Constraining the global mean surface temperature during 1850-1880 with new statistical physical model

TL;DR

This paper introduces a new statistical physical model to better estimate the early 19th-century global surface temperature, revealing that previous datasets overestimated anomalies and underestimated warming rates since pre-industrial times.

Contribution

A novel statistical physical model that explicitly quantifies external forcings as deterministic trends to improve early 19th-century temperature reconstructions.

Findings

Existing datasets overestimate early temperature anomalies.

The actual warming rate since pre-industrialization is higher than previously estimated.

The model provides a more accurate baseline for climate change assessments.

Abstract

As IPCC ARs stated, global warming is estimated based on the average from 1850 to 1900 (global average temperature of preindustrialization estimated from relatively sparse observations). Given the impossibility of massive increasing observation data in the early stages, accurately constraining this baseline has become an unresolved issue. Here we developed a new statistical physical model to quantify the contribution of external forcings to global warming as a "deterministic trend" of the surface temperature series (instead of as non-stationary processes that yield a stochastic trend) and constrained the reconstruction of the early time series (1850-1880). We find that the existing datasets slightly overestimated the temperature anomalies in this period, thus the speed of global warming since pre-industrialization is still underestimated.