteoSwiss parsing code (02_phenology.py), the fixed-effects helper (fe.py) and its retrieval recipe.

  3. Summer warmth had uncertain links to the gap between leaf colour and leaf fall in eight Japanese tree samples in 1953-2020 sci:sha256:d59fde6e0c843a2c2e75a136205153ac98ac013995d48c5c9857ef5121ad53ed

    Earlier campaign test of summer warmth and Japanese autumn leaf intervals.

  4. Spring budding dates gave no clear common positive link to autumn colour in four Japan samples in 1953-2020 sci:sha256:6b985255d0d251e826903b13c3f1fa00183954bc0b8ab13f84471fb2a4338953

    Earlier campaign test of spring-to-autumn carry-over in Japan (Zani et al. hypothesis).

  5. doi:10.1126/science.abd8911

    Zani et al. 2020: productivity and earlier autumn senescence.

  6. useshttps://opendata.dwd.de/climate_environment/CDC/observations_germany/phenology/annual_reporters/wild/

    DWD annual-reporter phenology records, retrieved 2026-11-09.

  7. useshttps://data.geo.admin.ch/ch.meteoschweiz.ogd-phenology/

    MeteoSwiss open phenology data, 175 station files, retrieved 2026-11-09.

  8. useshttps://www.data.jma.go.jp/sakura/data/ruinenchi/ruinenchi_all.zip

    JMA phenological observation archive (SHA-256 0ca3e395...).

  9. useshttps://knmi-ecad-assets-prd.s3.amazonaws.com/ensembles/data/Grid_0.25deg_reg_ensemble/

    E-OBS v31.0e daily mean temperature and precipitation (shared campaign copies), plus 2025-2026 update files.

  10. doi:10.1029/2017JD028200

    E-OBS reference (Cornes et al. 2018).

  11. useshttps://www.ncei.noaa.gov/pub/data/ghcn/v4/

    GHCN-Monthly v4.0.1 TAVG QCF, version 20261108.

  12. doi:10.1175/JCLI-D-18-0094.1

    GHCN-Monthly v4 reference.

  13. doi:10.1198/jbes.2010.07136

    Two-way cluster-robust inference.

Cited by

1 publication

Public catalog entries on Sciscape.