Examining a solar climate link in diurnal temperature ranges

TL;DR

This study investigates the proposed link between cosmic rays, cloud cover, and diurnal temperature range, finding no significant evidence to support a direct connection between solar activity and DTR variations over 60 years.

Contribution

It provides a comprehensive analysis dismissing previous claims of a solar-related influence on DTR, clarifying the statistical significance issues in earlier studies.

Findings

No significant correlation between DTR and long-term solar cycles.

DTR response to cosmic ray variations is below detection threshold.

Previous reports of a CR-DTR link are likely due to statistical noise.

Abstract

A recent study has suggested a link between the surface level diurnal temperature range (DTR) and variations in the cosmic ray (CR) flux. As the DTR is an effective proxy for cloud cover, this result supports the notion that widespread cloud changes may be induced by the CR flux. If confirmed, this would have significant implications for our understanding of natural climate forcings. Here, we perform a detailed investigation of the relationships between DTR and solar activity (total solar irradiance and the CR flux) from more than 60 years of NCEP/NCAR reanalysis data and observations from meteorological station data. We find no statistically significant evidence to suggest that the DTR is connected to either long-term solar periodicities (11 or 1.68 year) or short-term (daily-timescale) fluctuations in solar activity, and we attribute previous reports on the contrary to an incorrect estimation of the statistical significance of the data. If a CR-DTR relationship exists, based on the estimated noise in DTR composites during Forbush decrease (FD) events, the DTR response would need to be larger than 0.03{\deg}C per 1% increase in the CR flux to be reliably detected. Compared with a much smaller rough estimate of -0.005{\deg}C per 1% increase in the CR flux expected if previous claims that FD events cause reductions in the cloud cover are valid, we conclude it is not possible to detect a solar related responses in station-based or reanalysis-based DTR datasets related to a hypothesized CR-cloud link, as potential signals would be drowned in noise.