Satellite observations show how tropical forest carbon fluxes respond to changes in water from climate variability. A recent article in Reviews of Geophysics focuses on satellite-derived information on terrestrial carbon, and water storage and fluxes, with specific reference tropical regions. Here, the some of authors answer our questions about what has been observed and what this tells us about the tropical carbon cycle and its interaction with climate variability.
How have human activities in the tropics altered the carbon balance, and what have been some of the major the impacts on Earth’s climate?
Tropical forests contribute significantly more than other forests to the year-to-year variability of global terrestrial carbon balance. Human activities have both direct and indirect effects on the tropical forest carbon balance.
Direct effects include large-scale deforestation and fragmentation of tropical forests for clearing land for cultivation and grazing or extracting wood for timber and fire; these are largely propelled by economic drivers.
Indirect effects are mostly due to climate variations that introduce widespread anomalies of rainfall, resulting in severe and frequent droughts. These climate variations further impact the carbon cycling of tropical ecosystems, often by reducing forest productivity and carbon uptake and increasing tree mortality and carbon emissions.
Geographically, tropical forests cover a vast region across the globe that, unlike forests in the northern hemisphere, remain less managed, with limited access from ground and air.
Observations from satellites since the early 1980s have provided information about changes of climate and forest cover, allowing scientists to understand and model changes of tropical forest carbon cycle. Now, a new generation of satellites equipped with advanced technologies are providing measurements of forest structure, carbon stocks, productivity, and other carbon and water fluxes that have substantially changed our understanding the role of tropical forests in local and global climate.
New and emerging patterns from these measurements suggest that climate is increasingly becoming as important as deforestation in affecting the capability of tropical forests to take up atmospheric carbon dioxide.
These forests show more vulnerability to water stress as perceived in the past. This vulnerability is also not uniform across tropics and varies across continents and regions suggesting other hidden factors that moderate how changes in water variability affect carbon cycling.
The tropical carbon and water cycles are composed of many interlinked parts. Carbon is traded for water from the roots through plant stomata, and this transfer of carbon for water in turn depends on radiation, atmospheric humidity and CO2, soil moisture, and nutrients. Carbon is allocated to different parts of a tree (e.g., leaves, trunk, roots), while water is typically stratified in the soil depending on soil type, rooting depth, and amount of rainfall. The allocation of carbon to these different parts of a tree depends on the amount of water and where it is located.
The constellation of satellites now in orbit provide information on many parts of this interlinked system, including photosynthesis, atmospheric humidity, water variability in the soil column, rainfall, evapotranspiration, fires, and the total exchange of carbon dioxide between surface and atmosphere. Ecosystem models and measurements of forest structure and hydraulics are needed to piece together the puzzle of this interlinked problem.
From these measurements, we have learned that the carbon and water cycles in the forests of the Amazon are distinctly different than the forests of the Congo region and those in Asia, although following the same basic physical and ecological processes. Satellite observations have played a critical role in quantifying these variations of carbon and water cycles across the tropics.
How might this information be used to predict future trends?
The future distribution of carbon (trees, plants, roots, litter, soil carbon) and water in the tropics depends on the sensitivity of the different carbon and water pools to environmental drivers such as temperature, rainfall, and CO2 fertilization, convolved with the changes in temperature, rainfall, and CO2 that might occur with climate change. Satellite data can help quantify this sensitivity so that, given possible changes in these environmental drivers, we have an estimate of the future state of tropical carbon and water.
What are some limitations of the current modeling system in understanding the carbon and water cycles?
Different models can have the same amount of photosynthesis and respiration but for very different reasons because they allocate carbon and water differently to different pools (e.g. roots, leaves, soil moisture) or because they have slightly different methods for modeling the transfer of carbon and water through these pools. This equifinality leads to different sensitivities of atmospheric CO2 to changes in water, temperature, and the CO2 fertilization effect. Most models do not have the flexibility to change the carbon and water pools and corresponding exchange processes when confronted with new data, especially data that have different information about carbon and water states such as photosynthesis, total water storage, fire emissions, above ground biomass, and net biosphere exchange.
This reduced flexibility is a result of the extensive number of model parameters, which need to be “tuned” to ensure that it reproduces current observed atmospheric and terrestrial states such as leaf area index. Consequently, while we can learn about possible combinations of processes controlling the carbon and water cycle from existing state-of-the art models, it is challenging to learn about the most accurate combination from existing models. Ultimately, optimizing model parameters to ensure models match the wide array of in-situ and satellite measurements is a key to solving the equifinality problem. For present-day models, their large computational costs and complexity of the models have made parameter “optimization” exceedingly difficult.
What could be done to improve future models examining the relationship between the carbon and water cycles?
Satellite and ground observations, augmented with data collected from aircraft, can now measure a wealth of different carbon and water state variables or fluxes, which can in turn infer the processes and reservoirs that control the carbon and water cycles. However, models are needed that can integrate these new, extensive data sets not just for matching the state variables but also to update the processes controlling these state variables. These models need to be set up in such a way so that carbon and water processes and reservoirs can be adjusted to fit observations in a statistically robust (Bayesian) manner. New computational methods—including parallel computing and state-of-the-art model-data fusion techniques—are increasingly being used, along with satellite measurements spanning multiple decades—to tackle the challenge of model parameter optimization.
–John Worden, (john.r.worden@jpl.nasa.gov, 
0000-0003-0257-9549), Sassan Saatchi, and Anthony Bloom, Jet Propulsion Laboratory / California Institute for Technology, USA