Dendrochronology

By crossmatching. You measure ring-width sequences from modern trees (dated by counting from the present). Then measure older dead wood and align its ring pattern to the modern sequence. If a dead wood series has 200 rings and its pattern matches the 500-600 year old section of the modern chronology, you know that dead wood is 500-600 years old. Overlap multiple series to extend the chronology backward century by century.

Because ring width responds most strongly to water availability (moisture stress directly limits growth). Temperature is less directly linked to ring width because trees often grow along a temperature gradient where other factors (light, nutrients) co-vary with temperature. To reconstruct temperature, researchers use other tree-ring properties: stable isotopes, maximum density, or cellular anatomy—which respond differently to temperature. Precipitation reconstructions are generally more reliable.

Young trees (1-50 years) show high growth rates from juvenile wood formation and less competition. Old trees (100+ years) show lower growth due to aging physiology and competition. A chronology built from old wood may underestimate climate sensitivity because the old trees are inherently slow-growing. To correct, researchers standardize ring widths using age-related growth curves before building the chronology, or use regional growth rate models.

Yes, but with caveats. Tropical trees have less distinct annual rings because they lack a cold season. Many tropical species show multiple growth rings per year or no clear rings. However, in seasonally wet-dry tropics (monsoon regions), rings are visible and respond to seasonal rainfall. Tropical dendrochronology is less developed; chronologies are shorter and more scattered geographically. Combine with other proxies (corals, ice cores, speleothems) for broader coverage.

Verification skill (RE statistic, typical >0.4 for well-developed chronologies) indicates that 40-60% of climate variance is predictable from tree rings. This means paleoclimate reconstructions capture large-scale trends and extreme events well, but not small-scale noise. Confidence intervals typically ±0.5-1.0°C for temperature or ±15-20% for precipitation. Skill varies regionally and by climate variable; tropical and Southern Hemisphere chronologies are less developed and less skillful.