Variations in both width and density of annual rings from a network of tree chronologies were used to develop high-resolution proxies to extend the climate record in the Wrangell Mountain region of Alaska. We developed a warm-season (July–September) temperature reconstruction that spans A.D. 1593–1992 based on the first eigenvector from principal component analysis of six maximum latewood density (MXD) chronologies. The climate/tree-growth model accounts for 51% of the temperature variance from 1958 to 1992 and shows cold in the late 1600s–early 1700s followed by a warmer period, cooling in the late 1700s–early 1800s, and warming in the 20th century. The 20th century is the warmest of the past four centuries. Several severely cold warm-seasons coincide with major volcanic eruptions. The first eigenvector from a ring-width (RW) network, based on nine chronologies from the Wrangell Mountain region (A.D. 1550–1970), is correlated positively with both reconstructed and recorded Northern Hemisphere temperatures. RW shows a temporal history similar to that of MXD by increased growth (warmer) and decreased growth (cooler) intervals and trends. After around 1970 the RW series show a decrease in growth, while station data show continued warming, which may be related to increasing moisture stress or other factors. Both the temperature history based on MXD and the growth trends from the RW series are consistent with well-dated glacier fluctuations in the Wrangell Mountains and some of the temperature variations also correspond to variations in solar activity.

This article presents a late summer temperature reconstruction (AD 1675–1980) for the northeastern Mediterranean (NEMED) that is based on a compilation of maximum latewood density tree-ring data from 21 high-elevation sites. This study applied a novel approach by combining individual series from all sites into one NEMED master chronology. This approach retains only the series with a strong and temporally robust common signal and it improves reconstruction length. It further improved the regional character of the reconstruction by using as a target averaged gridded instrumental temperature data from a broad NEMED region (38–45°N, 15–25°E). Cold (e.g. 1740) and warm (e.g. 1945) extreme years and decades in the reconstruction correspond to regional instrumental and reconstructed temperature records. Some extreme periods (e.g. cold 1810s) reflect European-wide or global-scale climate conditions and can be explained by volcanic and solar forcing. Other extremes are strictly regional in scope. For example, 1976 was the coldest NEMED summer over the last 350 years, but was anomalously dry and hot in northwestern Europe and is a strong manifestation of the summer North Atlantic Oscillation (sNAO). The regional NEMED summer reconstruction thus contributes to an improved understanding of regional (e.g. sNAO) vs. global-scale (i.e. external) drivers of past climate variability.