Search NASA⌕ Search

SEARCH · Search NASA

Results for “Forest management”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
DOE OSTI2020

Soil organic carbon accumulation rates on Mediterranean abandoned agricultural lands

Secondary succession on abandoned agricultural lands can produce climate change mitigation co-benefits, such as soil carbon sequestration. However, the accumulation of soil organic carbon (SOC) in Mediterranean regions has been difficult to predict and is subject to multiple environmental and land management factors. Gains, losses, and no significant changes have all been reported. Here we compile chronosequence data (n = 113) from published studies and new field sites to assess the response of SOC to agricultural land abandonment in peninsular Spain. In this work we found an overall SOC accumulation rate of +2.3% yr -1 post-abandonment. SOC dynamics are highly variable and context-dependent. Minimal change occurs on abandoned cereal croplands compared to abandoned woody croplands (+4% yr -1 ). Accumulation is most prevalent within a Goldilocks climatic window of ~13–17 °C and ~450–900 mm precipitation, promoting >100% gains after three decades. Our secondary forest field sites accrued 40.8 Mg C ha -1 (+172%) following abandonment and displayed greater SOC and N depth heterogeneity than natural forests demonstrating the long-lasting impact of agriculture. Although changes in regional climate and crop types abandoned will impact future carbon sequestration, abandonment remains a low-cost, long-term natural climate solution best incorporated in tandem with other multipurpose sustainable land management strategies.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Modeling Spatial Distribution of Snow Water Equivalent by Combining Meteorological and Satellite Data with Lidar Maps

Abstract An accurate characterization of the water content of snowpack, or snow water equivalent (SWE), is necessary to quantify water availability and constrain hydrologic and land surface models. Recently, airborne observations (e.g., lidar) have emerged as a promising method to accurately quantify SWE at high resolutions (scales of ∼100 m and finer). However, the frequency of these observations is very low, typically once or twice per season in the Rocky Mountains of Colorado. Here, we present a machine learning framework that is based on random forests to model temporally sparse lidar-derived SWE, enabling estimation of SWE at unmapped time points. We approximated the physical processes governing snow accumulation and melt as well as snow characteristics by obtaining 15 different variables from gridded estimates of precipitation, temperature, surface reflectance, elevation, and canopy. Results showed that, in the Rocky Mountains of Colorado, our framework is capable of modeling SWE with a higher accuracy when compared with estimates generated by the Snow Data Assimilation System (SNODAS). The mean value of the coefficient of determination R 2 using our approach was 0.57, and the root-mean-square error (RMSE) was 13 cm, which was a significant improvement over SNODAS (mean R 2 = 0.13; RMSE = 20 cm). We explored the relative importance of the input variables and observed that, at the spatial resolution of 800 m, meteorological variables are more important drivers of predictive accuracy than surface variables that characterize the properties of snow on the ground. This research provides a framework to expand the applicability of lidar-derived SWE to unmapped time points. Significance Statement Snowpack is the main source of freshwater for close to 2 billion people globally and needs to be estimated accurately. Mountainous snowpack is highly variable and is challenging to quantify. Recently, lidar technology has been employed to observe snow in great detail, but it is costly and can only be used sparingly. To counter that, we use machine learning to estimate snowpack when lidar data are not available. We approximate the processes that govern snowpack by incorporating meteorological and satellite data. We found that variables associated with precipitation and temperature have more predictive power than variables that characterize snowpack properties. Our work helps to improve snowpack estimation, which is critical for sustainable management of water resources.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Effects of spatial variability in vegetation phenology, climate, landcover, biodiversity, topography, and soil property on soil respiration across a coastal ecosystem

Coastal terrestrial-aquatic interfaces (TAIs) are crucial contributors to global biogeochemical cycles and carbon exchange. A systematic evaluation of the interaction between coastal catchment properties and carbon dioxide (CO2) emission by soil respiration is significant for assessing carbon dynamics and predicting the future trajectory of atmospheric CO2 concentrations in coastal TAIs. The soil CO2 efflux in these transition zones is however poorly understood due to the high spatiotemporal dynamics of TAIs, as various sub-ecosystems in this region are compressed and expanded by complex influences of tides, changes in river levels, climate, and land use. We focus on the Chesapeake Bay region to (i) investigate the spatial heterogeneity of the coastal ecosystem and identify spatial zones with similar environmental characteristics based on the spatial data layers, including vegetation index (kNDVI), climate, landcover, diversity, topography, soil property, and relative tidal elevation; (ii) understand the primary driving factors affecting soil respiration within sub-ecosystems of the coastal ecosystem. Specifically, we employed hierarchical clustering analysis to identify spatial regions with distinct environmental characteristics, followed by the determination of main driving factors using Random Forest regression and SHapley Additive exPlanations. Maximum and minimum temperature are the main drivers common to all sub-ecosystems, while each region also has additional unique major drivers that differentiate them from one another. Precipitation exerts an influence on vegetated lands, while soil pH value holds importance specifically in forested lands. In croplands characterized by high clay content and low sand content, the significant role is attributed to bulk density. Wetlands demonstrate the importance of both elevation and sand content, with clay content being more relevant in non-inundated wetlands than in inundated wetlands. The topographic wetness index significantly contributes to the mixed vegetation areas, including shrub, grass, pasture, and forest. Additionally, our research reveals that dense vegetation land covers and urban/developed areas exhibit distinct soil property drivers. Overall, there is no one-size-fits-all approach to modeling carbon fluxes in coastal TAIs, and our study highlights the importance of further research and monitoring practices to improve our understanding of carbon dynamics and promote the sustainable management of coastal TAIs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

Functional but not taxonomic diversity increases productivity of Populus in the southeastern United States

Plant interactions like competition and facilitation impact ecosystem function and resilience. Improving our understanding of the relationships between these interactions and community productivity has important implications for managers of production systems in forestry and agriculture as well as conservation science. Populus spp. are an excellent model system for exploring how inter‐ and intraspecific interactions impact ecosystem functions, such as productivity, in forest plantations. In this study, we compared aboveground productivity of six Populus clones from three different taxa grown in monoclonal and mixed‐clonal plots. The different mixture treatments were intended to experimentally test aboveground biomass response to contrasting levels of taxonomic diversity and functional diversity based on nitrogen use characteristics of Populus clones. We hypothesize that functional diversity would be more important than taxonomic diversity in increasing aboveground productivity of mixed‐clonal plantings compared to monocultures. In addition, a subset of treatments was carried out on additional sites representing a productivity gradient in order to determine if the relationship between biodiversity and productivity in these systems diminished at more productive sites as suggested by the stress‐gradient hypothesis. We found that functionally diverse mixtures of clones had greater yield of aboveground biomass than the average of their constituent monocultures, while more taxonomically diverse mixes of clones did not differ from the average of their constituent monocultures. However, when reestablished on sites with extremely high or low productivity, the best performing clone mixture also did not differ from the average of its constituent monocultures. Our results suggest that intimate clone mixtures of Populus have the potential to significantly increase productivity, but results vary by mixture and by site. To capitalize on positive biodiversity effects on yield in production systems, targeted mixtures based on divergent functional traits linked to different use and acquisition strategies for site‐specific limiting resources are most likely to be successful.

BEF↗
DOE OSTI2026

A two-stage optical fusion framework for wildfire severity mapping across the conterminous United States

Accurate wildfire severity mapping (WSM) is essential for post-fire recovery planning, erosion risk assessment, ecosystem monitoring, and disaster risk reduction. Although Landsat and Sentinel optical imagery have been widely used for burn severity assessment, the added value of fusing multiple optical sensors has not been sufficiently quantified across diverse fire events, particularly since the launch of Landsat-9. This study evaluates whether multisensor optical fusion improves wildfire severity mapping relative to single-sensor baselines using Sentinel-2, Landsat-8, and Landsat-9 imagery across 40 wildfire events in the conterminous United States. We tested a two-stage fusion framework that combines feature-level fusion with pixel-level dimensionality reduction. First, feature-level fused datasets were created through early fusion by combining standardized post-fire bands from each sensor into a single predictor stack. Both raw reflectance bands and pairwise spectral transforms were retained to capture within- and cross-sensor spectral interactions. Second, Linear Discriminant Analysis was applied to both single-sensor and fused datasets to produce comparable low-dimensional feature spaces. Six machine-learning classifiers were then used to benchmark model performance with repeated spatially buffered train–test splits. Results show that Landsat-9 was the strongest single-sensor baseline. Among the fusion strategies, Sentinel-2 + Landsat-9 produced the most consistent improvement and reduced performance variability. Landscape-condition analysis further showed that this fusion was most beneficial in shrubland-dominated and high-terrain fires, where it achieved the highest overall mean accuracy and the fewest failures. In contrast, its benefits were less reliable in evergreen forests, mixed vegetation, and low- to moderate-elevation terrain. In operational settings, the Sentinel-2 + Landsat-9 configuration offers a practical solution for post-fire recovery planning, erosion-risk assessment, watershed management, and ecological monitoring when field observations are available and timely satellite-based information is needed.

Landsat↗
DOE OSTI2022

Revegetation affects the response of land surface phenology to climate in Loess Plateau, China

Land surface phenology (LSP), defined as the plant's growth rhythm retrieved from satellite sensing products, is proven to shift with climate change and affect the carbon cycles of terrestrial ecosystems. Global afforested area is largely increasing and consequently affecting local and global climate. However, how and to what extent revegetation affects LSP remains relatively unexplored. Here we investigated the difference in four LSPs (i.e., greenup, maturity, senescence, and dormancy) and the response of LSP to climate between restored and native vegetation on Loess Plateau, China, where a remarkable process of vegetation restoration happened during 1982–2015. Most study regions showed a longer growing season (LOS) over time, specifically, with a slight delay in greenup but a relatively large delay in senescence. We found that air temperature was the dominant factor affecting greenup and maturity, while precipitation mostly controlled the senescence and dormancy in the study area. Under similar climate conditions, the LSP of restored vegetation (i.e., restored forest and grassland) showed a significant difference (p < 0.05) from native ones during 1999–2015. Compared to the native forest, restored forest from cropland and grassland showed a delayed greenup date by 0.3 and 3.6 days (p < 0.05) and an advanced dormancy date of 6.6 and 9.0 days (p < 0.05), respectively. Furthermore, the restored vegetation became less sensitive to air temperature than native vegetation, while the restored forest was more sensitive to precipitation, and its growth was affected by the water limitation to a larger extent in the study area. In conclusion, our study highlights the necessity of considering land use management and its effect on the LSP change to better understand the effect of afforestation on global climate and carbon cycles.

Climate change↗
DOE OSTI2023

Unsaturated alluvium disposal modelling with improved geological realism

This report describes specific activities in the Fiscal Year (FY) 2023 associated with the Geologic Disposal Safety Assessment (GDSA) Repository Systems Analysis (RSA) work package funded by the Spent Fuel and Waste Science and Technology (SFWST) Campaign of the U.S. Department of Energy Office of Nuclear Energy (DOE-NE), Office of Spent Fuel and Waste Disposition (SFWD). The overall objective of the GDSA RSA work package is to develop generic deep geologic repository concepts and system performance assessment (PA) models in several potential host-rock environments, and to simulate and analyze these generic repository concepts and models using the GDSA Framework toolkit and other tools as needed. The specific GDSA goal addressed in this report is reference case development, simulation, and analysis for the unsaturated alluvium (UZ), one of the four potential host-rocks considered by the GDSA. Further, we aim to exercise the simulation tools and methodologies under development by GDSA for PA modelling.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2022

Space and Time Dynamics of Transpiration in the East River Watershed: Biotic and Abiotic Controls

Alpine forests have important impacts on water resources by affecting how much of the precipitation that falls on a watershed is routed to streamflow or returned to the atmosphere as transpiration – so-called “green water”. In many of the high-altitude watersheds that supply critical water resources to the western US, it remains a challenge to predict how much water forests will utilize limiting the ability to predict both short and long term information on downstream water resources. The uncertainty emerges because transpiration varies between tree species, across landscapes and over time in ways that cannot readily be predicted from physically-based models for evaporation. In this project, we utilized a technique called sap flux to measure the rate of water use for three common species of fir, spruce and aspen distributed across a hillslope in the East River Watershed from 2019-2021. We also made ~weekly measurements of the concentration of stable isotopic tracers in transpiration to understand not only how much water was being used but whether it originated from summer rain or snowmelt. Firstly, our results show that the species tend to use similar amounts of total water but they achieve this similar cumulative water flux both by using water at different times of the year and by using different water sources. This shows the importance of including species-level information into models used to make seasonal streamflow predictions. Secondly, our results show that trees lower on the hillslope use older waters all through the summer whereas trees higher on the hillslope rely more heavily on recent summer rain. This means trees lower on the hillslope are less sensitive to year-to-year changes in summer climate. Lastly, we found that sites with dense forest stands displayed the largest amounts of year-to-year changes in transpiration and thus these high localized spots in the watershed are the primary drivers of changes in “green water” use. The results provide an unprecedented spatial and temporal picture of forest water use that confirms a number of fundamental hypotheses on hillslope ecohydrology that had not previously been tested against observations. Predictions of streamflow that will benefit from this work are highly valuable for agriculture management, flood control and ecological restoration efforts.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2021

Taxonomic and Functional Diversity of Aquatic Macroinvertebrate Assemblages and Water Quality in Rivers of the Dry Tropics of Costa Rica

One of Costa Rica’s driest areas is the province of Guanacaste, in the Pacific Northwest, with almost no rain during the dry season from November to April. Due to this marked seasonality, the area is covered by dry tropical deciduous forest, considered the most threatened and least known tropical ecosystem in this area. This study analyzes and characterizes the assemblages of aquatic macroinvertebrates in water bodies within the Tempisque basin. Biological water quality was measured using the BMWP′-CR index. Macroinvertebrate assemblages were analyzed using abundance, richness, and functional feeding group approaches (FFG). Partial least square (PLS) analyses were performed, and the relationships between environmental factors and macroinvertebrate assemblages are also discussed. Macroinvertebrate assemblages were dominated numerically by mayflies, caddisflies, flies, and beetles. The BWMP′-CR index showed varying biological water quality, ranging from “very bad” to “excellent,” depending on rainfall and site management. Results suggest that tropical Mesoamerican rivers contradict the “river continuum concept” because predators and scrapers displace shredders in numbers. On the other hand, the study area shows a notable high richness of the Coleoptera genera. The class Rhynchocoela (Nemertea) is reported for the first time in Central America. The results indicate that the dry forest river ecosystem shows staggering biodiversity despite the surrounding agricultural land use, probably because of their older origin concerning tropical rain forests in Central America.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2025

What regulates decomposition in agroecosystems? Insights from reading the tea leaves

Litter decomposition is a critical Earth process, recycling nutrients and setting a portion of plant tissue on a path toward soil organic matter. Despite this importance, we still lack a good understanding of local factors that regulate decomposition, especially in agroecosystems where management plays an outsized role. Using a narrow range of climate and soils, we buried 1,308 pre-manufactured “litter bags” of differing residue quality (i.e., green and rooibos tea leaves) in 109 plots across several management practices to (1) explore the local controls on decomposition in agroecosystems and (2) test the robustness of the Tea Bag Index (TBI). We found that management practices intended to increase soil ecosystem services, that is, soil health, altered the decomposition of both teas. For example, adding nitrogen fertilizer and implementing perennial cropping decreased the extent of green tea decomposition (carbon-to-nitrogen ratio, or C:N = 12.8). No-tillage increased, but perennial cropping decreased, the rate of rooibos tea decomposition (C:N = 50.1). Cropped prairie accelerated green tea decomposition and increased the extent of red tea decomposition. A random forest regression model showed that soil temperature was the strongest predictor of green tea decomposition, but a soil health score also played a significant role in predicting the mass remaining. Soil texture and nutrient availability best predicted rooibos tea decomposition. Finer textured soils seemed to decelerate rooibos decomposition but increased the extent of decomposition. Furthermore, we demonstrated that the TBI metrics correlated somewhat well with empirically derived decomposition constants and were similarly sensitive to the effects of management. Still, the green tea stabilization factor had a substantial prediction bias. Our study increased our basic understanding of what regulates decomposition in agroecosystems. It also showed that the TBI can be a scientifically rigorous citizen science approach to monitoring changes in soil health.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2026

A Centralized AI Lakehouse Framework for Brain Tumor MRI Classification and Segmentation, University KPI Forecasting, and Water Potability Prediction

In many university and healthcare projects, models are built for very different data types such as tables, institutional time series, and medical images, but they are deployed as separate applications. In this work, that separation made testing and maintenance difficult because each module had its own pipeline and runtime requirements. This paper presents an integrated AI lakehouse-style implementation that runs three model pipelines inside one containerized backend. For medical imaging, we used MRI datasets from IEEE DataPort: a four-class classification set with 7012 images (5708 train/1304 test) and a segmentation set with 3063 image–mask pairs. The classification model (ResNet50 transfer learning) is evaluated using a proper train–validation–test protocol across multiple splits (80/10/10, 70/10/20, 60/10/30, and 10/30/60), achieving a test accuracy of 99.00% under the standard 80/10/10 split. Additionally, a patient-level evaluation is conducted using an external glioma dataset to provide a more realistic assessment without data leakage. The segmentation model (DeepLabV3-ResNet50) achieved 83.09% validation mIoU and 88.79% Dice score. For university KPI forecasting, we used annual IPEDS and NSF HERD data from 2010 to 2023 for three universities (BSU, EOU, and UAB). To examine the effect of preprocessing on forecasting performance, two case studies are conducted. In the first case, linear interpolation is applied to generate semester-level data. In the second case, the original annual data is used directly without interpolation. Random Forest regression and ARIMA models are evaluated using MAE, RMSE, MAPE, and R 2 . The results showed that interpolation improved apparent forecasting performance due to smoothing, while evaluation on the original annual data provided a more realistic assessment of model behavior. To further validate the framework on a larger dataset, an additional case study is conducted using a student dropout dataset. For water potability, we trained and compared multiple tabular classifiers on a large dataset (1,048,575 samples). A Random Forest model (100 trees, max depth 10) achieved 85.86% test accuracy and high recall for unsafe samples (0.8447). All modules are served via FastAPI and deployed together using Docker, with workflow automation routing requests to the correct endpoint. System-level benchmarking indicates that the backend maintains stable throughput and latency under concurrent requests.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2020

The Role of Crop Production in the Forest Landscape Restoration Approach—Assessing the Potential Benefits of Meeting the Bonn Challenge

Existing approaches and methodologies that investigate effects of land degradation on food security vary greatly. Although a relatively rich body of literature that investigates localized experiences, geophysical, and socioeconomic drivers of land degradation, and the costs and benefits of avoiding land degradation already exists, less rigorously explored are the global effects of restoring degraded landscapes for the health of the land, the climate, and world food security. The current scale of land degradation is such that the problem can be meaningfully addressed only if local successes are upscaled and a large number of landowners and land managers implement restoration activities. Significant global efforts to address degradation exist, but studies that evaluate the global benefits of these efforts generally do not account for global market forces and the complex web of relationships that determine the effects of wide-scale restoration on production and food security. This paper provides important insights into how a full integration of crop production in restoration efforts could impact food production levels, food availability, forest carbon stocks, and Greenhouse gas emissions.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2026

Response of soil nutrient pools and microbiomes to recurrent wildfire disturbance and varying burn severities in a mixed conifer forest

Wildfire is a pervasive disturbance in mixed-conifer forests, yet the relative influence of fire recurrence versus burn severity on soil biogeochemistry and microbial communities remains poorly quantified. We examined a natural gradient of fire history (0–3 prior fires) and burn severity (low–high) spanning 50 yr in a mixed-conifer ecosystem to assess how repeated fire shapes soil carbon (C) and nitrogen (N) pools, their isotopic signatures, mineral and particulate fractions, microbial community composition, carbon-use, CO₂ fluxes, and vegetation cover. Successive fires produced progressively higher bare-ground percentages and lower tree cover, which were tightly linked to declines in microbial diversity and reductions bulk %C, and %N. δ 13 C increased with fire frequency, indicating preferential loss of labile C through combustion or enhanced microbial oxidation, thereby explaining the observed net soil-C decline. Conversely, δ 15 N decreased and pH increased as tree density declined, reflecting altered N cycling and reduced acidification in post-fire soils. Fire recurrence, more than severity, corresponded with a marked shift in the bacterial community: for example, Xanthobacteraceae—key N-fixers and C-cyclers—diminished, while N-fixing Bacillaceae increased, underscoring the tightly coupled nature of soil nutrient dynamics and microbiome composition after repeated burns. Our results demonstrate that fire recurrence appears to be a stronger driver of post-fire soil ecosystem responses in this mixed-conifer forest, influencing both abiotic nutrient pools and the functional potential of the soil microbiome. These findings provide a more enhanced assessment and understanding to date of the biogeochemical consequences of repeated wildfire disturbance that can be used to inform management strategies aimed at preserving soil health in fire-prone landscapes.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2024

Bayesian calibration of stochastic agent based model via random forest

SAND2024-11403O The Bayesian Calibration of Stochastic Agent-based Model via Random Forest is a code that was developed in concurrence with an article by the same name that was written for a journal. The code reproduces simulation results and plots from the article. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

SciDAC↗
DOE OSTI2026

Exploring Wildfire & Energy data toward State Prioritization Index (WESPI)

Energy infrastructure can both induce and suffer risks from wildfires ranging from direct damage to energy assets such as substations and power lines to Public Safety Power Shutoffs. Recent wildfire events underscore the need for data-driven approaches that help states and utilities proactively plan for wildfire risk. Existing national tools such as Federal Emergency Management Agency (FEMA)’s National Risk Index (NRI) are valuable for community hazard planning. However, they are less suited for energy infrastructure, as they emphasize population and building exposure rather than system vulnerabilities. In this paper, we explore relationships between energy and wildfire data and present a Wildfire-Energy State Prioritization Index (WESPI). Our methodology combines data from the US Forest Service’s Fire Simulation (FSIM) dataset with energy resilience metrics, historical fire incidents, and geospatial data on transmission lines and fire stations. Correlation analyses suggest that FSIM burn probability is more strongly associated with power outage metrics (ρ = 0.32) than NRI wildfire frequency, and counties with a greater density of fire stations experience more frequent, but less intense wildfires. We further leverage data for burn probability, transmission line density, and fire station density to develop a Wildfire-Energy State Prioritization Index (WESPI) to highlight counties where wildfire hazard, infrastructure exposure, and limited suppression capacity converge. The index provides a consistent, scalable framework for state energy offices and utilities to screen counties for vegetation management, optimization of outage management system deployment, and to inform wildfire mitigation plans.

Critical infrastructure↗
DOE OSTI2024

GDSA Repository Systems Analysis Investigations in FY 2024

The Disposal Research and Development (R&D) Program of the US Department of Energy (DOE) office of Nuclear Energy (NE-8) Spent Fuel and Waste Science and Technology (SFWST) Campaign is to conduct R&D on disposal of spent nuclear fuel (SNF) and high-level waste (HLW). The goal of the Geologic Disposal Safety Assessment (GDSA) within this project is to develop a disposal system modeling and analysis capability that supports the integrated modeling of coupled processes controlling disposal system performance of deep geologic repositories, including uncertainty. This report describes specific activities in the Fiscal Year (FY) 2024 associated with the GDSA Repository Systems Analysis (RSA) work package. The overall objective of the GDSA RSA work package is to develop generic deep geologic repository concepts and repository system performance models in crystalline, argillite, salt, and unsaturated alluvium potential host-rock environments, and to simulate and analyze these generic repository concepts and models using GDSA Framework toolkit, and other tools as needed.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2021

Soil carbon sequestration potential and the identification of hotspots in the Eastern Corn Belt of the United States

Soil carbon sequestration is a significant CO 2 mitigation strategy but precise assessments of sequestration require spatially explicit modeling of potential changes in soil organic carbon (SOC) in response to soil, climate, land condition and management interactions. We assessed the SOC sequestration potential of the Eastern Corn Belt (ECB) in the USA (Indiana, Kentucky, Maryland, Michigan, Ohio, Pennsylvania, and West Virginia) in response to the adoption of conservation farming practices and land use change (pasture and forestation) using the SOCRATES model. Input data was provided through an intersection of the STATSGO, National Land Cover Database and a PRISM climate surface. At the end of the 20th century, the 15.3 Mha of cropped soils in the ECB contained 632 Tg C, an estimated reduction of 52% since the introduction of agriculture in the mid-1800s. Complete adoption of no-tillage practices on Prime Cropland would potentially recover 147 Tg SOC over 20 years, whereas a continuation of conventional tillage would produce a loss of 35 Tg SOC. Sequestration hotspots (> 500 Gg increase in SOC) under no-tillage cover 2.3 Mha providing 28 Tg C over 20 years. The conversion of Marginal (non-Prime) agricultural lands to Forests would yield an additional 13 Tg C in SOC and 381 Tg C in aboveground biomass. The rehabilitation of Minelands to Forest would yield an additional 4 Tg C in SOC and 42 Tg C in biomass. Here, opportunities to sequester C in the ECB via tillage and reforestation are substantial and should be incorporated into regional and national climate change mitigation strategies.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2022

Classification of animal sounds in a hyperdiverse rainforest using convolutional neural networks with data augmentation

To protect tropical forest biodiversity, we need to be able to detect it reliably, cheaply, and at scale. Automated detection of sound producing animals from passively recorded soundscapes via machine-learning approaches is a promising technique towards this goal, but it is constrained by the necessity of large training data sets. Using soundscapes from a tropical forest in Borneo and a Convolutional Neural Network model (CNN), we investigate i) the minimum viable training data set size for accurate prediction of call types (‘sonotypes’), and ii) the extent to which data augmentation and transfer learning can overcome the issue of small and imbalanced training data sets. We found that even relatively high sample sizes (>80 per sonotype) lead to mediocre accuracy, which however improved significantly with data augmentation and transfer learning, including at extremely small sample sizes (3 per sonotype), regardless of taxonomic group or call characteristics. Neither transfer learning nor data augmentation alone achieved high accuracy. Our results suggest that transfer learning and data augmentation could make the use of CNNs to classify species’ vocalizations feasible even for small soundscape-based projects with many rare species. Retraining our open-source model requires only basic programming skills which makes it possible for individual conservation initiatives to match their local context, in order to enable more evidence-informed management of biodiversity.

54 ENVIRONMENTAL SCIENCES↗