Interplay of Turbulence Regimes, Deep Convection, and Tree Mortality in the Amazon Rainforest

The Science
With focus in the Central Amazon (Fig A), at the Tropical Silviculture Experimental Station -EEST (located about 60 km northwest of Manaus, 2°36′S, 60°12′W, 130 m ...) also known as ZF2, we investigated the influence of seasonality and proximity to the forest canopy on nocturnal turbulence regimes in the roughness sublayer (Fig B). Since convective systems of different scales are common in this region (Fig C), we also analyzed the effect of extreme wind gusts (propagated from convective downdrafts) (Fig D) on the organization of the turbulence regimes, and their potential to cause the mortality of canopy trees.

The Impact
Two different turbulence regimes were identified at three heights above the canopy: a weakly stable (WS) and a very stable regime (VS). The threshold wind speeds that mark the transition between turbulence regimes were larger during the dry season and increased as a function of the height above the canopy. Downdrafts occurred only in the WS and favored a fully coupled state of wind flow along the canopy profile.

Summary
Our study data include high-frequency winds, temperature and ozone concentration at different heights during the dry and wet seasons of 2014. In addition, we used critical wind-speed data derived from a tree-winching experiment and a modeling study conducted in the same study site. This study provides three novel contributions. The first was the identification of different turbulence regimes and their patterns in terms of seasonality and proximity to the forest canopy in the nocturnal roughness sublayer. The second was the assessment of the effects of near surface wind gusts (propagated from downdrafts) on the organization of turbulence regimes. Finally, it provides evidence on the occurrence of extreme wind gusts associated with convective downdrafts, with potential to promote damage and mortality of canopy trees. These aspects highlight the strong interactions between atmospheric and biospheric processes and mechanisms regulating forest structure and dynamics.

Figure. (A) Study area – Central Amazon, and high frequency sensor. (B) Relationship between turbulence velocity scale () and mean wind speed ( ) at different heights in the dry and wet season during the nighttime. Triangles indicate the threshold wind speed () at which the very stable changed to weakly stable regime. (C) GOES 13 image during 13 April 2014 at 0300 UTC when downdrafts reached the micrometeorology tower at the study area (red dot). (D) horizontal wind speed () over the sensor during 13 April.

Contact: Daniel Magnabosco Marra (Max Planck, Biogeochemistry), dmarra@bgc-jena.mpg.de

Funding
R Negron-Juarez was supported by the Next Generation Ecosystem Experiments-Tropics (NGEE-Tropics), funded by the U.S. Department of Energy, Office of Science. This study is part of the Wind–Tree Interaction Project (INVENTA) and the Amazon Tall Tower Observatory (ATTO), funded by the German Federal Ministry of Education and Research (BMBF, contracts 01LB1001A and 01LK1602A), the Brazilian Ministry of Science, Technology and Innovation (MCTI/FINEP, contract 01.11.01248.00) and the Max Planck Society (MPG), Germany.

Publication: Mendoca A, Dias-Junior C, Acevedo O, Santana R, Costa F, Negron-Juarez R, Manzi A, Trumbore S, Magnabosco Marra D. Agricultural and Forest Meteorology, 339, 109526 (2023), DOI: https://doi.org/10.1016/j.agrformet.2023.109526

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.

The Science
Fan-shaped dead forest patches were found over the entire Amazonia. These patches, over 37 hectares, affect the role the Amazon forests played in the world’s carbon cycle. Scientists found that frequent storms result in these dead forest patches, but how does the process happen? In this study, we explored the three characteristics of storms, including their passing over time, cloud top temperature, and associated precipitation, to identify their relationship with the size of the dead forests. We found that long-lived storms with thicker and tall clouds, result in bigger sizes of dead forest patches. Moreover, forests in the western Amazonia are more vulnerable to storms than forests on the other parts of the Amazonia.

The Impact
My research explores how extreme storms impact tree loss in tropical forests, especially in the Amazon. These storms, responsible for 50-90% of annual rainfall in the tropics, often result in toppling trees, which disrupts the forest’s ability to store carbon, a crucial ability to fight climate change. Previously, these phenomena were studied separately, but my work connects them. By analyzing satellite data, I’ve started uncovering relationships between the characteristics of extreme storms and tree loss sizes. This understanding can improve climate models, providing more accurate predictions about our changing environment.

Summary
Our research delved into the relationship between large-scale storm systems known as mesoscale convective systems (MCSs) and the phenomenon of ‘windthrow’ – when storms uproot trees – in the Amazon rainforest. We examined 38 pairs of windthrow and their associated MCS events to identify the specific storm characteristics influencing the extent of windthrow. We found that MCSs with a longer storm duration tended to result in more extensive windthrow. We found a positive correlation between the storm’s duration and the area of forest affected. The depth of convection clouds within the storm also played a role. Deep convection caused larger windthrow across the entire Amazon. In contrast, shallow convection led to medium-sized windthrows in western Amazonia and smaller ones in central Amazonia. Interestingly, we observed that rainfall wasn’t uniformly distributed among forest disturbances of the same size, suggesting the need for more precise precipitation data to establish a clearer relationship with windthrow sizes. This study offers detailed case studies on windthrows and corresponding MCS features. It helps reduce the uncertainty of previous research due to data mismatches between MCSs and windthrows, offering fresh insights into how land and atmosphere interact. These findings are important for refining our climate models and, ultimately, our understanding of climate change impacts on the ecosystem.

Figure. The workflow of correlating a windthrow event on the land surface to its associated convective storm in the atmosphere using remote sensing images from both land and meteorological satellites. Image courtesy of Yanlei Feng.

Contact: Yanlei Feng (UC Berkeley PhD graduates, currently at Carnegie Institute for Science), ylfeng@berkeley.edu

Funding
This research was supported as part of the Next Generation Ecosystem Experiments-Tropics, funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research under contract number DE-AC02-05CH11231. We acknowledge the World Climate Research Programme, which, through its Working Group on Coupled Modelling, coordinated and promoted CMIP6. We thank the climate modeling groups for producing and making available their model output, the Earth System Grid Federation (ESGF) for archiving the data and providing access, and the multiple funding agencies who support CMIP6 and ESGF.

Publication: Feng, Y., Negrón‐Juárez, R.I., Chiang, J.C. and Chambers, J.Q., 2023. Case studies of forest windthrows and mesoscale convective systems in Amazonia. Geophysical Research Letters, 50(12), p.e2023GL104395. https://doi.org/10.1029/2023GL104395

ForestGEO scientists quantified the proportion of total biomass losses from damaged but surviving trees across seven tropical forests.

The Science
Damage (i.e., branchfall, trunk breakage, wood decay) is a ubiquitous feature in forest ecosystems. Yet, traditional forest inventories assume tree mortality as the only source of biomass losses. While previous studies have shown that damage is an important condition preceding tree death, the contribution of non-lethal damage (i.e., from surviving trees) to total forest biomass (and therefore carbon) losses remained unclear. ForestGEO scientists combined field-based measurements of tree completeness with vertical volume profile models obtained from terrestrial laser scanning to show that 42% (range 12%–76% across forests) of total aboveground biomass loss is due to damage to living trees across seven tropical forests.

The Impact
Ground-based biomass stocks and fluxes are widely used to estimate carbon budgets, to quantify forest carbon offsets, and to calibrate and validate remote sensing products employed to obtain biomass estimates at regional and global scales. This study shows that biomass loss from damage to living trees constitutes an important and overlooked component of biomass loss. These results contrast with the typically low forest biomass losses estimated only from tree mortality and suggest that forest carbon turnover may be higher than previously thought. Since forest disturbance rates are expected to increase under climate change the biomass loss to damage is likely to become more important.

Summary
Forest carbon losses constitute a significant source of uncertainty in vegetation models. These estimates are typically calculated based on the biomass of dead trees, without accounting for losses via damage to living trees: branchfall, trunk breakage, wood decay. In this study, forest ecologists employ multiple annual records of tree survival and structural completeness to compare aboveground biomass (AGB) loss via damage to living trees to total AGB loss (mortality + damage) in seven tropical forests widely distributed across environmental conditions. Researchers find that 42% (3.62 Mg ha-1 yr-1; 95% CI 2.36–5.25) of total AGB loss (8.72 Mg ha-1 yr-1; CI 5.57–12.86) is due to damage to living trees. They also find that conventional forest inventories overestimate stand-level AGB stocks by 4% (1-17% range across forests) because assume structurally complete trees, underestimate total AGB loss by 29% (6-57%) due to overlooked damage-related AGB losses, and overestimate AGB loss via mortality by 22% (7-80%) because of the assumption that trees are undamaged before dying. These results indicate that forest carbon fluxes are higher than previously thought. Damage on living trees is an underappreciated component of the forest carbon cycle that is likely to become even more important as the frequency of forest disturbances increases.

Figure. Damaged but surviving trees are highly frequent across forest ecosystems. Photo credits: C.Y. Chia-Hao (left), D. Zuleta (center, right).

Contact: Daniel Zuleta, Ph.D., Ecologist (Postdoctoral fellow) at Forest Global Earth Observatory, Smithsonian Tropical Research Institute in Washington, DC (dfzuleta@gmail.com)

Funding
This project was supported as part of the Next Generation Ecosystem Experiments–Tropics and was funded by the Office of Biological and Environmental Research (BER) within the U.S. Department of Energy’s (DOE) Office of Science. Data collection was supported by the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Institution.

Publication: Zuleta, D., et al. “Damage to living trees contributes to almost half of the biomass losses in tropical forests” Global Change Biology in press (2023). https://doi.org/10.1111/gcb.16687.

Drivers and impacts of timber extraction, fires, drought and habitat fragmentation

The Science
This review paper analyzed forest degradation processes using existing data that included fire, edge effects, extreme drought, and timber extraction between 2001 and 2018 in Amazon forest. The synthesis focused on causes and impacts, possible futures, and some interventions required to slow these forest degradation processes. The analyses demonstrated that the combined effects of all degradation processes impacts an area larger than the total area deforested during this same period. Forest carbon loss from these processes is also comparable to carbon loss from direct deforestation.

The Impact
The analysis highlights specific processes that can be targeted to reduce the negative effects of forest degradation processes, including carbon emissions that occur beyond the direct effects of deforestation. Reducing degradation processes will require engaging with a diverse set of stakeholders, the monitoring of different disturbance processes, and the development of appropriate policies and programs.

Summary
Degradation processes in the Amazon, including areas degraded by fire, edge effects, timber extraction, and extreme drought are impacting an area equivalent to 38% of all remaining forests in the region, with a related carbon emission of 0.06 to 0.21 Pg C yr-1. This degradation also reduces dry-season evapotranspiration by up to 34%, and results in large losses to biodiversity. Even with reductions in deforestation rates, degradation will remain a dominant source of carbon emissions primarily due continued degradation of forests along the edges of agricultural areas. The quantification of carbon, water and energy balance changes from land-use activities in tropical forests need to be included in Earth system models to accurately account for second order degradation processes.

Contacts: Brian Benscoter (DOE BER Program Manager) at Brian.Benscoter@science.doe.gov; Jeff Chambers (Lawrence Berkeley National Laboratory) at jchambers@lbl.gov

Funding
Jeff Chambers and Charlie Koven from LBNL were supported as part of the Next Generation Ecosystem Experiments–Tropics to participate in the associated meeting in Manaus, Brazil that led to the development of this review paper.

Publications
Lapola, D.M., et al. (2023) The drivers and impacts of Amazon forest degradation, Science, 379. https://doi.org/10.1126/science.abp8622.

Representing diurnal shifts in leaf-level water use efficiency may be key to modeling tropical forest gas exchange

The Science
To understand how tropical ecosystems will respond to global change, researchers must correctly represent the relationship between water loss and carbon gain in leaves, known as the water use efficiency (WUE). There are still significant uncertainties associated with the dynamics of WUE over different timescales, from over the day, to the changes experienced over the full lifespan of a leaf. Here we collected data to assess the possible physiological and mechanistic factors which influence WUE dynamics. While WUE does differ between leaves of different phenological stages, the trend was not consistent across species. However, we identified a unidirectional increase in WUE of approximately 2.5 times over the course of the day in five of the six species studied.

The Impact
One of the major roadblocks to accurate representation of transpiration in climate models is an understanding of the physiological factors which most strongly contribute to variation in leaf level WUE. In this study, we demonstrate that including leaf age as a primary driver of WUE did not help to improve or explain variation in modeled transpiration. However, models which accounted for diurnal (within-day) changes in WUE improved the representation of transpiration. These findings provide a roadmap for future investigation into the physiological traits which most strongly influence transpiration over space and time and underscore the need to closely consider the model assumptions, like constant WUE, implicit in many of the models used to project the future of tropical forests.

Summary
A primary source of uncertainty in terrestrial biosphere model projections of ecosystem-scale carbon uptake and water cycling is the relationship between CO₂ assimilation and water loss via stomatal conductance. In models, this relationship is governed by two terms, the stomatal slope (g₁) and intercept (g₀). Accurate mechanistic representation of how the g₁ and g₀ parameters vary over time is crucial, particularly in wet tropical broadleaf forests where trees have a near consistent annual pattern of leaf production and senescence, and precipitation and humidity are strongly seasonal. These stomatal parameters are estimated using leaf-level gas exchange by two alternative methods: (1) a response curve where the environmental conditions are modified for a single leaf, or (2) a survey approach, where repeated measurements are made on multiple leaves over a diurnal range of environmental conditions.

In this study we found that stomatal response curves and survey style measurements produce statistically different estimations of stomatal parameters, which resulted in large (between 26% and 125%) differences in simulated fluxes of water. Furthermore, we found that g₁ varies both diurnally and to a lesser degree with leaf age. Taken together, these results show that models which use stomatal parameters derived from response curves significantly underestimated canopy level transpiration, and that while leaf traits do vary among leaf phenological stage, models, which tend to only include mature vegetation parameterizations, perform similarly to those that explicitly simulate three leaf age stages.

Contact: Kenneth Davidson, Brookhaven National Lab (kdavidson@bnl.gov)

Funding
This work was supported by the Next-Generation Ecosystem Experiments (NGEE) – Tropics project, which is funded by the Biological and Environmental Research (BER) Program within the US Department of Energy’s (DOE) Office of Science and through DOE contract no. DE-SC0012704 to Brookhaven National Laboratory.

Publications
Davidson KJ, Lamour J, Rogers A, Ely KS, et al. “Short-term variation in leaf-level water use efficiency in a tropical forest” New Phytologist, 237, 2069-2087, (2023), [DOI: 10.1111/nph.18684]

The importance of plant water and carbon dynamics to fire behavior and effects and a framework to link remotely sensed estimates to fire models.

The Science
The condition of living woody plants can change fire behavior. Plants can have different levels of dryness throughout the seasons and in different parts of a landscape based water and nutrients in the soil. Lower levels of live fuel moisture in plants can be linked to faster fires and change the way fires burn and how likely plants are to die after a fire. Using live fuel moisture measurements from remote sensing tools, such as airborne systems, and linking these to models of fire behavior and effects we can improve our understanding of how fires may change in the future.

The Impact
Fire behavior models have long used general fuels in broad groups. Now with new types of models and remote sensing measurements of fuels and fires, we can capture more realistic fuels and how they change in both their condition, such as live fuel moisture, and their structure. This information is critically important for fire management in conditions of drought and warming. Linking how living plants change through time and across a landscape to how fires might behave will give us information we need to better support communities in a world with more fire.

Summary
Wildfires have been recognized as a global crisis, but current fire models do not capture how living plants change in response to changing climate. With drought and warming temperatures increasing the importance of living plants as a factor in changing fire behavior, and new capabilities of models, we are able to capture these complex processes and interactions. We provide a renewed focus on capturing live woody plants in fire models. Living plant conditions and properties influence fire combustion and heat transfer and often dictate if a plant will survive. These interactions provide a mechanistic link between living plants and fire behavior and effects that can be included in new models. We include a conceptual framework linking remotely sensed estimates of plant condition to models of fire behavior and effects, which could be a crucial first step toward improving models used for global fire forecasting. This process-based approach will be essential to capturing the influence of physiological responses to drought and warming on live fuel conditions, strengthening the science needed to guide fire managers in an uncertain future.

Contact: L.T. Dickman, Los Alamos National Laboratory, lee@lanl.gov

Funding
LTD, AJ, RRL and SS were supported by the Los Alamos National Laboratory (LANL) through its Center for Space and Earth Science (CSES). Center for Space and Earth Science is funded by LANL’s Laboratory Directed Research and Development (LDRD) program under project no. 20210528CR. AJ and ZJR received additional funding from LANL LDRD under project no. 20210689ECR. RPF and STM were supported by SERDP project RC18-1346 and an NSERC Discovery Grant. AB acknowledges funding from the Austrian Science Fund (FWF, project P32203) and from the University of Innsbruck (Early-Stage Funding, grant W-171705). MY receives funding from the Australian Research Council, the Australian Research Data Commons, The SmartSat Cooperative Research Centre and Singtel Optus Pty Limited. JKS was supported by the National Center for Atmospheric Research, a major facility sponsored by the National Science Foundation (NSF) under Cooperative Agreement no. 1852977, with additional support from NASA Arctic Boreal Vulnerability Experiment Grant 80NSSC19M0107. JKS, SPS and CX were also supported as part of the Next-Generation Ecosystem Experiments – Tropics, funded by the US Department of Energy, Office of Science, Office of Biological and Environmental Research. SPS was also partially supported by the NASA Surface Biology and Geology Mission Study (NNG20OB24A) and through the United States Department of Energy contract no. DE-SC0012704 to Brookhaven National Laboratory. CMH acknowledges US Department of Defense (DoD) Strategic Environmental Research and Development Program (SERDP) Project RC19-1119. IA declares support from NSF WIFIRE Commons under grants 2040676 and 2134904. VRD acknowledges funding from MICINN projects RTI2018-094691-B-C31; EU H2020 (grant agreements 101003890). RAP acknowledges support from US Department of Defense Strategic Environmental Research and Development Program’s Closing Gaps Project RC20-1025. USDA Forest Service personnel were supported by annual Forest Service appropriations.

Publication: L.T. Dickman, et al., “Integrating plant physiology into simulation of fire behavior and effects.”, New Phytologist (2023). nph.18770, [https://doi.org/10.1111/nph.18770].