Greener and more fragile at the same time
Global forests are gaining productivity while losing resilience. Multi-source remote sensing covering 2001 to 2021 puts more than a third of them in both columns at once, and rising vapour pressure deficit — the shortfall between the moisture the air can hold and the moisture it holds — comes out as the primary driver of the resilience loss. Productivity is the number most forest accounting rests on, and it moves in the reassuring direction while the capacity to come back from a bad year moves the other way.[1]
Resilience erosion weakened the coupling between gross primary productivity and ecosystem photosynthetic efficiency in boreal forests, while vapour pressure deficit strengthened the covariation between productivity and water use efficiency in tropical and boreal forests. A tighter carbon-water number looks like better coordination; the paper instead puts it down to stomatal constraint on transpiration variability, the leaf closing its pores and narrowing the range it can operate over. In intact forests the same pressure weakened that coupling, which is a reminder that one atmospheric signal does not produce one response.[1]
Structure decides who pays first. Non-intact forests were twice as sensitive to resilience loss as intact ones, which reads as structural simplification multiplying a signal that arrives everywhere. The alternative reading deserves to stay on the table: non-intact forests sit disproportionately in drier and more disturbed places, so part of the doubled sensitivity could be where they are rather than what they have become.[1]
Amazonia outruns its own average
The 95th percentile of maximum temperatures in Amazonia's driest period increased by 0.49 °C per decade, against 0.21 °C per decade for the central trend of mean temperatures for the year. Over 700,000 km² of central-north Amazonia has taken dry-season extremes of at least 0.75 °C per decade, at least 3.22 °C across 43 years. A basin that is often summarised by one annual anomaly turns out to be doing most of its moving in the hot tail of the dry season.[2]
Vapour pressure deficit is the primary driver of forest resilience loss, and temperature-linked measures of water deficit change faster than central trends in Amazonia; the two measurements name the same quantity from different ends. What the atmosphere demands of a leaf is the term that is moving, and monitoring built on annual means and on productivity is measuring the parts that are still behaving, then reporting them as health.[1], [2]
What moves the number
2005 marked a turn in an earlier column here, which read the Amazon droughts through warming-driven atmospheric demand overtaking rainfall deficit as the dominant intensifier. The basin-wide quantiles now attach a rate to that turn and put it somewhere: the driest period, the hot tail, central-north Amazonia. Rainfall remains a real variable; it has stopped being a sufficient one.[2], [3]
Atmospheric aridity is not something management can change on a decade's timescale, so the usable lever is the structural simplification that doubles sensitivity — and the Amazon authors land in the same place when they conclude that adaptation measures must include preventing the deforestation and disturbances that amplify risks. That lever is held by forest agencies, concession regulators and the people who live in the basin, not by the emissions pathway alone. The test that settles it is whether forest monitoring programmes start publishing a resilience term next to the productivity term, because a canopy that greens and destabilises at once reads as improvement in any ledger that counts only growth.[1], [2]