Uncertainty growth in global ensemble forecasts: attribution and sensitivity to resolution

[thumbnail of rodwell_uncertainty_resubmitted_two_column.pdf]
Text
- Accepted Version
· Restricted to Repository staff only until 14 March 2027.

Please see our End User Agreement.

It is advisable to refer to the publisher's version if you intend to cite from this work. See Guidance on citing.

Add to AnyAdd to TwitterAdd to FacebookAdd to LinkedinAdd to PinterestAdd to Email

Rodwell, M. J., Tsiringakis, A., Gray, S. L. ORCID: https://orcid.org/0000-0001-8658-362X, Methven, J. ORCID: https://orcid.org/0000-0002-7636-6872 and Wood, D. (2026) Uncertainty growth in global ensemble forecasts: attribution and sensitivity to resolution. Monthly Weather Review. ISSN 1520-0493 doi: 10.1175/MWR-D-26-0030.1

Abstract/Summary

The jet stream is a key aspect in mid-latitude synoptic-scale predictability. We explore the diabatic and dynamic mechanisms that amplify jet stream level uncertainty in global ensemble weather forecasts. Uncertainty growth requires positive ensemble covariances between the prevailing (potential vorticity) anomalies and their associated tendencies. In an example of mesoscale convection over North America, we find that diabatic tilting of horizontal vorticity (associated with mean vertical shear) can result in strong non-conservative growth. We suggest that a similar mechanism operates during the extratropical transition of a tropical cyclone. Conservative growth, due to dynamical processes, tends to be weaker and more ubiquitous. However, it can be stronger at the ‘necks’ of troughs (where horizontal winds are near normal to mean PV contours) and may lead to uncertainty in cut-off formation. We argue that these growth-rate mechanisms were responsible for unusually large vector wind uncertainties in a storm over Europe. Our case study results are consistent with a three-month systematic analysis, with small-scale (18–72km) non-conservative processes being responsible for the largest growth rates (leading to occasional forecast ‘bust’ situations) and larger-scale conservative processes being more important for mean uncertainty growth. We discuss implications for the parametrization of model uncertainty, and suggest that a similar diagnostic approach would be valuable for understanding uncertainty growth in data-driven forecasts.

Altmetric Badge

Dimensions Badge

Item Type Article
URI https://centaur.reading.ac.uk/id/eprint/131660
Identification Number/DOI 10.1175/MWR-D-26-0030.1
Refereed Yes
Divisions Science > School of Mathematical, Physical and Computational Sciences > Department of Meteorology
Publisher American Meteorological Society
Download/View statistics View download statistics for this item

University Staff: Request a correction | Centaur Editors: Update this record