Time Scales and Gaps: What Measurement Density Reveals About Earth’s History

Much of our understanding of Earth’s past is derived from stratigraphic records exposed in rock outcrops or recovered from drilled cores. These records span immense time intervals, from thousands to billions of years, and form the basis of geochronologies used to reconstruct geological, climatic, and environmental change. However, as a new study published in Nature Communications Earth & Environment shows, these records are far from uniform. Co-authored by Andrej Spiridonov, Professor of Geology and Paleontology at Vilnius University, the research demonstrates that pronounced gaps, clusters, and variations in geological records fundamentally shape how Earth’s history can be interpreted.

The study, Time scales and gaps, Haar fluctuations and multifractal geochronologies, was carried out by an international team of researchers from Canada, Germany, Chile, and Lithuania. Among the authors is Shaun Lovejoy, Professor of Physics at McGill University (Canada) and one of the paper’s lead authors. Using 23 geochronological datasets spanning the Holocene, Quaternary, Phanerozoic, and Precambrian, the team focuses not only on the data themselves, but also on how densely those data are distributed through time.

Measurement density as a key quantity

“Geological records are often treated as if they were evenly sampled through time, but in reality they are highly uneven. These variations are not just technical limitations – they carry information that fundamentally affects how we interpret Earth’s past”, says Prof. Spiridonov. 

In quantitative geochronology, samples are typically taken at regular depth intervals, and properties such as rock density or isotopic composition are measured. To be useful, these samples must then be assigned ages, creating a chronological framework. Yet this process is complicated by highly variable sedimentation rates, erosion, and missing strata, sometimes entire geological intervals are absent from the record.

As Prof. Lovejoy explains in his accompanying blog text, the traditional way of dealing with this complexity has been to identify key reference points or “events” with established ages and interpolate between them. While this allows individual samples to be dated, it can obscure the uneven structure of the underlying record.

The new study addresses this issue by analysing measurement density – the number of measurements per unit time – across a wide range of time scales. This quantity turns out to be highly informative. Rather than behaving in the expected rather uniform manner, measurement density exhibits systematic fluctuations that follow scale-invariant, or multifractal, patterns.

From classical statistics to multifractals

To characterise these fluctuations, the authors apply Haar fluctuation analysis, a method that reveals how variability changes with time scale. Their results show two distinct regimes. At shorter time scales, up to around one million years in the geological records they analysed, measurement density fluctuations resemble the familiar “bell curve” of classical statistics. At longer time scales, however, the fluctuations become so large that classical statistical approaches no longer apply.

Instead, the density follows a multifractal behaviour, meaning that its statistical properties vary across scales but do so in a structured, predictable way. Importantly, this behaviour is observed consistently across many different types of records, including pollen data, lake sediments, ice cores, loess, speleothems, and marine sediments.

According to the authors, identifying the transition from classical to multifractal behaviour provides new insight into the dynamics of Earth’s system. It shows that variability in geological and climatic records is not merely noise, but reflects deep, hierarchical organisation across time scales.

Gaps, bias, and a new paleoclimate indicator

One of the key implications of this work is the recognition of measurement density as a new paleoindicator. Traditionally, gaps in the geological record have been treated as a problem or a limitation. This study demonstrates that the distribution of those gaps, and the uneven sampling they produce, contains valuable information.

The authors find that measurement density is typically correlated with primary paleoclimate indicators, such as paleotemperature fluctuations. For example, in ice cores, periods of high variability tend to be sampled more densely, while quieter intervals are often underrepresented. Without accounting for this effect, statistical interpretations of paleoclimate data can become biased.

By explicitly analysing measurement density, it becomes possible, in principle, to “unbias” paleoclimate records and other spectra. In this way, information about missing data and gaps can be transformed from an apparent absence of evidence into positive knowledge about the structure of Earth’s past variability. However, the methodological work required to implement such corrections has yet to be carried out. At present, removing the bias would require detailed knowledge of the measurement density and its relationship to the specific paleo-indicator being analysed.

How much time is enough?

The study also addresses a fundamental question: how long a geological record is needed to capture the full range of Earth’s behaviour. The authors estimate that revealing the outer scale of Earth system variability requires records spanning at least half a billion years, and possibly longer. This helps explain why shorter time scales often fail to capture both stable and highly chaotic regimes.

Interestingly, evidence from the longest carbonate records hints at a possible shift in behaviour at time scales beyond roughly 500 million years, where variability may begin to decrease rather than continue increasing. However, further analysis is needed before firm conclusions can be drawn.

As Prof. Lovejoy emphasises, these findings have implications not only for understanding Earth’s past, but also for how future planetary change is modelled. Recognising the multifractal structure of geochronological data provides a more realistic framework for interpreting variability, extremes, and long-term trends.

For further reading: https://communities.springernature.com/posts/multifractal-geochronologiesn