Project Selections for DSI-24

15 Projects Awarded Under the Decadal Survey Incubation (DSI) Program

08/19/2025 – The National Aeronautics and Space Administration (NASA) solicited Decadal Survey Incubation proposals (through the NASA Science Mission Directorate Research Opportunities in Space and Earth Sciences – 2024 NNH24ZDA001N-DSI A.55 Decadal Survey Incubation Program: Science and Technology) to accelerate readiness of high priority observables in the Planetary Boundary Layer (PBL) and Surface Topography and Vegetation (STV) Targeted Observables (TO) as outlined in the 2017 Decadal Survey, which are not yet feasible for cost-effective space flight implementation. PBL and STV science goals call for exploring next generation measurement approaches that could be ready for spaceborne implementation in the next decade. This program element supported the development of Earth observing instrument sensor systems, advanced information systems, and enabling science studies to further advance PBL and STV. The ultimate intent of DSI is to enable the next generation of possible measurement approaches, system architectures, and mission concepts to address PBL and STV needs.

NASA received a total of 97 proposals in response to this NRA and selected 15 for funding. The total funding to be provided for these investigations is approximately $14 million over three years. The awards are as follows:

 —-

Back to Top

—

PBL: Analyzing HAMMR-HD and TEMPEST-D (along-track) Data to Quantify the Capability of Future Broad-Spectrum Microwave Radiometers for PBL Thermodynamic Characterization
Shannon Brown, Jet Propulsion Laboratory

The Planetary Boundary Layer (PBL) incubation study team report has identified hyperspectral microwave sounders as a key component of a future observation system focused on characterizing the thermodynamic properties of this atmospheric layer in which we all live. The thermodynamic structure of the PBL is encoded in the shape of the microwave spectrum around the 22 GHz water vapor line and in the wings/windows of the 50-60, 118 and 183 GHz absorption lines. Existing microwave sounder systems sub-optimally sample the PBL since they are configured to uniformly sample the full tropospheric/stratospheric temperature/water vapor profile with limited (10s) channels. Current systems in development at NASA (under ESTO) and within other organizations use new broadband spectrometer technology that allow – for the first-time – complete sampling of the spectrum below 200 GHz returning all available PBL information in the microwave spectrum. Our algorithms must now keep up with the technology. To prepare for these future systems, we must develop and test advanced algorithms that efficiently use spectral information in new ways.

Our study focuses on two unique data sets — one spaceborne, one airborne – for demonstrating the capability (in terms of recovering the PBL thermodynamic structure) of the new hyperspectral sounders that will be flying later this decade.  The spaceborne dataset was acquired with the TEMPEST-D CubeSat water vapor sounder flying in a unique along-track scanning configuration where each spot on the ground is sampled over a +/- 70 degree range of emission angles. The multi-angle, multi-frequency observations give ‘hyperspectral-like’ dense vertical sampling of the water vapor in the PBL. The airborne dataset is from HAMMR-HD, which is anticipated to complete a 30-day stratospheric balloon flight by Fall 2025. HAMMR-HD has been recently upgraded to sample the PBL relevant portions of the microwave spectrum with channels at 18.7, 23.8, 34.0 GHz and broadband spectrometers sampling from 48-72 GHz and 113-183 GHz at 6MHz resolution. It can be considered an airborne prototype for the NASA AURORA pathfinder project, sharing the same 113-183 GHz spectral coverage planned for that sensor. It is the only microwave sensor currently configured to operate in these PBL relevant portions of the microwave spectrum and a unique resource for quantifying capability and informing future PBL instrument/observatory design.

We will evaluate the PBL information content in the non-linear shape of the multi-angle, broad-spectrum spaceborne and airborne observations.  We will first focus on quantifying the PBL thermodynamic information content in the microwave spectrum in these newly available spectral regions and multi-angle observation geometries, focusing on new ways to disentangle surface emissivity and temperature and atmospheric opacity and temperature, which is the foundation of any geophysical retrieval algorithm. From there, we will develop and test new PBL retrieval algorithm approaches that exploit this information. Our algorithms will advance beyond current optimal estimation approaches, which will struggle with the large amount of input data. We will evaluate machine learning (ML) algorithms that more efficiently deal with large datasets. We will be able to test and validate our algorithms using TEMPEST-D and HAMMR-HD. The study output is a quantification of the PBL retrieval capability from these new microwave hyperspectral sounders relative to the science goals outlined in the PBL study report. We will specifically focus on identifying additional information, from either other PBL remote sensing instruments or ancillary sources, that significantly improve the retrieval quality from the passive microwave sensors. This will help define sensor combinations for a future mission.

Back to Top

—

STV: 3D Forest Structure Retrievals using AI and Multi-modal Satellite Time-series Data
Antonio Ferraz, Jet Propulsion Laboratory

Satellite remote sensing offers a robust method for mapping 3D forest structure at scales that are critical for supporting forest management efforts and to monitoring progress toward international and US-promoted initiatives aiming at preserving carbon stocks, biodiversity, and habitat integrity.

Mapping forest structure from space relies on baseline 3D reference measurements, typically acquired by LiDAR sensors from airborne (Airborne Laser Scanning, ALS) or spaceborne platforms like GEDI. These LiDAR estimates are then spatially interpolated using satellite wall-to-wall imagery correlated with forest structure, such as reflectance data (e.g., Landsat and Sentinel-2) or radar backscatter (ALOS-PALSAR and Sentinel-1). Machine Learning (ML) has been widely used to correlate spatial patterns in the satellite imagery to the forest 3D structure.
However, these methods face challenges related to significant pixel-level uncertainties and biases. For instance, forest height in tall forests (greater than 40 meters) is often systematically underestimated, while short forests are overestimated. Uncertainties in retrieving forest 3D structure arise primarily from three factors:

1.       The limitations of baseline reference data. ALS data, while highly accurate, is often not representative across broader spatial domains. On the other hand, GEDI is available across large areas but may exhibit considerable uncertainties, particularly for forest understory metrics and in regions with complex topography.
2.         The limitations in the sensitivity of spaceborne spectral reflectance and radar backscatter measurements, which struggle to capture structural variation in taller and denser forests effectively.
3.         Imbalanced Learning problem due to the underrepresentation of less common forest in the data, leading ML models to overfit the “average” middle-sized forests.

To mitigate these uncertainties, a comprehensive approach is required, including:
1.         Developing next-generation observing systems and sensors with enhanced sensitivity to 3D forest structure, and
2.         Creating advanced, data-driven ML methodologies to better interpret complex spatiotemporal patterns present in satellite remote sensing data.

Our project focuses on studying the potential of ML models to explore multi-modal satellite time-series towards extracting spectral-spatial-temporal patterns across multiple scales in both the spatial (local to landscapes) and temporal (monthly, seasonal, and multi-annual) domains. In particular, we explore the added value of two Foundational Models, developed by NASA and  Arizona State University (ASU), that are pre-trained in massive satellite time-series and promise to improve classification and interpolation problems, in particular for tasks with scarse and noisy calibration data.

We hypothesize that our framework will significantly reduce uncertainties in retrieving 3D forest structure by exploring the variability in forest responses to both spatial and temporal environmental factors, such as topography, phenological cycles, post-wildfire recovery, pest infestations, drought conditions and other changes. In general, we expected to address the imbalanced learning problem by revealing forests that are unique in their spatiotemporal behavior, allowing the ML model to learn differences between “average” forests from rare forests.

A key component of our approach is ensuring the accuracy and representativeness of the calibration data used to train the ML models. We will evaluate the performance of calibrating the models using GEDI data (which offers widespread spatial coverage but lower accuracy and shorter temporal longevity) compared to ALS data (which provides higher accuracy and longer temporal longevity but has limited spatial and temporal coverage).

Back to Top

—

PBL: A Passive-Active, Multi-Sensor Approach to Earth’s Planetary Boundary Layer (PBL)  – A Demonstration from the Westcoast & Heartland Hyperspectral Microwave Sensor Intensive Experiment (WH2yMSIE)
Antonia Gambacorta, NASA Goddard Space Flight Center

The primary goal of this proposal is to mature and demonstrate a passive-active multi-sensor data fusion algorithm from novel and conventional technology, to improve the retrieval of thermodynamic profiles (temperature and water vapor) in the Planetary Boundary Layer (PBL). This will be accomplished using data from the Westcoast & Heartland Hyperspectral MW Sensor Intensive Experiment (WH2yMSIE).

Under the PBL DSI-2021 program, our team successfully developed a simulation-based AI/ML-based algorithm combining passive (e.g., hyperspectral microwave) and active (e.g., lidar) sensors. We conducted signal-to-noise trade studies, which quantified the enhanced information content for PBL retrievals. The data fusion technique we developed has guided future pathways, such as the design of the Advanced Ultra-high Resolution Optical Radiometer (AURORA) Pathfinder, addressing both technology and uncertainty gaps in current programs.

This proposal seeks to apply and validate our previous research using real-world data from the WH2yMSIE campaign. Specifically, we will develop a transfer learning method applicable to a combination of measurements from CoSMIR-H, AMPR, CPL, and eMAS, collected during WH2yMSIE. These datasets will help validate the algorithms through data fusion and address co-registration challenges that are critical for operational scalability.

This proposal research will meet the following objectives: 1) Pursue a pilot study to demonstrate the feasibility of passive and active data fusion techniques using real measurements (e.g., computational efficiency, co-registration challenges, scalability into future operational systems). 2) Demonstrate enhanced PBL profiling by assessing improvements in PBL temperature, water vapor, and height compared to existing operational products (e.g., JPSS); 3) Deliver a Science and Application Traceability Matrix (SATM) to support PBL sensor architecture decision-making.

This proposal aligns with the overarching goals of the Decadal Survey Incubation (DSI) program, particularly: “(1) innovation in the research, development, and demonstration of new measurement technologies in preparation for future integrated observing system architectures, and (2) science activities that support maturation of measurement concepts, retrieval algorithms, […] and/or integrated observing system approaches.”

The primary focus of this proposal is the “”PoR Focus Area,”” as it addresses “”specific topics listed in the announcement, such as: “The refinement of multi-instrument retrieval approaches” … “that builds upon previous efforts that have illustrated the capability to advance ability to perform science in the PBL from space. This includes, but is not limited to, work funded by NASA under ROSES-21 DSI.”

A secondary focus area is “Maturing Existing Spaceborne PBL Capabilities and Investments” as it addresses specific topics listed in the announcement, such as: “The refinement of multi-instrument retrieval approaches” … “that builds upon previous efforts that have illustrated the capability to advance ability to perform science in the PBL from space. This includes, but is not limited to, work funded by NASA under ROSES-21 DSI.”

Back to Top

—

PBL: Synergistic Retrievals as a Pathway to Improving Thermodynamic Observations of the Planetary Boundary Layer
Robert Knuteson, University Of Wisconsin, Madison

The focus of this proposal is on advancing the science of the planetary boundary layer through the exploitation of simultaneous measurements of space-borne, airborne, and ground-based sensors using both passive and active remote sensing techniques. The proposal is in response to NASA ROSES 2024 A.55 Decadal Survey Incubator subelement “”PBL Science””. The goal of this proposal is to advance the scientific basis of measurement requirements necessary for moving the DSI PBL Incubation technology towards Targeted Observable (TO) status in the next Decadal Survey. The ultimate intent of the NASA DSI is to enable the next generation of possible measurement approaches, system architectures, and mission concepts to address PBL community needs.

We anticipate that the NASA DSI PBL program will leverage past spaceborne measurements, current and future airborne measurements, and past, present, and future ground-based measurements to accomplish this goal. Our PI team has 30+ years of experience in the development and operational implementation of the hyperspectral infrared sensors that span the period of record from the launch of NASA AQUA AIRS in 2002 through the most recent launch of the NOAA-21 (JPSS-2) Cross-track Infrared Sounder currently in orbit. Prior to and during this satellite development the University of Wisconsin-Madison Space Science and Engineering Center has operated the High-resolution Interferometer Sounder (HIS) and the follow-on Scanning-HIS from NASA high altitude (ER-2 and WB-57) and middle altitude (DC-8) airborne platforms. Most recently, the S-HIS participated in the Westcoast & Heartland Hyperspectral Microwave Sensor Intensive Experiment (WHyMSIE) field campaign which was highlighted in the A.55 announcement of opportunity.

The PI group also has intimate knowledge of ground-based uplooking hyperspectral infrared instruments designed at SSEC called the Atmospheric Emitted Radiance Interferometer (AERI) and deployed by the Department of Energy at their Atmospheric Radiation Measurements (ARM) sites to provide high vertical resolution thermodynamic structure information. Within this proposal we bring together truth datasets from the DOE ARM sites with an Optimal Estimation (OE) methodology that allows us to synergistically combine measurements from different sensor types with known measurement uncertainties. This includes, but is not limited to, the combination of airborne hyperspectal infrared with hyperspectral microwave datasets, e.g. from WH2yMSIE, or airborne hyperspectral infrared with ground-based hyperspectral infrared, or hyperspectral infrared in synergistic combination with active sensors, e.g. water vapor differential absorption lidar (DIAL) or Radio Occultation (RO). The scientific emphasis of these studies is to more fully characterize the uncertainties of remote sensing measurements across a broad range of climatic atmospheric regimes both for individual sensors and in various combinations. The PI team anticipates contributing to the refinement of measurement requirements for vertical, horizontal, and temporal scales that are needed to formulate a comprehensive Science and Applications Traceability Matrix (SATM) for the DSI PBL TO as an essential component of any future integrated observing system architecture.

Back to Top

—

STV: Maturing STV Mission Concept towards Earthquake Science and Vertical Land Motion Measurement
Zhen Liu, Jet Propulsion Laboratory

Global surface topography and change directly relate to a broad range of solid Earth processes including earthquakes, fault movements and vertical land motion (VLM). The 2021 STV science team report has identified a number of knowledge and methodology gap areas for solid Earth science, calling for gap-filling investigations to inform the architecture and system design of a future STV mission.

We have made initial progress in quantifying the measurement needs for mapping earthquake displacement and fault slip by exploiting the existing optical, topography, and radar data through the analysis of selected earthquake events [Antoine and Liu, 2024]. Our results show <1m resolution and submeter accuracy is preferred for reliable measurement of earthquake displacement and fault zone geometry while 1-3m resolution is a potential trade-off zone subject to further refinement. Topography vertical accuracy has greater effects than ground resolution on the measurement’s quality. In this project, we propose to build on the results from previous analyses and carry out further gap-filling trade studies with an overarching goal of maturing the STV mission concept towards earthquake and fault displacement science and VLM measurement. We have three objectives:

1. Perform a global trade-study to assess ground resolution and vertical accuracy needs for earthquake topography change and fault displacement measurement. We propose to perform further analysis informed by our selected event studies to refine the trade-off range for resolution and accuracy requirements. We will leverage the compilation of fault zone displacement and slip of earthquakes on a global scale [e.g. Milliner et al., 2024], coupled with forward simulations based on a global earthquake rupture model database to identify the acceptable resolution/accuracy and desired instrument measurement needs for future STV mission on a global scale.

2. Develop and exploit the fusion of lidar, stereo optical, global/regional topography and InSAR for fault slip science and 3-D landscape change. Lidar and stereo optical imagery are complementary techniques to construct Digital Surface/Terrain Models (DSM/DTM) and map fault creep and landscape change. Our initial test of using optical stereo data to map fault creep shows its promise to detect fault creep but results have large uncertainties subject to multiple confounding factors, dominantly vegetation effect. We propose to use the central San Andreas fault (CSAF) as a target area to further examine the combined use of multiple sensors for uncertainty reduction and noise correction with the aim of identifying optimal fusion strategy for the study of fault displacement science and 3-D land surface change. The CSAF is also covered by STV precursory coincident datasets (PCD) and planned ASCENT campaign.  We will consider using PCD and ASCENT when available as additional training/validation datasets to develop a machine learning approach to correct vegetation effects on mapping fault creep and geomorphology processes.

3. Assess the resolution and accuracy needs for global baseline topography (GBT) and its application in mapping VLM. A future STV mission likely involves snapshots (less frequent) of GBT measurement coupled with more frequent repeating measurements at targeted scales with affordable cost. We aim to demonstrate the resolution and vertical accuracy needs for mapping VLM processes through exploiting existing global Digital Elevation Models (e.g., NASADEM, COP30, FABDEM, etc.) and regional DEM products (3DEP, EarthDEM, etc.) at selected target areas (Central Valley and the Pacific coast) in California. We will quantify the measurement accuracy of existing global and regional DEMs at different resolutions and investigate the capability of combined use of the global DEM with higher resolution regional DEM for measuring these slow VLM processes that do not require very frequent measurement, but vertical accuracy is a key factor.

Back to Top

—

PBL: Improving PBL Temperature and Water Vapor Information by Combining Hyperspectral MW and IR Sounders with Active Lidar Measurements
Xu Liu, NASA Langley Research Center

We propose to study the synergistic use of multiple measurements from passive hyperspectral IR, MW and active Lidar measurements to improve retrievals for PBL sounding.  Our objective is to establish the framework of a multi-sensor data fusion solution using different configurations of IR, MW, and lidar measurements, as well as demonstrate and validate the potential improvement of this approach for both airborne and space-borne PBL sounding.  Differential Absorption Lidar (DIAL)-enabled high vertical resolution water vapor profiles will complement sounder data. Emerging hyperspectral MW sounders will complement IR observations under cloudy conditions. The study will include advanced radiative transfer modeling, multi-sensor physical retrieval algorithm development, information content analysis, validation through simulation and WH2yMSIE POR data, and space-based PBL measurement trade studies.  Specifically, we propose to achieve the following five objectives:

1) We will complete the principal component based radiative transfer model (PCRTM) for hyperspectral MW remote sensors.  PCRTM, which is well established for IR and solar instruments,  provides effective and optimized means to use all spectral channels for the temperature and water vapor retrieval. The hyperspectral MW PCRTM for CoSMIR-H will be delivered as an open-source tool to support the WH2yMSIE campaign. A more general version of satellite-based MW PCRTM will also be delivered and can support the PBL as well as the general remote sensing community.

2) We will complete comprehensive trade studies to explore the information content of PBL sounding using different instrument configurations for a wide range of atmospheric and surface conditions. The studies include PCRTM-based high-fidelity simulations for hyperspectral IR and MW measurements, along with the estimation for averaging kernels and error covariance matrices based on realistic and varying assumptions for measurement uncertainties and a priori constraints (e.g. DIAL H2O profile).

3) We will extend the optimal estimation (OE) based Single Field-of-view (SFOV) Atmospheric Sounder Product (SiFSAP) retrieval algorithm to include hyperspectral MW retrieval applications. The SiFSAP algorithm has been used to process data from the airborne NAST-I and current space-based IR hyperspectral sensors.  The extended SiFSAP algorithm can be used to process CoSMIR-H data as well as carry out the joint NAST-I+CoSMIR-H retrievals. We will then apply the algorithm to the WH2yMSIE POR data to evaluate its performance.

4) We will study the use of high vertical resolution water vapor information from DIAL to enhance the SiFSAP retrieval by fusing multiple data sources in the OE framework.

5) Based on lessons learned using WH2yMSIE data, we will extend the information content analysis and simulation-retrieval study to future space-based instruments. The goal is to quantify the PBL sounding performance for different instrument configurations and therefore provide guideline information for instrument design and CONOPS of future space-borne PBL missions.

The proposed work will advance PBL goals and objectives defined by the DSI Program which aims to “build upon the current state of understanding in PBL and push advancements in future integrated observing system architectures that will enable science for the next decade.”  By analyzing POR data from multiple instruments used in the WH2yMSIE campaign, this work will provide a framework for extending the airborne measurements to satellite-based PBL sounding. Through improved temperature and water vapor sounding products, the study will enable more accurate PBL height determination, offering concrete solutions to enhance PBL research. Ultimately, the success of this effort will establish a robust framework for merging passive MW and IR sounder data with active lidar measurements towards better PBL observations.

Back to Top

—

STV: Volcano Topography Science and Applications Observation needs for STV
Paul Lundgren, Jet Propulsion Laboratory

Volcano topography change is among the largest on Earth, spanning a wide range of spatial and temporal scales. Volcanoes increase topography through extrusion of new lava flows or domes, and remove topography through explosive eruptions, caldera and sector collapse. Volcano topography also changes by deposition of ash or pyroclastic flows or by remobilization of recent deposits from lahars and landslides. The spatiotemporal variability of these processes suggests a need for high-resolution topography and dense temporal sampling during volcanic unrest.

We propose to use a variety of existing satellite and airborne topography datasets for select volcanoes in combination with numerical simulations to understand the impacts of data quality and spatiotemporal sampling on volcano science and applications. We focus on three key volcano science and application subtopics: 1) lava flow and lava dome flow volume and pathway forecasting; 2) pyroclastic flow and lahar deposition/erosion; 3) forecasting lava effusion rate trends. Several volcanoes will serve as test cases where high quality data products and recent activity allow us to assess data resolution and temporal sampling needs. Locations and events include: 1) the Kīlauea, Hawaii, 2018 eruption, and the on-going post-2021 sequence of eruptions in the Reykjanes Peninsula, Iceland; 2) silicic dome forming eruptions at Great Sitkin, Alaska; and Nevados de Chillán, Chile; 3) pyroclastic flows and lahars at Fuego volcano, Guatemala; and Tunugurahua, Ecuador; 4) time-varying effusion rates for eruptions at Bardarbunga, Iceland; Kīlauea; Nevados de Chillán; Cordon Caulle, Chile; Piton de la Fournaise, Reunion, France. Datasets include: TanDEM-X, NASA UAVSAR GLISTIN-A, EarthDEM, Maxar, Capella, Umbra, and locally acquired airborne Lidar and structure-from-motion (SfM) for Kīlauea, Great Sitkin, Fuego, Tungurahua, Nevados de Chillán, and Reykjanes.

We propose to build off our previous efforts. We developed a lava flow modeling code (flow DEM) that underpins both a theoretical basis for repeat topography resolution and noise effects on flow volume and effusion rate uncertainty (Roman and Lundgren, submitted). We are applying this to simulation analyses based on actual flows from the 2018 Kīlauea eruption. For pyroclastic flow (PF) analysis at Fuego volcano we use the VolcFlow software (Kelfoun, 2017) to simulate effects of pre-existing topography resolution based on varying resolution DEMs acquired both from satellite and SfM observations. We will extend the flow and hazard analysis to lahars (remobilization of volcanic deposits through high rainfall or glacier melting), a major volcanic hazard that preliminary analysis has shown to be highly sensitive to DEM resolution in both natural and urban terrains, using the Kestrel software (Vasconez et al., 2024). We will cross-validate flow models using the VICTOR framework (Co-I Lev), that will allow us to add greater realism to simulating multi-fluid flows like lava and pyroclastic flows. Using models of lava effusive rates based on previously studied eruptions (e.g. Coppola et al., 2017), we will explore forecasting strategies and impact of data latency on the evolution of eruptions and for estimating magma supply at subsurface storage zones.

Our proposal responds to A.55 2.2.1 STV Science component of the NRA. Through the combination of exemplary topography observations from recent radar, lidar, and photogrammetry observations combined with physical forecasting models we will quantify the product needs and develop the analysis tools during the current decade that will support the development of STV observing capabilities in the next decade.

Back to Top

—

STV: Advancing Surface Topography Science over Land Ice Masses with a Hybrid Data Fusion and Modeling Approach
Brooke Medley, NASA Goddard Space Flight Center

We propose work to address Cryospheric knowledge and methodology gaps to inform and advance next generation Surface Topography and Vegetation (STV) concepts through a science investigation that leverages Program of Record (POR) ice topography datasets, surface process modelling of ice sheets and glaciers, and additional Earth System Observatory observations to build a data fusion framework for generating enhanced spatiotemporal resolution ice elevation products.  Development of the data fusion framework will then allow us to investigate several different STV architectures to refine and relax the original science needs developed by the 2021 STV Incubation Team.  While an “everywhere all at once” mapping of global baseline topography is desired, it is unrealistic.  The work proposed here will maximize the scientific impact of sparsely sampled measurements of ice topography through time and space by benefitting from our understanding of ice surface behavior through state-of-the-art process modelling to intelligently derive spatiotemporally complete snapshots of ice-sheet-wide and glacier-wide surface topography.

The relevance of the proposed work to STV science and development is exhibited through several complementary and connected efforts including:

Ice surface height through time is driven by multiple processes that act across a variety of time and length scales.  Mesoscale snowfall events produce similar height changes in time across large areas of the ice sheet, whereas post-depositional wind redistribution can impart a local signal as snow if preferentially blown around subtle topographic features across the ice sheet.  At the same time, ice streams and glaciers deliver solid ice to the ocean through ice flow, which can drive additional dynamic height changes that are localized in space where warming ocean waters are delivered to the ice fronts.  Therefore, changes in ice elevation are correlated in space and time to different degrees depending on the driving force magnitude and location.  As a result, we can leverage incomplete space-time ice elevations to derive complete space-time gridded ice elevation change through fusion with process models and STV-adjacent observations.  Investigation and exploitation of the space-time length scales driving ice elevation change over the ice sheets can act to relax the scientific measurement needs for a future STV mission through reduction in time and/or space sampling.

This proposal aims to further refine ice elevation measurement needs through investigation of spatiotemporal length scales from the POR and simulated height changes, design and creation of a Machine Learning model to predict spatiotemporally complete ice elevations from sparse STV observations, and extended application of the aforementioned analyses to smaller scale ice caps and glaciers.  Such refinement of the ice elevation measurement needs for a future STV mission is critical as ice elevation is a primary driver at present to several needs in the Science and Applications Traceability Matrix.  Relaxation of these needs could help reduce cost, lengthen the lifespan, and/or extend the scientific applicability of an STV mission.

Back to Top

—

STV: Surface Topography and Biomass Resolution Needs for Enabling Wildfire and Vegetation-Atmosphere Modeling and Forecasts
Mostafa Momen, University of Houston

Wildfires are one of the most devastating natural disasters by causing more than $100 billion in costs to the US economy since 2010. The wildfire spread strongly depends on the topography and biomass canopy shape (fuel distribution). While wildfire models significantly rely on surface topography and vegetation (STV) data, they suffer from insufficient resolution of these data. The impacts of the STV input data on real wildfire forecasts and modeling are not comprehensively established using 3D high-resolution vegetation-resolving simulations. Furthermore, a systematic understanding of the impacts of spatial heterogeneity in fuel loading on wildfire behavior and spread remains a persisting challenge. The primary objective of this proposal is to address these knowledge gaps and demonstrate the STV resolution needs for a baseline topographic and biomass map to enable accurate wildfire, plume dispersion, and vegetation-atmospheric forecasts.

To address this objective, we will develop a new unique framework called INNOVATE (Intercomparison of Numerical Nodal-dependence of Vegetation-Atmosphere-Terrain Exchanges). This project will bring together an interdisciplinary research team with expertise in numerical modeling of environmental flows (PI) and remote sensing (Co-I). Our working hypothesis is that the coarse resolution of the existing input STV datasets of weather models causes large uncertainties in wildfire likelihood and plume dispersion forecasts. Furthermore, the complex interplay between the 3D vegetation structure and bare Earth topography together with atmospheric fluxes plays an essential role in wildfire spread. To test these hypotheses, we will use the Weather Research and Forecasting model (WRF) with a fire code (WRF-Fire) to simulate wildfire spotting likelihood and emission dispersions. Through the INNOVATE framework, we will conduct a downscaling technique to connect coarse-resolution weather models (~1 km) to very high-resolution (~1 m) 3D vegetation-resolving large-eddy simulations (LESs). We will systematically change the resolution of the bare surface topography and vegetation fuel data in WRF as well as model grid cells to characterize their resolution impacts on wildfire behavior and its generated plume concentrations for five past major wildfire events. Finally, we will ingest high-resolution 3D vegetation and bare topographic Light Detection and Ranging (LiDAR) data into our in-house LES code. Then, we will systematically vary the horizontal and vertical resolution of these input data such as biomass density, 3D vegetation structure, bare Earth topography, and atmospheric conditions to comprehensively characterize their resolution needs for accurate wildfire and vegetation-atmosphere modeling.

Significance and Relevance: At the completion of this task, we expect to have demonstrated the resolution needs for baseline topographic and vegetation structure maps to enable accurate wildfire forecasts. Furthermore, the project will answer some key open questions in the NASA STV incubation study regarding the relation between the vegetation structure and fire risk as well as the impacts of forest fuel loading on fire behavior. The proposal is specifically relevant to the following elements of this solicitation: Demonstrating resolution needs for a global baseline topographic map to enable various STV science and applications objectives, and identifying where this global baseline map or other STV products could provide a foundational enabling data product for other disciplines and observing systems that rely on topographic data. This proposal aims to help define what kind of vegetation-related products the STV mission should consider for future missions by collaborating with the NASA labs.

Back to Top

—

STV: Satellite (multi)stereo-lidar Fusion for High-resolution Bathymetry and Coastal Topography
Monica Palaseanu, USGS Reston

Nearshore continuous bathymetry is considered a “critical gap” in the STV Study Team Report, and no single technology identified can meet the current needs in terms of spatial resolution (1-3 m), vertical accuracy (decimeter-centimeter level), precision, and temporal repeat (monthly to weekly). Beyond STV, the current topo-bathymetry data scarcity has real impacts for millions. Island Nations in the Pacific, South-East Asia, and Caribbean, including US territories and the Compact of Agreement need accurate, high-resolution topo-bathymetric DEMs (TBDEMs) for adaptive planning and sustainable sharing of coastal resources, including fishing rights.

Our proposal answers topical science and applications questions and addresses all four major types of gaps identified as barriers to meeting the STV Incubation Study objectives for coastal areas, as well as specific gaps mentioned for cryosphere and hydrology. This project aims to deliver a robust satellite bathymetry processing module, which leverages intelligent fusion of high-resolution stereo optical imagery capabilities with available STV POR Datasets (especially altimeters) to meet the current coastal research community needs for resolution, vertical accuracy and even temporal repeat by leveraging the high-resolution and sub-weekly average revisit time potential of stereo imagery. No other satellite derived bathymetry method can deliver a seamless coastal topo-bathymetric DEM, a key minimum requirement for coastal science driven applications, from change and time-series analysis (geomorphic, vegetation), sedimentary budget transport (shoreline erosion, accretion, landslides), hazards (inundation, storm surge, infrastructure, safety), ecosystems (sandy beaches and coastal bluffs, estuaries, wetlands and marshes, coral reefs, kelp forests, tidepools, and barrier islands) to economic (fishery, tourism, housing, infrastructure).

To accomplish this, we will develop a novel stereo-lidar data fusion algorithms to produce seamless topo-bathymetric products with decimeter accuracy, by incorporating water surface elevation from in situ observations and global tidal models, intelligent 3D co-registration with available satellite altimetry, and jitter pointing correction. Key gap-filling activities included in this project are: 1. Single open-source methodology to derive seamless coastal topo-bathymetric DEM; 2. Answers for high-resolution (1-3 m), high vertical accuracy (at least decimeter level up to ~ 1 Secchi disk, and less than 0.5 m for very shallow waters less than 5 m depth), and potentially high temporal frequency (weekly revisits) needs for coastal topo-bathymetric DEM; 3. Collection, analysis and fusion of multi-sensor spaceborne, airborne, and in situ datasets, including but not limited to lidar and tide gauge data to increase vertical and geolocation accuracy, and eliminate systematic errors (jitter); 4. Quantification of bathymetric uncertainty, accuracy in derived geophysical information to understand their propagation in benthic habitat maps/products and time-series change estimates; and 5. A global index of optical satellite (including stereo) bathymetry retrievability split by seasonality (dry/wet).

This work will be the first complete framework for a space-borne STV system that can deliver a seamless high-resolution topo-bathymetric DEM for data-scarce areas such as Island Nations in the Pacific, South-East Asia, and Caribbean, including U.S territories and the Compact of Agreement.

Back to Top

—

PBL: Data System Strategy for a Multi-mission Integrated Observing System
Alexey Shiklomanov, NASA Goddard Space Flight Center

This proposal nominates Alexey Shiklomanov to the PBL Data System Survey and Assessment Tiger Team, with support from a GSFC-based team with a variety of expertise in PBL instrumentation, modeling, and informatics. Dr. Shiklomanov has a combination of (1) technical expertise in software architecture and Earth Science informatics; (2) scientific expertise in remote sensing, land surface modeling, and data science; (3) extensive professional network at GSFC, across NASA centers, and in the broader Earth Science informatics community; and (4) demonstrated ability to manage complex projects with large, diverse teams that makes him a strong candidate for the Tiger Team. Dr. Shiklomanov played a leadership role in the Earth Information System (EIS) project, which had a strong focus on using advanced computational infrastructure (including HPC and cloud) to produce new data products and deliver them to a variety of users. Dr. Shiklomanov also spent 2 years as Program Scientist for the NASA Earth Science Data Systems (ESDS) Program, where he established a deep understanding of ESDS technology and organization as well as the challenges of agency-scale data management and distribution. Dr. Shiklomanov currently serves in a similar capacity as Data Integration Strategist for the GSFC Earth Science Division front office. Finally, Dr. Shiklomanov leads efforts around data distribution and data service development at the Global Modeling and Assimilation Office (GMAO). Dr. Shiklomanov will be supported as Tiger Team PI by a strong interdisciplinary team, including disciplinary experts in land surface (Shawn Serbin, David Shean) and atmospheric processes (Antonia Gambacorta), mission data processing (JP Swinski), community-driven informatics (Tasha Snow), operational science data processing (David Giles), and project management (Jeff Piepmeier).

Our team will work with the PBL Tiger Team to help define a PBL mission data pipeline with system agility, efficiency, cloud-optimization, openness, and community co-production at its core. Combining this single data system and other mission and agency data systems, we additionally aim to design an intelligent, integrated, affordable multi-mission, multi-agency, multi-node (GEO-LEO-Suborbital-Surface) framework. This framework will ensure we bring together all temporally coincident variables required to characterize the PBL with interoperable data formats, projections, and grids for science discovery, satellite validation, data assimilation, and applications such as air quality. Building a Findable, Accessible, Interoperable and Reusable (FAIR) framework for production and data dissemination among agency-led missions, domestic and international partners, and agency teams to construct each PBL product will harness all available compute resources (including commercial cloud and on-premises systems), follow open source best practices for collaborative development and documentation, and use a community-driven communication platform to accelerate development and build community. Our work will build upon our team’s prior experience with science data processing and community development for ICESat-2 (e.g., SlideRule, CryoCloud) as well as ongoing formulation activities for the Atmosphere Observing System (AOS) mission science data processing system and an existing PBL micro data system framework developed for the WH2yMSIE campaign. Through the PI, the team will work together with other members of the PBL Tiger Team to leverage NASA and non-NASA expertise to ensure transferability and that our recommendations incorporate leading-edge and holistic data system practices.

Back to Top

—

PBL: Toward an Optimal Combination of Passive Orbital and Active Suborbital Measurements of the PBL Thermodynamic Vertical Structure
Joao Teixeira, Jet Propulsion Laboratory

Some of the critical Planetary Boundary Layer (PBL) science questions discussed in the NASA PBL Study Team report are related to important interactions between the PBL thermodynamic vertical structure and horizontal mesoscale variability. To address these questions, a new PBL satellite mission would require both high vertical and horizontal resolutions for temperature and water vapor profiles. Unfortunately, there is no single instrument that, from a space-based perspective, will be able to provide the required high vertical and horizontal resolutions.

A strategy to overcome this issue is to take advantage of the optimal combination of instruments with different characteristics. In this proposal, we address the broad issue of merging information provided by (a) instruments with high vertical resolution but lower horizontal resolution and horizontal sampling (e.g., differential absorption lidar [DIAL], differential absorption radar [DAR], radio occultation [RO]) in orbital and/or suborbital platforms, with (b) instruments with high horizontal resolution and sampling but lower vertical resolution such as hyperspectral infrared (IR) or microwave (MW) instruments in orbital platforms.

We will use (i) synthetic PBL temperature and water vapor profiles from large eddy simulation (LES) models for a variety of key PBL physical regimes; and (ii) observations from the current orbital and suborbital program of record (POR).

In particular, we will develop an approach to merge IR sounder measurements from space with suborbital DIAL observations. In order to achieve this objective, we will focus on two critical tasks: 1) the development of a machine learning (ML) approach using autoencoders and deep neural networks to estimate temperature profiles from suborbital DIAL measurements of water vapor profiles over the tropical and subtropical oceans, using our LES data as the training datasets for the ML algorithm; and 2) using the DIAL water vapor profile suborbital measurements (and corresponding ML temperature profiles) as prior information for an optimal estimation (OE) IR algorithm.

This approach will allow us to produce a full three-dimensional (3D) depiction of the PBL temperature and water vapor structure by combining these methodologies (ML and OE) and these PBL measurements (POR suborbital DIAL and POR orbital IR), ideally suited for studying PBL/mesoscale interactions. These results also inform the potential benefits of combined DIAL and IR instruments for future orbital missions.

Back to Top

—

PBL: Multi-Instrument High-Resolution Characterization of the Planetary Boundary Layer using Tomography Techniques
Kuo-Nung Wang, Jet Propulsion Laboratory

The three most important observables identified in the 2021 PBL study team report to characterize the planetary boundary layer (PBL) are: temperature, water vapor, and PBL height (PBLH). While the report clearly defines the required temporal and spatial resolution as well as accuracy, the extent to which the current POR can meet these requirements remains an open question. The lack of a generalized algorithm to combine multiple observations, the absence of rigorous uncertainty quantification (UQ) for the POR within the PBL, and the limited understanding of the information content when POR data are combined, represent the biggest challenges in determining the sufficiency of current and future PBL observation datasets.

We propose to develop a 3D tomography algorithm to better address these challenges. Tomography can combine radiance-based remote sensing such as (hyperspectral) microwave radiometer (MWR)/infrared radiometer (IR) with path-integrated observations such as GNSS-RO and ground GNSS line-of-sight measurements. Distinct from data assimilation, tomography provides observation-oriented 3D temperature and moisture solutions whose temporal/spatial resolutions are independent of a physical model–which can be highly uncertain for the PBL. A new PBLH detection algorithm based on multi-variate approach will be developed to provide accurate PBLH estimation. The results will be validated using the published WHyMSIE dataset and global radiosonde observations (e.g. IGRA) as well as with regional observation sites (e.g. ARM stations). Errors in temperature, water vapor, PBLH retrieval, and information content of each voxel will be quantified using POR data. The sufficiency and limitation of POR toward reaching the requirement stated in the PBL study team report will be evaluated throughout the study.

Our objectives are: (1) Design and implement the advanced tomographic algorithm to combine various remote sensing techniques such as MWR, IR, GNSS-RO, and data collected from ground GNSS stations. (2) Quantify the spatial/temporal resolution, solution error, and information content by use of tomography, single value decomposition, and sensitivity analysis. (3) Assess the feasibility and identify the limitations of using current POR for PBL characterization purposes, and search for potential scenarios to aid the development of future missions.

Back to Top

—

STV: An Agile Radar System for High-Resolution 3-D Surface Topography and Vegetation Structure Measurements from Stratospheric and Distributed Platforms
Lauren Wye, Aloft Sensing, Inc.

Interferometric synthetic aperture radar (InSAR) is a critical sensing modality for many Earth science investigations and has the capability of addressing key measurement gaps identified in NASA’s Surface Topography and Vegetation (STV) Study Team Report (STR). When paired with the emerging category of stratospheric platforms (high-altitude pseudo satellites, or HAPS), a properly designed InSAR system provides a new measurement capability able to address the most stringent horizontal, vertical, and revisit aspirations for bare surface topography, water surface topography, and vegetation structure.

For example, for bare surface topography, the critical values are 10 cm accuracy and 0.5 to 1.0 m resolution with repeat measurements as fast as one day. To achieve these aspirational requirements from space requires a large constellation of capable InSAR or LiDAR systems, an approach that is almost certainly too large in scope for a near-term STV mission. However, many of the most stringent requirements are driven by regional phenomena: volcanos, faults, and landslides among them. Suborbital HAPS can provide cost-effective broad regional coverage with tailored revisit times and reconfigurable formations, and when coupled with a compact X-band radar and innovative algorithms, can achieve these stringent measurement aspirations from a single platform or pair of platforms.

Aloft Sensing, Inc. (Aloft) has developed an innovative Synthetic Aperture Radar (SAR) sensor small enough to be accommodated on the lightest of stratospheric HAPS vehicles, yet capable enough to meet the most demanding STV requirements. Funded by prior NASA ESTO IIP and DSI awards, Aloft’s X-band Radar (AXR) system is enabled by Aloft’s patented, ultra-precise Position, Navigation, and Timing (AloftPNT) algorithms, allowing SAR coherency to be maintained over long temporal and spatial apertures, including across multiple distributed platforms. The combination of our unique sensor architecture, AloftPNT, and emerging HAPS capabilities enables Aloft to provide perhaps the only realizable, single measurement methodology to simultaneously satisfy the STV accuracy, resolution, and revisit requirements.

In this effort, Aloft proposes to extend our current stratospheric AXR SAR capabilities to include single-pass InSAR, enabling rapid repeat, aspirational-level accuracy that fills critical STV observation gaps. We start by establishing this capability from a single stratospheric vehicle, demonstrating a first-of-its-kind measurement system for bare earth topography measurements. We then propose to extend to multiple platforms, allowing longer baselines that can transform the measurement of surface water extent, and ultimately, through the use of three or more platforms, provide unique vegetation structure measurements currently unavailable from any existing measurement technique.

The outcome of this effort is a new and innovative suborbital measurement system that significantly advances the state of the art. The resulting single-pass InSAR capability has immediate applicability to the most challenging STV domains, with particular relevance to regional studies of volcanism, pre- and post-hazard monitoring and assessment, surface water studies, and advanced vegetation structure characterization. The demonstrated performance has the additional benefit of informing future STV system architecture designs and positioning STV observation for low risk and high readiness in the next Decadal Survey.

Entry TRL: 3, Exit TRL: 6. As a stand-alone SAR sensor, AXR currently exists at TRL-8; it is a fully functional SAR and repeat-pass InSAR sensor with laboratory, environmental, aircraft, and stratospheric flight testing. As a single-pass InSAR HAPS instrument, AXR is at TRL-3. We expect to reach system TRL-8 for the single-pass InSAR HAPS system and TRL-6 for the 3-D structure measurement system.

Back to Top

—

STV: The Concurrent Artificially-intelligent Spectrometry and Adaptive Lidar System (CASALS): A new STV sub-orbital capability
John Yorks, NASA Goddard Space Flight Center

The Concurrent Artificially-intelligent Spectrometry and Adaptive Lidar System (CASALS) is a swath-mapping altimetry lidar that employs a novel transmitter and grating method in combination with a state-of-the-art, high-speed, photon-sensitive detector array. Previous NASA altimetry lidars have acquired Earth’s surface height measurements in either 1D within single laser footprints or 2D by overlapping footprints along the ground track. Recent lidars have utilized a single wavelength, inhibiting simultaneous observations of bare surface land topography, ice topography, vegetation structure, and shallow water bathymetry. CASALS, enabled by orders of magnitude enhancement of overall photon efficiency compared to current spaceflight lidars, provides 3D swath mapping of Earth surface heights and vegetation waveforms with fine cross-track spatial resolution that are especially advantageous for characterizing the structure of forests. Given this, and the potential to add visible (~520 nm) capabilities to the already existing near infrared (1040 nm) instrument, CASALS is a strong candidate for the lidar measurements desired as part of a future Surface Topography and Vegetation (STV) space mission. The CASALS instrument concept was successfully demonstrated via rooftop testing in September 2024 and airborne flights in November 2024, but the cross-track swath was limited to a width of 50-100 m, the software and some hardware are considered lower fidelity, it only operates at a single wavelength, and no science data processing software exists to demonstrate the science performance.

The overall goals of this proposed work, submitted to the STV Technology sub-element, is to (1) increase the CASALS science utility for STV orbital and suborbital activities and (2) reduce the CASALS implementation risk for a future STV space mission through four objectives:

  1. Increase the swath width of airborne CASALS to 1 km (assuming an aircraft altitude of 10 km), providing the STV community’s desired spatial resolution and coverage.
  2. Ruggedize the instrument hardware and software for more efficient airborne operations, addressing the lidar-related technology gaps that can reduce SWaP and enable SmallSat solutions.
  3. Create CASALS Level 1 and 2 science data products, supporting the Airborne Surface, Cryosphere, Ecosystem, and Nearshore Topography (ASCENT) and STV goal of improving our knowledge and quantifying limitations of the various measurement approaches.
  4. Mature visible (i.e., ~520 nm) transmitter capabilities for bathymetry observations as a potential addition to future spaceborne CASALS concepts.

The planned milestones and schedule enable a wide-swath, more efficient airborne version of CASALS to fly in the ASCENT campaign in Hawaii during summer of 2027 and matures the CASALS system as a viable spaceborne lidar option for the STV mission. The wide-swath, high fidelity version of the CASALS instrument has an entry TRL of 4, with a planned exit TRL of 6.

Back to Top