{
  "abstract": "Root-mean-square error (RMSE) remains the default training loss for data-driven precipitation models, despite precipitation being semicontinuous, zero inflated, strictly nonnegative, and heavy tailed. This Gaussian-implied objective misspecifies the data-generating process because it tolerates negative predictions, underpenalizes rare heavy events, and ignores the mass at zero. We suggest replacing RMSE with the Tweedie deviance, a likelihood-based and differentiable loss from the exponential–dispersion family, with variance function V ( μ ) = μ p . For 1 < p < 2, it yields a compound Poisson–gamma distribution with a point mass at zero and a continuous density over y > 0, matching observed precipitation characteristics. In this paper, we (i) estimate p from the variance–mean power law and show that precipitation across temporal aggregations is far from Gaussian and that the Tweedie power p increases with accumulation length toward a gamma limit and (ii) demonstrate consistent skill gains when training deep data-driven models with Tweedie deviance instead of RMSE. In diffusion-model downscaling over Beijing, Tweedie loss improves wet-pixel MAE and extreme recall (∼0.611 vs 0.566 at the 99th percentile). In convolutional long short-term memory (ConvLSTM) nowcasting over Kolkata, Tweedie loss yields improvements in wet-pixel MAE and dry-pixel hit rates, with improvements autoregressively compounding over lead time (for MAE, from 3.5% at t + 1 growing to 8.7% at t + 4). Extreme recall at the 99.9th percentile is not significantly higher at t + 1 but becomes significantly higher by t + 3 and t + 4. Because the Tweedie deviance is continuous in p , it adapts smoothly across scales, offering a statistically justified, practical replacement for RMSE in precipitation-based learning tasks.",
  "abstract_shown": true,
  "affiliations": [
    "Department of Meteorology, University of Reading, Reading, United Kingdom",
    "National Centre for Atmospheric Sciences, University of Reading, Reading, United Kingdom"
  ],
  "article_type": null,
  "arxiv": null,
  "author_total": 1,
  "authors": [
    {
      "affs": [
        0,
        1
      ],
      "name": "Kieran M. R. Hunt",
      "orcid": "0000-0003-1480-3755"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/a11j7thj3qd33.bib",
  "cited_by": 9,
  "doi": "10.1175/aies-d-25-0083.1",
  "fulltext_url": null,
  "id": "a11j7thj3qd33",
  "issns": [
    "2769-7525"
  ],
  "landing_url": "https://doi.org/10.1175/aies-d-25-0083.1",
  "language": "en",
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "journal",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "openalex",
      "1902|Precipitation Measurement and Analysis"
    ]
  ],
  "openalex": "W4417065230",
  "pdf_url": null,
  "pmcid": null,
  "pmid": null,
  "published": "2026-07-29",
  "publisher": "American Meteorological Society",
  "ris": "https://research.draconitic.com/paper/a11j7thj3qd33.ris",
  "sources": [
    {
      "first_seen": "2026-11-04T13:21:15Z",
      "native_id": "W4417065230",
      "source": "openalex",
      "url": "https://doi.org/10.1175/aies-d-25-0083.1"
    }
  ],
  "subjects": [
    {
      "field": "earth",
      "origin": "openalex:1902|Precipitation Measurement and Analysis",
      "subfield": "atmospheric"
    }
  ],
  "title": "Stop Using Root-Mean-Square Error as a Precipitation Target!",
  "title_text": "Stop Using Root-Mean-Square Error as a Precipitation Target!",
  "type": "peer-reviewed",
  "updated_at": "2026-11-07T13:21:01Z",
  "url": "https://research.draconitic.com/paper/a11j7thj3qd33",
  "venue": "Artificial Intelligence for the Earth Systems",
  "versions": []
}