A historical climate dataset for southwestern Australia, 1830–1875

Abstract

Southwestern Australia is a high priority area for climate research as there has been a significant, anthropogenically influenced winter rainfall decline detected in the region since 1970. Here we present the oldest daily historical climate dataset for Perth, southwestern Australia, to provide an extended record for analysing pre‐industrial climate variability and extremes from the region. This newly digitized record contains sub‐daily observations of temperature, barometric pressure, wind direction and weather remarks, including rain days, from 1830 to 1875. Following quality control and homogeneity adjustments, we demonstrate that the historical dataset reliably resolves modern characteristics of Perth's weather and climate variability, including the seasonal cycle and the ability to capture daily extremes like storm and heatwave events. We compare covariations in temperature, pressure, wind and rain days to cross‐verify our historical dataset against modern observations for Perth. We also use the historical dataset, modern Australian Bureau of Meteorology observations, the Twentieth Century Reanalysis, and historical documents to further validate our results. The multivariate method introduced in this study demonstrates that using multiple variables and data sources is a useful approach to reconstruct past variability, extremes and their societal impacts from historical observations. This is the first long instrumental record to be developed for southwestern Australia; an important region for observing pre‐industrial Indian and Southern Ocean variability and extremes. The new record from Perth provides 19th century data from a data sparse region of the Southern Hemisphere that can now be compared with daily observations from South Africa, New Zealand and southeastern Australia. This comparison will improve our understanding of contemporary changes in Southern Hemisphere atmospheric circulation.

Australian climate warming: observed change from 1850 and global temperature targets

“…There are challenges in assessing Australian temperature change prior to 1910, as there are very limited data available outside mainland south-east Australia before the mid-1870s, apart from some for Western Australia (Gergis et al 2021) and limited records from the Northern Territory. There are also substantial inhomogeneities in the dataset as a result of the wide range of instrument exposures prior to the introduction of the Stevenson thermometer screen as a standard.…”

New perspectives on historical climatology

“…Reconstructing historical climates from archives of societies remains a core mission of historical climatology. Recent years have brought numerous new studies of early weather measurements (Brugnara et al, 2020; Camuffo et al, 2019, 2020, 2021), further digitization and publication of early instrumental records (Ashcroft et al, 2021; Burt & Burt, 2019; Gergis et al, 2021; Pappert et al, 2021; Rodrigo, 2019a), as well as applications of early instrumental data to analyze climate and weather events, such as the cooling impact of volcanic eruptions (Brönnimann et al, 2019; Picas & Grab, 2020; Rodrigo, 2019b). Research has continued into descriptive records as well, including weather chronicles (e.g., Filipiak et al, 2019), ships' logbooks (e.g., García‐Herrera et al, 2018), and official records (e.g., Haolong et al, 2020).…”

Assessing homogeneity of land surface air temperature observations using sparse‐input reanalyses

“…Even in 1850, there exist numerous stations in the Northern Hemisphere. The same cannot be said for the Southern Hemisphere (Figure 20, bottom panel) where the earliest records in the ISTI databank currently arise from the single station in Tasmania (Gergis et al ., 2021), although data rescue efforts can, hopefully, improve this situation in the future (Brönnimann et al ., 2019).…”