Abstract Tracking climatic changes throughout the Ordovician is crucial to a better understanding of the coevolution of life and environment on Earth. Ordovician climate fluctuations have been the subject of a vigorous and productive body of work over the past two decades. Here we present a synthesis of studies that have focused on reconstructing Ordovician climate and environment via direct geochemical proxy datasets and/or numerical modelling approaches. Many new insights have been gained on potential causes of events: the timing and potential causes of the major radiation of Ordovician marine life, changes in weathering and the transition from a greenhouse-to-icehouse state, and cooling and (de)oxygenation during the end-Ordovician Glaciation. Marked improvements in sample resolution/distribution of traditional palaeotemperature (oxygen isotopes), palaeoredox (sulfur isotopes) and weathering proxy records (strontium and neodymium isotopes), as well as the development of new palaeoenvironmental proxies (e.g. clumped, uranium, molybdenum and thallium isotopes; iodine, iron, trace metal geochemistry), have led to better constraints on Ordovician climate and environment. These recent works have led to a more nuanced understanding of the Ordovician Earth System, which allows the global community to focus future efforts on answering remaining questions regarding palaeoclimate and environment, as well as embark upon new investigations.

Two prominent, and apparently globally distributed, δ 13 C excursions have been documented from the Upper Ordovician, namely the early Katian Guttenberg isotope carbon excursion (GICE) and the latest Ordovician Hirnantian isotope carbon excursion (HICE). The former excursion, which has lower δ 13 C values than the HICE, is now recorded from dozens of localities in North America and Baltoscandia, and it appears to be present also in China. In North America the GICE ranges from the uppermost Phragmodus undatus Midcontinent Conodont Zone to near the top of the Plectodina tenuis Midcontinent Conodont Zone, an interval corresponding to the lower part of the Diplacanthograptus caudatus Global Graptolite Zone. The base of the GICE lies somewhat above the Millbrig K-bentonite. In Baltoscandia the GICE occurs in the upper Diplograptus foliaceus through the lower Dicranograptus clingani Graptolite Zones, and in the upper Amorphognathus tvaerensis Conodont Zone. Its base is a few meters above the widespread Kinnekulle K-bentonite. In Baltoscandia and in Oklahoma the GICE ranges through a part of the Spinachitina cervicornis Chitinozoan Zone. In North America the GICE is regionally in a transgressive-regressive succession. The bathymetric conditions in the GICE interval in Baltoscandia were somewhat complex and have been the subject of different interpretations, but there is no obvious correlation between the GICE and apparent sea level changes. A review of the relations between the GICE and potential climatic and water temperature indicators, such as lithofacies, faunas, and 18 O geochemistry, does not suggest a close correlation to specific environmental conditions. The cause of formation of the GICE is enigmatic, but there is no direct evidence that it was coeval with a period of extensive glaciation in the Gondwana. The GICE is a powerful chemostratigraphic tool that is useful for detailed local and even transatlantic correlations.