Past studies of cyclones (hurricanes, typhoons, tropical cyclones) disturbance showed that meteorological, topographical, and biological factors affect the patterns of forest disturbance intensity but left open the extent to which these findings were representative across different global cyclone regions. Using remote sensing data and machine learning models, we examined how these factors change over spatial scales and assessed their consistency across four major cyclones: Katrina (August 2005), Rita (September 2005), Yasi (February 2011), and María (September 2017). Our results revealed that the factors which best explained forest disturbance intensity pattern varied across these regions. Wind speed and precipitation were the dominant factors contributing to the variation in impacts of Katrina; terrain features, especially elevation, explained most of the variation in disturbance intensity of Rita; pre-disturbance vegetation condition was significant predictors of effects of Yasi; these factors played equal roles in explaining the disturbance intensity variation of María. A 40 m/s (144 km/h) wind speed threshold was proposed to split low- and high-level forest disturbance intensity. Other than wind speed, few generalizations can be made on features across multiple regions. We built several generalized hurricane impact models, which worked well with the test data from cyclones used for model development (R2 = 0.89). However, these models did not have good predictions on other cyclones, such as Michael (October 2018) and Laura (August 2020). This study showed that each cyclone interacted with the landscape in a unique way and the challenges remained in building a generalized cyclone impact model.

Widely recognized as one of the worst natural disaster in Puerto Rico's history, hurricane María made landfall on September 20, 2017 in southeast Puerto Rico as a high-end category 4 hurricane on the Saffir-Simpson scale causing widespread destruction, fatalities and forest disturbance. This study focused on hurricane María's effect on Puerto Rico's forests as well as the effect of landform and forest characteristics on observed disturbance patterns. We used Google Earth Engine (GEE) to assess the severity of forest disturbance using a disturbance metric based on Landsat 8 satellite data composites with pre and post-hurricane María. Forest structure, tree phenology characteristics, and landforms were obtained from satellite data products, including digital elevation model and global forest canopy height. Our analyses showed that forest structure, and characteristics such as forest age and forest type affected patterns of forest disturbance. Among forest types, highest disturbance values were found in sierra palm, transitional, and tall cloud forests; seasonal evergreen forests with coconut palm; and mangrove forests. For landforms, greatest disturbance metrics was found at high elevations, steeper slopes, and windward surfaces. As expected, high levels of disturbance were also found close to the hurricane track, with disturbance less severe as hurricane María moved inland. Results demonstrated that forest and landform characteristics accounted for 34% of the variation in spatial forest spectral disturbance patterns. This study demonstrated an informative regional approach, combining remote sensing with statistical analyses to investigate factors that result in variability in hurricane effects on forest ecosystems.

This study identifies 38 cases of windthrows in the Amazonia to explore the relationship between windthrows and the characteristics (storm passing time, cloud top temperature, and maximum precipitation) of mesoscale convective systems (MCSs) that produced them. Most of windthrow cases in this study occurred in August and September. The storm passing time is positively correlated with the size of windthrows. MCSs with colder cloud top temperature (with a mean at 206 K)—indicating deeper convection—resulted in large windthrows, while those with warm cloud top (with a mean above 230 K) resulted in relatively small windthrows except for windthrows in the western Amazonia. No significant relationship is found between maximum precipitation intensity and the area of windthrows.

A significant fraction of anthropogenic CO2 emissions is assimilated by tropical forests and stored as biomass, slowing the accumulation of CO2 in the atmosphere. Because different plant tissues have different functional roles and turnover times, predictions of carbon balance of tropical forests depend on how earth system models (ESMs) represent the dynamic allocation of productivity to different tree compartments. This study shows that observed allocation of productivity, biomass, and turnover times of main tree compartments (leaves, wood, and roots) are not accurately represented in Coupled Model Intercomparison Project Phase 5 ESMs. In particular, observations indicate that biomass saturates with increasing productivity. In contrast, most models predict continuous increases in biomass with increases in productivity. This bias may lead to an over-prediction of carbon uptake in response to CO2 or climate-driven changes in productivity. Compartment-specific productivity and biomass are useful benchmarks to assess terrestrial ecosystem model performance. Improvements in the predicted allocation patterns and turnover times by ESMs will reduce uncertainties in climate predictions.

Forest mortality caused by convective storms (windthrow) is a major disturbance in the Amazon. However, the linkage between windthrows at the surface and convective storms in the atmosphere remains unclear. In addition, the current Earth system models (ESMs) lack mechanistic links between convective wind events and tree mortality. Here we find an empirical relationship that maps convective available potential energy, which is well simulated by ESMs, to the spatial pattern of large windthrow events. This relationship builds connections between strong convective storms and forest dynamics in the Amazon. Based on the relationship, our model projects a 51 ± 20% increase in the area favorable to extreme storms, and a 43 ± 17% increase in windthrow density within the Amazon by the end of this century under the high-emission scenario (SSP 585). These results indicate significant changes in tropical forest composition and carbon cycle dynamics under climate change.

Earthquakes can produce significant tree mortality, and consequently affect regional carbon dynamics. Unfortunately, detailed studies quantifying the influence of earthquake on forest mortality are currently rare. The committed forest biomass carbon loss associated with the 2008 Wenchuan earthquake in China is assessed by a synthetic approach in this study that integrated field investigation, remote sensing analysis, empirical models and Monte Carlo simulation. The newly developed approach significantly improved the forest disturbance evaluation by quantitatively defining the earthquake impact boundary and detailed field survey to validate the mortality models. Based on our approach, a total biomass carbon of 10.9 Tg·C was lost in Wenchuan earthquake, which offset 0.23% of the living biomass carbon stock in Chinese forests. Tree mortality was highly clustered at epicenter, and declined rapidly with distance away from the fault zone. It is suggested that earthquakes represent a significant driver to forest carbon dynamics, and the earthquake-induced biomass carbon loss should be included in estimating forest carbon budgets.

Storage carbon (C) pools are often assumed to contribute to respiration and growth when assimilation is insufficient to meet the current C demand. However, little is known of the age of stored C and the degree to which it supports respiration in general. We used bomb radiocarbon ((14)C) measurements to determine the mean age of carbon in CO2 emitted from and within stems of three tropical tree species in Peru. Carbon pools fixed >1 year previously contributed to stem CO2 efflux in all trees investigated, in both dry and wet seasons. The average age, i.e., the time elapsed since original fixation of CO2 from the atmosphere by the plant to its loss from the stem, ranged from 0 to 6 years. The average age of CO2 sampled 5-cm deep within the stems ranged from 2 to 6 years for two of the three species, while CO2 in the stem of the third tree species was fixed from 14 to >20 years previously. Given the consistency of (14)C values observed for individuals within each species, it is unlikely that decomposition is the source of the older CO2. Our results are in accordance with other studies that have demonstrated the contribution of storage reserves to the construction of stem wood and root respiration in temperate and boreal forests. We postulate the high (14)C values observed in stem CO2 efflux and stem-internal CO2 result from respiration of storage C pools within the tree. The observed age differences between emitted and stem-internal CO2 indicate an age gradient for sources of CO2 within the tree: CO2 produced in the outer region of the stem is younger, originating from more recent assimilates, whereas the CO2 found deeper within the stem is older, fueled by several-year-old C pools. The CO2 emitted at the stem-atmosphere interface represents a mixture of young and old CO2. These observations were independent of season, even during a time of severe regional drought. Therefore, we postulate that the use of storage C for respiration occurs on a regular basis challenging the assumption that storage pools serve as substrates for respiration only during times of limited assimilation.

Leaf area index (LAI) is a key parameter that affects the surface fluxes of energy, mass, and momentum over vegetated lands, but observational measurements are scarce, especially in remote areas with complex canopy structure. In this paper we present an indirect method to calculate the LAI based on the analyses of histograms of hemispherical photographs. The optimal threshold value (OTV), the gray-level required to separate the background (sky) and the foreground (leaves), was analytically calculated using the entropy crossover method (Sahoo, P.K., Slaaf, D.W., Albert, T.A., 1997. Threshold selection using a minimal histogram entropy difference. Optical Engineering 36(7) 1976–1981). The OTV was used to calculate the LAI using the well-known gap fraction method. This methodology was tested in two different ecosystems, including Amazon forest and pasturelands in Brazil. In general, the error between observed and calculated LAI was ∼6%. The methodology presented is suitable for the calculation of LAI since it is responsive to sky conditions, automatic, easy to implement, faster than commercially available software, and requires less data storage.

Windthrows (trees uprooted and broken by winds) are common across the Amazon. They range in size from single trees to large gaps that lead to changes in forest dynamics, composition, structure, and carbon balance. Yet, the current understanding of the spatial variability of windthrows is limited. By integrating remote sensing data and geospatial analysis, we present the first study to examine the occurrence, area, and direction of windthrows and the control that environmental variables exert on them across the whole Amazon. Windthrows are more frequent and larger in the northwestern Amazon (Peru and Colombia), with the central Amazon (Brazil) being another hot spot of windthrows. The predominant direction of windthrows is westward. Rainfall, surface elevation, and soil characteristics explain the variability (20%-50%) of windthrows but their effects vary regionally. A better understanding of the spatial dynamics of windthrows will improve understanding of the functioning of Amazon forests.