We have a new paper in Nature Communications, led by Fernando Espirito-Santo, that describes the full spectrum of natural disturbance regimes in the Amazon for the first time. How often does a sing large tree fall over, or a bunch of trees, compared to when a storm blow down knocks over a larger area of trees. Such questions are important if we are to understand what drives tree death in the tropical rainforest, and also affect how we interpret the apparent increase of biomass we see in the RAINFOR plot network. The paper combined analyses at multiple scales: Landsat satellite data over the whole region, airborne lidar data and intermediate scales, and fine-scale tree mortality data from forest plot networks. The end result is the first complete description of the disturbance regime at multiple scales in a tropical forest region (Figure 3). This is truly a thing of beauty.
In terms of carbon sinks, the analysis shows that large stochastic disturbances are too infrequent to have a significant influence on the carbon sink observed. The big disturbances appear too rare to cause most of the forest plots to be recovering from stochastic disturbances.