Climate analysis & attribution
Considerable yet contrasting regional imprint of circulation change on summer temperature trends across the Northern hemisphere mid-latitudes
In one sentence. Evaluates statistical and ML decomposition methods on nudged simulations; they get the sign of circulation-driven summer warming right but underestimate its size, which is up to half of 1979–2023 warming.
Abstract
Abstract. The increase in summer temperature and heat extremes is well documented and attributed to anthropogenic climate change. There is, however, still a vivid debate about the influence of atmospheric circulation changes on summer temperatures and heat extremes. Over the northern hemispheric mid-latitudes, considerable regional differences in summer temperatures have been observed. These differences have been linked to atmospheric circulation changes using statistical methods, but it remains challenging to evaluate such methods on multi-decadal time scales. Here, we evaluate different decomposition methods and systematically investigate circulation-induced summer temperature trends. For the evaluation of statistical and machine learning decomposition methods we use climate model simulations without external forcing but with atmospheric circulation nudged towards the circulation of a freely running simulation forced by anthropogenic emissions and land use changes. We train the decomposition methods on the free-running forced simulation and compare its circulation-induced trends to the trends simulated in the nudged circulation simulation. Most decomposition methods show skill in estimating the sign of circulation-induced trends but all methods underestimate the magnitude of these trends. The application of tested decomposition methods confirms that circulation changes have contributed substantially to an increase in summer heat over several mid-latitude regions, including Europe. In this hotspot region, we estimate that up to half of the warming over the period 1979–2023 is due to circulation changes.
Labels
- Topic
- Climate analysis & attribution
- Type
- Evaluation or intercomparison of existing models
- Method
- Classical ML (trees, GP, linear, kernels)
- Time scale
- Climate projections / centennial
- Domain
- Regional
- Data
- CMIP / climate model output
- Status
- Research
- Identifiers
- DOI:10.5194/wcd-7-89-2026 OpenAlex:W7124242408
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.