In this portion of a larger collaborative study, the early-career PI led a cutting-edge investigation of an annual-resolution geologic climate record from the end of the last ice age. The geologic record targeted by this work is a stalagmite, or cave formation, that grew in Hulu Cave, China. Stalagmite “H82”, as the cave sample is called, has had a significant impact on the study of past regional and global climate since it was first analyzed in 2001; it is a key record in the chronology used for radiocarbon dating. We set out to analyze a seasonal-resolution climate proxy record from this keystone record in order to quantify and interpret changes in the East Asian monsoon system during a time of remarkable climate change around the globe.

Caves are natural laboratories where calcite stalagmites can preserve a “proxy” record of temperature and/or rainfall above the cave in their chemistry as they grow. Furthermore, they can be directly dated by radiometric techniques that provide a remarkably precise chronology for the proxy records. Here, as in many stalagmite studies, we measured the ratio of two stable, naturally-occurring isotopes of oxygen in stalagmite H82 as our climate proxy record. Extensive prior work on H82 demonstrates that the ratio of ¹⁸O and ¹⁶O (symbolized as d¹⁸O) is inversely proportional to the amount of incoming solar radiation in the Northern Hemisphere. When this was first reported in 2001, it was an important corroboration of a prior hypothesis that the strength and extent of monsoon rainfall belts around the globe follow insolation. This discovery was quickly followed by a rapid increase in the study of stalagmites as climate archives, which in turn led to significant advances in our understanding of the rate, extent, and magnitude of paleoclimate changes around the globe over the last 500,000 years.

While it was known that H82 contained annual growth banding, the banding had not been used to develop the initial d¹⁸O record, chronology, or climate interpretations. In large part, this was due to the small size of stalagmite growth bands—sampling for d¹⁸O typically uses a dental drill at mm-scale, but growth bands are commonly only 0.04 mm wide. For this project, PI Orland used a combination of microscopic imaging and specialized instrumentation, called an ion microprobe, to make 0.01 mm-diameter spot analyses of d¹⁸O at seasonal-resolution across the annual bands of H82. First, we tested whether we could measure similar seasonal changes in the d¹⁸O of individual growth bands as are caused by monsoon rainfall above the cave. 

Interestingly, we measured no seasonal pattern of d¹⁸O variation in the annual bands of H82. While this prevented us from interpreting changes in the seasonality of the monsoon at Hulu Cave, likely due to mixing of annual rainwaters as they percolated into the cave, it also meant that we could better reconstruct an annual d¹⁸O record without worrying about exactly which season we targeted in each growth band. At this point, the project focus shifted to reconstructing climate at annual resolution across the entire sample. We developed and tested automated sampling protocols for the ion microprobe that allowed for nearly 3x greater throughput, with only a negligible sacrifice in d¹⁸O precision. By itself, this method development is an important contribution, making speleothems an additional source of high-resolution climate reconstructions as scientists work to understand global climate dynamics on human timescales.

Our annual-resolution d¹⁸O data from H82 not only bring an unprecedented level of detail to the structure and rate of climate changes recorded during this period, but also provide an opportunity to investigate periodic climate variation. Recent decadal-resolution observations of atmospheric CH₄ measured in a polar ice core (which follows northern hemisphere monsoon strength on millennial timescales) had 50-80 year periodicity; if a similar period were observed in the d¹⁸O of H82 it might indicate a common climatic cause. Notably, frequency analysis of the new H82 record was not coherent with the CH₄ data, and instead showed more power in higher-frequency cycles. While this disparity may be a result of misaligned age models for the two records, it does corroborate hypotheses that solar cycles (20 year and shorter periods) affect regional monsoon behavior.

The new d¹⁸O record from H82 and the methodological developments that made it possible are significant results. Here, we tested a new analytical approach that greatly increased the speed of geochemical analysis and allowed us to produce a 5,800-year reconstruction (with at least one analysis per year from 16,500-10,700 years ago) of past climate change as major ice sheets rapidly shrank following the peak of the last ice age. The new record not only provides important new data about past climate variability on decadal timescales, but also confronts some of the challenges that will be faced as scientists push to reconstruct past climate on human timescales.

Last Modified: 12/27/2023
Modified by: Ian J Orland