Beyond MRV: high-resolution forest carbon modeling for climate mitigation planning over Maryland, USA
Explore the source record for details and available documents.
G Hurtt↗
SEARCH · Search NASA
Results for “Forest Modeling”
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.
Geometric-optical bidirectional reflectance modeling of a conifer forest canopy
A geometric-optical forest canopy model that treats conifers as cones casting shadows on a contrasting background explains the major anisotropies in bidirectional reflectance measurements of a conifer forest canopy. The model uses parallel-ray geometry to describe the illumination and viewing of conifers as three-dimensional cones. Both computer simulation and analytical closed-form expressions are implemented. The results show a good qualitative agreement with the directional reflectance measurements of the conifer stand, indicating that the three-dimensional nature of the canopy is a key factor in determining its directional reflectance.
Li, X.↗
Intermountain West Wildland Fires: Mapping Tree Mortality and Burn Patches using NASA Earth Observations to Determine Fire Risk and Inform Fire Management Practices
Within the intermountain west, monitoring fuel loads is a major concern for wildland fire management efforts. To address this concern, we partnered with the U.S. Forest Service to inform the agency which forested areas should be prioritized for prescribed burning and fuel reduction near human communities in the Bridger-Teton National Forest, Wyoming. We created burn maps, fuel load maps, and a tutorial document to identify forest impact trends and provide the partner with the tools to replicate project methods for expansion to other wildfire crisis strategy sites. These end products were made using two NASA Earth observations: Landsat 8 Operational Land Imager and Shuttle Radar Topography Mission. Based on our random forest analysis, our maps identified 998 acres within the Wildland Urban Interface that are predicted to have high fuel loading and high burn severity within the Bridger-Teton National Forest. Forested areas closer to heavily populated areas such as Jackson, Kelly, Moran, New Forks Lake, and Star Valley Ranch should be prioritized for fuel reduction. However, our random forest model analysis was limited to using vegetation and topographical indices with no field data for model validation. Therefore, future studies should use field data for model validation to improve model accuracy and additionally incorporate Global Ecosystem Dynamics Investigation data into models to create better predictions of forested areas with high fuel load and high burn severity.
Remote Sensing↗
A theory of forest dynamics: Spatially explicit models and issues of scale
Good progress has been made in the first year of DOE grant (number sign) FG02-90ER60933. The purpose of the project is to develop and investigate models of forest dynamics that apply across a range of spatial scales. The grant is one third of a three-part project. The second third was funded by the NSF this year and is intended to provide the empirical data necessary to calibrate and test small-scale (less than or equal to 1000 ha) models. The final third was also funded this year (NASA), and will provide data to calibrate and test the large-scale features of the models.
Pacala, S.↗
A composite L-band HH radar backscattering model for coniferous forest stands
The radar backscattering model developed by Richards et al. (1987), has been improved and further tested in this research. The trunk term may now be calculated from the exact solution to the electromagnetic wave equations instead of the corner reflector equation. Rough surface models have been introduced into the radar model, so that the forward reflectance and the backscattering from the ground surface are now calculated from the same model and, thus, are consistent. The number of trees in an individual pixel is assumed to be Poisson distributed, with tree height in a stand log-normally distributed. The simulated results show that the match of backscattering coefficients for eight forest stands between SIR-B image data and the simulated results are satisfying, and that the trunk term now seems to be convincingly established as the dominant term in the L-band HH radar return from coniferous forest stands.
Sun, Guoquing↗
Impacts of Disturbance History on Forest Carbon Stocks and Fluxes: Merging Satellite Disturbance Mapping with Forest Inventory Data in a Carbon Cycle Model Framework
Forest carbon stocks and fluxes are highly dynamic following stand-clearing disturbances from severe fire and harvest and this presents a significant challenge for continental carbon budget assessments. In this work we use forest inventory data to parameterize a carbon cycle model to represent post-disturbance carbon trajectories of carbon pools and fluxes for specific forest types growing in high and low site productivity class settings. We then apply these trajectories to landscapes and regions based on forest age distributions derived from either the FIA data or from Landsat time series stacks (1985-2006) for 54 representative scenes throughout most of the conterminous United States. We estimate the net carbon uptake in forests caused by post-disturbance growth and decomposition ("regrowth sink") for forested regions across the country. At the landscape scale, the prevailing condition of positive net ecosystem productivity (NEP) is in stark contrast to local patches with large sources, particularly in the west where fires and clear cuts create contiguous disturbed patches. At the continental scale, regional differences in disturbance rates reflect management patterns of high disturbance rates in the Southeastern and South Central states, and lower disturbance rates in the Northeast and Northern Lakes States. Despite low contemporary disturbance rates in the Northeast and Northern Lakes States (0.61 and 0.74% y(exp −1)), the regrowth sink there remains of moderate to large strength (88 and 57 g C m(exp −2) y(exp −1)) owing to the continued legacy from historical clearing. Large regrowth sinks are also found in the Southeast, South Central, and Pacific Southwest regions (85, 86, and 95 g C m(exp −2) y(exp −1)) where disturbance rates also tend to be higher (1.59, 1.38, and 0.93% y(exp −1)). Overall, the Landsat-derived disturbance rates are elevated relative to FIA-derived rates (1.19 versus 0.93% y(exp −1)) particularly for western regions. The differences only modestly adjust regional- and continental-scale carbon budgets, reducing NEP from forest regrowth by about 8%.
carbon↗
Radar polarization signatures of vegetated areas
A simple model is presented for the prediction of the full polarization signature of vegetation resembling tall grass. This polarization signature can be used to detect the presence of vegetation even in those cases in which the vegetation layers are comparatively thin. Also presented is a model which predicts the polarization dependence of different tree types. Attention is given to the cases of pine and deciduous forest model predictions; both types of forest can be expected to contain terms representing the scatter from the ground, as well as forward, double reflections from the ground and limbs/trunk.
Van Zyl, Jakob J.↗
Coherent Effects in Microwave Backscattering Models for Forrest Canopies
In modeling forest canopies, several scattering mechanisms are taken into account: (1) volume scattering; (2) surface-volume interaction; and (3) surface scattering from forest floor. Depending on the structural and dielectric characteristics of forest canopies, the relative contribution of each mechanism in the total backscatter signal of an imaging radar can vary. In this paper, two commonly used first-order discrete scattering models, distorted born approximation (DBA) and radiative transfer (RT) are used to simulate the backscattered power received by polarimetric radars at P-, L-, and C-bands over coniferous and deciduous forests. The difference between the two models resides on the coherent effect in the surface-volume interaction terms. To demonstrate this point, the models are first compared based on their underlying theoretical assumptions and then according to simulation results over coniferous and deciduous forests. It is shown that by using the same scattering functions for various components of trees (i.e., leaf, branch, stem), the radiative transfer and distorted Born models are equivalent, except in low frequencies, where surface-volume interaction terms may become important, and the coherent contribution may be significant. In this case, the difference between the two models can reach up to 3 dB in both co-polarized and cross-polarized channels, which can influence the performance of retrieval algorithms.
Saatchi, Sasan S.↗
The derivation of sub-canopy surface terrain models of coastal forests using synthetic aperture radar
Radar data acquired by the Shuttle Imaging Radar-B mission covering a portion of the Mouths of the Ganges forests were used to create a terrain model for use in determining tidal flow and eventual nutrient transport from the forest to the marine habitat. Results show that good digital topographic terrain models of wet coastal forests can be generated using multiple sets of L-band SAR and ancillary tide elevation data. The dominance of the interaction phenomenon in the radar backscatter of flooded forests can be used to create sub-canopy inundation maps which when merged with tide surface data can be used to generate reasonable topographic models. Ideally models could be improved by using multiple sets of data at a constant incidence angle over the total tide range. The optimal angle for the SAR depends upon the characteristics of the forest. The range of 46 to 57 deg seems applicable to the 12.5 m tall closed canopy in this example. Such models can be an extremely valuable tool for studying and mapping the mangal ecosystem.
Imhoff, M. L.↗
Remote sensing of forest ecosystem dynamics: Measurements and modeling
The development of an integrated approach to the modeling of forest dynamics encompassing submodels of forest growth and succession, soil processes and radiation interactions, is reported. Remote sensing technology is a key element of this study in that it provides data for developing, initializing, updating, and validating the models. The objectives are reviewed, the data collected and models in use are discussed, and a framework for studying interactions between the forest growth, soil process and energy interaction components, is described. Remote sensing technology used in the study includes optical and microwave field, aircraft and satellite borne instruments. The types of data collected during intensive field and aircraft campaigns included bidirectional reflectance, thermal emittance and multifrequency, multipolarization synthetic aperture radar backscatter. Synthetic imagery of derived products such as forest biomass and NDVI (Normalized Difference Vegetative Index), and collections of ground data are being assembled in a georeferenced data base. These data are used to drive or test multidiscipline simulations of forested ecosystems. Enhancements to the modeling environment permit considerable flexibility in configuring simulations and selecting results for reporting and graphical display.
Williams, Darrel L.↗
Quantifying Boreal Forest Structure and Composition Using UAV Structure from Motion
The vast extent and inaccessibility of boreal forest ecosystems are barriers to routine monitoring of forest structure and composition. In this research, we bridge the scale gap between intensive but sparse plot measurements and extensive remote sensing studies by collecting forest inventory variables at the plot scale using an unmanned aerial vehicle (UAV) and a structure from motion (SfM) approach. At 20 Forest Inventory and Analysis (FIA) subplots in interior Alaska, we acquired overlapping imagery and generated dense, 3D, RGB (red, green, blue) point clouds. We used these data to model forest type at the individual crown scale as well as subplot-scale tree density (TD), basal area (BA), and aboveground biomass (AGB). We achieved 85% cross-validation accuracy for five species at the crown level. Classification accuracy was maximized using three variables representing crown height, form, and color. Consistent with previous UAV-based studies, SfM point cloud data generated robust models of TD (r(sup 2) = 0.91), BA (r(sup 2) = 0.79), and AGB (r(sup 2) = 0.92), using a mix of plot- and crown-scale information. Precise estimation of TD required either segment counts or species information to differentiate black spruce from mixed white spruce plots. The accuracy of species-specific estimates of TD, BA, and AGB at the plot scale was somewhat variable, ranging from accurate estimates of black spruce TD (+/−1%) and aspen BA (−2%) to misallocation of aspen AGB (+118%) and white spruce AGB (−50%). These results convey the potential utility of SfM data for forest type discrimination in FIA plots and the remaining challenges to develop classification approaches for species-specific estimates at the plot scale that are more robust to segmentation error.
aboveground biomass↗
Navigating Team Dynamics: Automated Detection of Micro-Behaviors Between Team Members Through Longitudinal Interaction Data
The success in future long term space exploration missions will depend on the cooperation, coordination, and mutual understanding among the crew members. Micro-behaviors are momentary, subtle linguistic and paralinguistic indicators of thinking and feeling toward another member of the team (Cortina et al., 2001; Smith & Griffiths, 2022) that can significantly impact team dynamics and influence the overall team performance (Paromita & Chaspari, 2024). Due to their interactive nature, micro-behaviors have a sender (i.e., the team member expressing the micro-behavior) and a target (the team member impacted by the micro-behavior). Detection of these behaviors can assist in avoiding possible conflict among crew members and promoting the overall team success. Our prior research focused on an initial proof of concept of machine learning (ML) models and natural language processing (NLP) techniques that were used for automatically detect micro-behaviors among crew members of the US National Aeronautics and Space Administration’s (NASA) Human Exploration Research Analog (HERA) Campaigns 4 and 5 missions (Paromita et al., 2023). Results underscored the importance of incorporating contextual information in the ML models in the form of sentiment analysis, type of task, and dyadic interaction among team members. Here, we expand the scope of our prior work in two ways. First, we assess ML/NLP methods on new behavioral annotations coded using an adapted version of Smith & Griffins (2022) theoretical framework in terms of Violation (i.e., presence of valenced behavior, uplifting/positive or discouraging/negative), Intensity (i.e., force of behavior in terms of how uplifting or discouraging is the behavior), and Intent (i.e., motive of the behavior in terms of whether it was deliberate or unintentional). Second, we expand the design of the ML model to preserve information about the role of each team member within the occurrence of the micro-behavior (in contrast to the previous model that only considered the sender and the target without determining the team member role). This allows to consider all team members' contributions in the conversation and model long-term dependencies in the dialogue. Our experiments for this study are conducted on data from 5 teams of the NASA HERA C4 (NASA grant NNX16AQ48G (PI: Bell)). Conversations were extracted from the 1.5 hour Team Interaction Battery (TIB) task that occurred 5 times in-mission per crew. This resulted in a total of 13,058 conversational turns (i.e., 17.8% uplifting, 3.3% discouraging, 75.76% neutral, 3.14% nulls). Our findings with the revised behavioral coding and ML/NLP models indicate a 43.66% macro F1-score (i.e., 38.29% precision (P), 50.8% recall (R)) for a dialog state-tracking model that includes information from the sender only, and a 40.9% F1-score (i.e., 38.7% P, 43.36% R) for the same model that includes information from both the sender and the target of the micro-behavior. These are significantly higher compared to simple random forest models that classify behaviors strictly based on speech content and do not consider iterative team dynamics, achieving a 36.07% F1-score (i.e., 39.04% R, 33.53% P). Our findings demonstrate potential ways to leverage large conversational datasets to better capture complex team dynamics. We will discuss future directions including proposed models that can incorporate additional mission days and tasks beyond the TIB for objectively quantifying team behavior at high temporal resolution in space exploration missions.
Projna Paromita↗
Northern Colorado Water Resources: Using Earth Observations to Assess Relationships between Snowpack and Wildfires for Water Utility Management
In recent years, wildfires in the western U.S. have increased in size, frequency, and severity. Colorado is no exception to this trend, with many recent wildfires raising concerns. The Cameron Peak and East Troublesome Fires, the two largest recorded wildfires in Colorado history, burned over 400,000 acres in 2020. Wildfires pose a significant threat to high-elevation ecosystems, especially in terms of disturbances to watershed resources. Colorado’s annual water supply is dependent upon the melt and runoff of high-elevation snowpack, thus understanding wildfire effects is crucial. This project partnered with Northern Water, a water management company in Colorado, to explore fire impacts on Colorado Front Range watersheds and water supply forecasts. We investigated the use of remote sensing to study the influences of wildfire on snow depth, as Northern Water had previously relied on field-based observations of snowpack for their water supply forecasts. This project utilized airborne LiDAR data from Airborne Snow Observatories, Inc., Landsat 8 OLI derived products, as well as in situ snow survey data from Colorado State University to assess and model changes in snowpack over time. We uncovered key drivers of change in snowpack such as burn severity and snow zone and created a random forest model to determine and investigate their relationships. Despite temporal data availability limitations, analysis also showed that remotely-sensed snow depth measurements were strongly correlated with in situ snow depth measurements, revealing the accuracy and feasibility of using remote sensing technologies to study landscape-scale snowpack characteristics for water utility management purposes.
airborne LiDAR↗
Spectral response curve models applied to forest cover-type discrimination
The potential of remote sensing systems to provide a cost-effective inventory tool in the case of forest resources is currently of interest to a variety of natural resources management agencies. A number of studies have been performed regarding the use of Landsat data for mapping forest resources in Michigan. The present paper is concerned with current research, which has been directed toward the development and evaluation of computer-implemented classifications for the identification and characterization of coniferous forest types in Michigan's northern Lower Peninsula. Attention is given to the characteristic response curves from Landsat MSS data, spectral response curve models, and forest cover-type discrimination. It is found that spectral response curve models can be used to evaluate and explain the characteristic spectral responses of coniferous forest types on a snow-covered, winter Landsat scene.
Hudson, W. D.↗
Design of Materials with Alchemite
Machine learning models that establish the relationships between materials processing and properties can enable inverse design of materials through active learning. Alchemite is a commercial software that can perform inverse materials design on sparse data. Here we evaluate Alchemite’s performance on a dataset of shape memory alloys and a dataset of heat exchangers compared to baseline random forest models. Alchemite had higher accuracy when making predictions on sparse data and was more accurate or nearly as accurate as random forests on complete datasets while also quantifying uncertainty. The software was also used to suggest processing steps and design parameters to optimize properties and performance; however, physical validation of the suggested design parameters was beyond the scope of this work. Several useful design insights were gained about the impact of the design parameters on properties and performance including the importance of dopant choice and amount for shape memory alloys and the importance of height and weight on the thermal resistance of heat exchangers.
Machine learning↗
Evaluating Meteorological Dust Events and Machine-Learning Based Dust Identification in Geostationary Satellite Imagery
NASA scientists in the Short-term Prediction Research and Transition Center (SPoRT) developed a physically-based machine learning approach to identify dust in satellite imagery with a focus on night-time dust detection (Berndt et al. 201; DustTracker-AI). NASA/NOAA Geostationary Environmental Operational Satellite-16 (GOES-16) imagery was used for training and model inputs. The training, testing and validation data set consists of 28 events in the Southwest United States, capturing dust and null events in the region from 2018-2020.With 83 distinct images and millions of pixels a random forest model was trained and validated, correctly labeling 85% of dust pixels.For the first time, the model was run in near-real time production during the spring of 2022 and dust probability visualizations were made available to NOAA National Weather Service (NWS) forecasters to assess its utility for dust forecasting. Results indicated the model helped increase the confidence in the presence of dust and enabled dust tracking for a longer period of time into the night-time hours. Forecaster assessment and running the model in near real-time allowed for the team to determine the types of events missed, captured, and false alarms. To gain additional context on model performance,the SPoRT team sought to gather more detailed information on the training database(e.g., meteorological characteristics and drivers). The goal of this project was to identify the meteorological drivers for the dust events and create a database which synthesized information from observations, forecaster discussions, and analyses pertaining to the dust events to understand the types of events currently used to train the model. A more detailed meteorological synopsis was created for each dust event in the training, testing, and validation datasets. Following the completion of the database and documentation, the classification details revealed that 88% of the dust events were synoptically driven while mesoscale events were less prevalent in model datasets. Meteorological conditions found such as mixing layer depth and wind velocity had mean values of 645mb and 21kt respectively.With conditions of deep mixed layers and moderate to strong surface winds a mesoscale thunderstorm outflow event was considered and subsequently added to the model training data set to test the impact of additional mesoscale training data. The model was retrained and then qualitatively tested on a sample thunderstorm outflow case that the original model was unable to identify. Preliminary results showed potential that the addition of more mesoscale events included in the training data could help to better identify indistinct and localized dust events.
Connor Welch↗
Utilizing Earth Observations to Model Probable Coastal Wetland Extent, Sea-Level Rise Inundation Risk, and Assess Impacts on Historic Hawaiian Lands
Climate induced sea-level rise poses a risk to coastal areas on the Island of Hawai’i, and many of the island’s historic cultural lands are in danger of becoming overtaken by wetlands or inundation. In partnership with the County of Hawai’i, State of Hawai’i Department of Land and Natural Resources, and Arizona State University, NASA DEVELOP mapped wetland extent and short-term sea-level rise inundation risk. We utilized Earth observations over a 10-year span (2013 – 2022) that included the NASA MEaSUREs Gridded Sea Surface Height Anomalies and MEaSUREs Group for High Resolution Sea Surface Temperature datasets, United States Geological Survey (USGS) Hawaii Digital Elevation Models (DEM), and in situ tidal gauge data. Flood risk index values were acquired for 5 known Hawai’i flood events between 2019 – 2021 from the Global Flood Mapper tool on Google Earth Engine. We used a random forest model to predict short-term sea-level rise inundation risk along the entire coast of Hawai’i. Current wetland extents and probabilistic locations of new wetlands were modeled with the most recently available data from PlanetScope Surface Reflectance optical imagery (2022), USGS 3D Elevation Program (3DEP) 10m DEM (2020), temperature and precipitation data from the Hawai’i Climate Atlas, and soils data from the Hawai’i Soil Atlas (2014) using the Wetland Intrinsic Potential tool. Results indicated locations that had the highest probability of wetland creation. The end products aimed to help the partners prioritize efforts to meeting regulation requirements for wetlands protection, evaluate the inundation risk to historical features, and support decision-making for their Shoreline Setback and Climate Adaption plans.
Lisa Tanh↗
Northern Forest Ecosystem Dynamics Using Coupled Models and Remote Sensing
Forest ecosystem dynamics modeling, remote sensing data analysis, and a geographical information system (GIS) were used together to determine the possible growth and development of a northern forest in Maine, USA. Field measurements and airborne synthetic aperture radar (SAR) data were used to produce maps of forest cover type and above ground biomass. These forest attribute maps, along with a conventional soils map, were used to identify the initial conditions for forest ecosystem model simulations. Using this information along with ecosystem model results enabled the development of predictive maps of forest development. The results obtained were consistent with observed forest conditions and expected successional trajectories. The study demonstrated that ecosystem models might be used in a spatial context when parameterized and used with georeferenced data sets.
Ranson, K. J.↗