Current Projects at CReSIS

Lead Institution: Pennsylvania State University

KU-PI: Leigh Stearns, Co-I: Carl Leuschen, Jilu Li

Sponsor: NSF

03/16-2018-02/28/2023

This project contributes to the joint initiative launched by the U.S. National Science Foundation (NSF) and the U.K. Natural Environment Research Council (NERC) to substantially improve decadal and longer-term projections of ice loss and sea-level rise originating from Thwaites Glacier in West Antarctica. Global costs of rapid sea-level rise to infrastructure (e.g., houses, roads, farms, ports) are likely to be large. A possible source of water for sea-level rise is the West Antarctic Ice Sheet, and Thwaites Glacier in particular. Ice sheets and glaciers contain vast quantities of water (in the form of ice) that is continually shed to the ocean, and continually replenished by snowfall (from water that evaporates from the oceans). If the amount of ice that Thwaites Glacier loses to the ocean over the next decades is much greater than the amount it receives as snowfall, then sea level in all the world's oceans would rise, possibly as much as a meter (approximately 3 feet). In order to estimate how likely such a catastrophic scenario would be, we need to better understand the surface over which Thwaites Glacier slides. If we can better characterize that layer ("is to smooth? Is it rough? Is it soft? Is it hard?"), then computer models of Thwaites would be much improved and we can make better projections of the amount of ice that Thwaites Glacier would shed to the ocean.

The objectives of the project are to learn whether basal conditions allow for rapid retreat of the Thwaites Glacier grounding line or whether retreat may re-stabilize near its current grounding line. These objectives will be achieved by using dedicated ice-flow modeling to guide targeted field surveys and experiments over two seasons, and to measure the most important unknown quantities and incorporate them into the models. Numerical models will be used to generate hypotheses for basal conditions that are testable through geophysical surveys and to project future behavior of Thwaites Glacier after assimilating the resulting data. The geophysical methods include seismic, radar, gravity, and electrical surveys that together will allow for a fuller characterization of the bed. We will conduct field surveys in areas representative of different parts of the glacier, including across the margins, near the grounding line, and along the central axis of the glacier into its catchment.

Lead Institution: New York University

KU-PI: John Paden

Sponsor: NSF

06/01/2018-05/31/2021

This project will observe, quantify and model the Thwaites ice-ocean system in the grounding zone, to firmly establish the physics linking ocean forcing and ice-sheet response. The time-dependent cavity will be thoroughly surveyed and instrumented with ocean monitoring devices. Melting will be observed by a network of autonomous sensors and from space over an extended period. The response of the glacier will also be observed. Our enhanced understanding of melting beneath TG’s ice shelf, its grounding zone and its connection with the glacier flow will be built into state-of-the-art coupled ice sheet and ocean models. These physics-rich, high-resolution models will allow the potential sea-level contribution of TG to be bounded with unprecedented fidelity.

We propose a suite of integrated activities: (1) multi-year oceanographic time series from beneath TG’s ice shelf to quantify melting processes that need inclusion in ocean models, (2) analogous measurements on the glacier to validate processes governing grounding-line retreat, (3) coupling of these in situ measurements with novel, high-resolution space-borne observations, (4) building this new understanding into state-of-the-art ocean and ice sheet models to correctly simulate the TG system, (5) coupling the models and running with realistic present-day ocean forcing to project the state of TG basin over the next hundred years. The international team will consist of experienced marine and glacier scientists using a range of techniques, from the well-established through to the cutting-edge. The outcome of the project will be a thorough understanding of the TG system in the critical zone extending from a few kilometers inland of the grounding line, through the grounding zone, and out under the ice shelf.

PI: John Paden

Sponsor: National Science Foundation

09/01/2018-08/31/2022

The objective of this research is to investigate artificial intelligence (AI) solutions for data collected by the Center for Remote Sensing of Ice Sheets (CReSIS) in order to provide an intelligent data understanding to automatically mine and analyze the heterogeneous dataset collected by CReSIS. Significant resources have been and will be spent in collecting and storing large and heterogeneous datasets from expensive Arctic and Antarctic fieldwork (e.g. through NSF Big Idea: Navigating the New Arctic). While traditional analyses provide some insight, the complexity, scale, and multidisciplinary nature of the data necessitate advanced intelligent solutions. This project will allow domain scientists to automatically answer questions about the properties of the data, including ice thickness, ice surface, ice bottom, internal layers, ice thickness prediction, and bedrock visualization. The planned approach will advance the broader big data research community by improving the efficiency of deep learning methods and in the investigation of methods to merge data-driven AI approaches with application-specific domain knowledge. Special attention will be given to women and minority involvement in the research and the project will develop new course materials for several classes in AI at a Hispanic and minority serving institute.

In polar radar sounder imagery, the delineation of the ice top and ice bottom and layering within the ice is essential for monitoring and modeling the growth of ice sheets and sea ice. The optimal approach to this problem should merge the radar sounder data with physical ice models and related datasets such as ice coverage and concentration maps, spatiotemporal meteorological maps, and ice velocity. Rather than directly engineering specific relations into the image analysis that require many parameters to be defined and tuned, data-dependent approaches let the machine learn these relationships. To devise intelligent solutions for navigating the big data from the Arctic and Antarctic and to scale up the current and traditional techniques to big data, this project plans several approaches for detecting ice surface, bottom, internal layers, 3D modeling of bedrock and spatial-temporal monitoring of the ice surface: 1) Devise new methodologies based on hybrid networks combining machine learning with traditional domain specific knowledge and transforming the entire deep learning network to the time-frequency domain. 2) Equip the machine with information that is not visible to the human eye or that is hard for a human operator to consider simultaneously, to be able to detect internal layers and 3D basal topography on a large scale. Using the results of the feature tracking of the ice surface in radar altimetry, the research effort will also develop new data-dependent techniques for predicting the ice thickness for following years based on deep recurrent neural networks.

Lead Institution: Georgia State

KU-PI: Haiyang Chao

Sponsor: NSF/USDA

09/02/2018-11/30/2022

Accurate predictions of wildfire spread are critical for effective wildfire management to support decision makings of fire managers and to ensure safety of firefighters. However, the lack of real time wildfire and wind data, both of which change in space and time, makes it difficult to achieve operational wildfire spread prediction. Unmanned Aircraft System (UAS) is emerging in many civilian applications and shows great potential in wildfire management. This project aims to develop and evaluate a collaborative human-UAS wildfire spread prediction and situational-awareness system for wildfire management. The UASs will work side-by-side with fire managers and ground firefighters to perform collaborative tasks. This new paradigm brings new research challenges from multiple aspects. First and foremost, the UASs must achieve sufficient autonomy in their mission so that they can autonomously collect the most useful information in dynamic wildfire environments. Besides wildfire sensing, the UASs also need to pay close attention to firefighters' safety by monitoring their vicinity. The second challenge is associated with effective teaming and collaboration between humans (fire managers and firefighters) and UASs. In particular, there is a need for humans to interact with and direct UASs' autonomy based on their domain knowledge and expert opinions for more effective wildfire management. To address these challenges, this project will includes four tasks: (1) fire sensing and wind estimation using a team of UASs to enable data-driven wildfire spread prediction, (2) UAS coordination and path planning algorithms governing UAS autonomy to sense dynamic wildfires while monitoring firefighters' safety risk, (3) teamed human-UAS collaboration, including human-directed autonomy and a human-UAS interaction interface to support human awareness of UAS operation, and (4) evaluation of the proposed research by flying a team of UASs over prescribed fires.

This project has the potential to transform wildfire management by enabling operational wildfire spread prediction and situation awareness for firefighters through teamed human-UASs collaboration. Using UASs to sense fire characteristics and wind parameters will fill the critical gap of real time data collection and data assimilation for operational wildfire spread prediction. The multi-UAS autonomy algorithms allow UASs to effectively collect the most useful information about dynamic wildfires and to monitor the safety of firefighters and other people on the ground. The approach of human-directed autonomy supports humans in-the-loop to optionally direct UAS teams to certain locations and tasks for effective human-UAS collaboration. Besides wildfire management, this research will also benefit other emergency response applications where humans and autonomous robots increasingly work together. The PIs will develop new and unique education and outreach programs, including a Wildfire-UAS Field Trip program and an annual outreach workshop series to provide interdisciplinary training to undergraduate/graduate students and to outreach to broader communities and the general public.

PI: Carl Leuschen

Sponsor: Heising Simons

04/01/2019-03/31/2023

The primary goal of the project is to improve our knowledge of the bedrock topography of Helheim Glacier. Bed topography is a primary input for ice sheet models, and filling in existing gaps will improve modelling results and understanding of ice sheet processes near the terminus. Previous airborne radar sounding measurements of Helheim Glacier from large fixed-wing manned aircraft have been unable to completely map the bed topography, especially near the terminus where the glacier is rapidly changing. Previous airborne survey using lightweight UAS over Russell Glacier have shown the utility of a high-frequency (operating near 30 MHz) radar sounder operating on a compact UAS in mapping the bed when compared to previous surveys.

PI: Fernando Rodriguez-Morales

Sponsor: Honeywell

12/01/2020-08/30/2021

The RADAR 2021 Consortium was created to address the fact that the need for increased Radio Frequency (RF) radar functionality for future tail number programs is driving the need for advanced radar technology development and miniaturization, because there will be little or no increase in radar weapon volume. New RF radar and packaging technologies, new Synthetic Aperture Radar (SAR) technology, and new Additive Manufacturing (AM) methods are being developed in order to be able to keep up with RF Integrated Circuit (RFIC) technology advances. This project consists of seven university collaborators.

PI: Emily Arnold

Sponsor: NSF

08/01/2019-07/31/2024

The ice sheet mass loss being observed in Greenland and Antarctica directly contributes to global Sea Level Rise (SLR). By the end of this century, scientists predict that changes in the polar ice sheets could contribute anywhere from tens of centimeters to almost two meters in SLR. This large uncertainty in future SLR predictions is due, in part, to insufficient measurements of bedrock topography and surface crevasses in the most critical regions of the ice sheets. These measurements are used by scientists in ice sheet models to predict contributions to SLR. The observational gaps in bedrock topography and surface crevasses limit scientists abilities to accurately model changes in the dynamic ice sheets. This project addresses this data need by equipping a small drone helicopter with a radar suite to produce fine-grid measurements of ice thickness, bed topography, and crevasses in critical regions of the ice sheet. Rising seas will have huge social and economic impacts on the entire global population especially to the estimated 150 million people living in coastal regions at elevations within 1 m of current sea level. The uncertainties in SLR predictions greatly inhibit our ability to properly plan for and adapt to our changing climate. The broader impacts of this work are not limited to reducing uncertainty in SLR predictions. This project also involves the training of post-secondary students in developing next-generation remote sensing technologies to better prepare them for 21st century careers. By integrating research and education, post-secondary students will gain practical experience via classroom design, build, and test projects. Through these projects, students will be exposed to the environmental and social issues that are driving the need for this new technology. The intellectual merits of this work encompass both the technological development of the new sensor-platform and the glaciological studies this tool will enable. The primary technological research goal is to extend the application of drones in environmental remote sensing by: 1) using a novel approach for antenna integration and multi-pass distributed array processing that overcomes major payload limitations of small drones, and 2) demonstrating an autonomous platform that is easier to operate yet has sufficient payload capabilities and is robust enough to conduct measurements in polar environments. The vehicle’s flight capabilities will enable crevasse mapping and bed topography data collection with a combined spatial extent and resolution that will allow scientists to study: 1) the effects of measurement resolution on modeling ice sheet dynamic processes; 2) the significance of bed topography on glacial behavior at multiple time scales; and, 3) crevassing mechanisms and correlating crevasse attributes to calving events.

Lead Institution: Alabama University

KU-PI: David Braaten

Sponsor: NOAA

06/01/2018-06/30/2021

The University of Kansas proposes to collaborate with the University of Alabama in the development of remote sensing technology to conduct measurements of snow and soil moisture as part of the newly established Remote Sensing Center (RSC) at the University of Alabama. David Braaten at KU will contribute to four of the six project focus areas, which are (1) Sensors & Miniaturization; (4) Field Programs & Data Collection; (5) Data Processing & Data Products; and (6) Models & Synthesis.

KU-PI: Shawn Keshmiri Business Collaborator: DarCorp

Sponsor: NASA SBIR

08/14/2019-08/14/2021

University of Kansas team will be developing a hybrid ML/AI learning flight controller deigned based on the RTA architect. The ML/AI learning flight controller will be supported by a secondary LQR safety flight controller, the RTA monitors the aircraft states and would switch from ML/AI to LQR controller in case safety thresholds are violated. KU team will also expand ML/AI autopilot policy and reward functions to explore new control schemes, to improve aircraft performance, and to reduce reaction times in safety-critical situations such as emergency landing or when aircraft is in loss of control conditions (LoC). The goal is to find methods to train policy and reward functions to be adaptive and robust around new performance goals (e.g. fuel efficiency, maintaining minimum velocity, etc.). The KU team will also develop required hardware and software and will perform validation flight tests.

PI: Daniel Gomez-Garcia

Sponsor: NASA

05/12/2020-05/11/2021

The Center for Remote Sensing of Ice Sheets (CReSIS) proposes to integrate an ultra-wideband snow probing radar onto the Vanilla VA001 Unmanned-Aerial-System (UAS), which is a ultra-long endurance UAS, to measure snow thickness over sea-ice and map near-surface internal layers of snow over land-ice in polar regions. The Vanilla VA001 aircraft is a long-endurance UAS capable of multi-day flights. Its payload capability and long-endurance feature makes it an ideal aircraft to perform large-coverage snow-probing measurements over sea-ice and land-ice at lower operating costs than those required for manned airborne operations.

PI: John Paden

Sponsor: NSF

05/15/2020-07/31/2022

The University of Kansas will modify the existing airborne accumulation radar to measure vertical velocity and produce fine resolution polarimetric 2D images of the ice base during the field seasons in year 1 and year 2. This accumulation radar (600-900 MHz ice sounder) was developed and deployed for the NSF-NERC Thwaites MELT project. The radar was deployed successfully three times on a British Antarctic Survey Twin Otter, but its chassis and antenna are not built for extended ground based work. The primary modifications are to 1) replace the chassis with a larger and more robust chassis to support ground deployments and provide more space for electronics required for multichannel operation, polarimetric operation, and higher power, 2) develop a polarimetric antenna sled, and 3) develop 1600W power amplifiers to replace the 400 W power amplifier in the current system. This radar will support improved vertical velocity measurements and produce 2D images of the ice base during the field seasons in year 1 and year 2.

Lead Institution: University of Washington

KU-PI: Leigh Stearns

Sponsor: NASA

07/22/2020-07/21/2023

The University of Kansas will provide a database of iceberg and sea ice concentrations around Greenland, derived bi-monthly, from 2015-present. In specific fjords, we will collect data at higher temporal resolutions (6 days), and derive iceberg trajectories for large icebergs. We will spearhead machine learning algorithms to create clusters of fjords, based on their geometric and climate characteristics.  

Lead Institution: University 

KU-PI: Leigh Stearns

Sponsor: NASA

07/22/2020-07/21/2023

The University of Kansas will spearhead the glaciology component of this project. That includes synthesizing glaciological data with topographic and climate products to assess how glaciers in HMA currently contribute to river discharge, and how they will contribute to discharge estimates in the future

PI: John Paden

Sponsor: NASA

06/22/2020-06/21/2022

We propose to integrate a Multichannel Snow Radar (MSR) onto the NASA P-3 Orion airborne science platform. CReSIS built the MSR under a previous NASA call associated with Operation IceBridge (OIB). The multichannel 2-18 GHz capability enables tomographic techniques that will generate 3D maps of snow distribution with unprecedented detail and coverage on sea ice and reliable direct sounding of snow on land. Over land ice, the additional antenna directivity will increase signal penetration and reduce off-nadir clutter which will improve detectability of deeper layers and precise mapping of interfaces. The current single-channel snow radar has been operated successfully on 25 missions as a part of NASA OIB since 2009. Where the surface layering is sufficiently flat, the snow radar does an excellent job of mapping snow thickness. However, over deformed sea ice and over moderately rough land snow, the current snow radar’s broad antenna footprint is not able to unambiguously resolve the top and bottom snow interfaces. The cross-track resolving capability of the MSR will refine snow thickness estimates and increase coverage over deformed and rough snow surfaces

KU PI: Shawn Keshmiri

Sponsor: NSF

08/15/2020-07/31/2023

PI: Leigh Stearns

Sponsor: National Science Foundation

07/15/2021-06/30/2023

Three military airborne campaigns (USAF: 1940s-1960s, USCG: 1970s, USN: 1970s-1980s) acquired tens of thousands of aerial photographs of Greenland. These photographs, which have never been digitized or used for science, can yield insight into glacier behavior – across the ice sheet – prior to satellite observations. We propose to digitize these images, render glacier surface elevations and terminus positions for the perimeter of Greenland, and estimate surface velocities where the data permits. In doing so, this project will expand both the temporal length and resolution of Greenland outlet glacier observations.

PI: John Paden

Sponsor: National Science Foundation

09/01/2021-08/31/2024

Earth’s polar ice sheets play a critical role in shaping sea level over geological time, yet ice-sheet response to contemporary climate change remains highly uncertain. Forty years of spaceborne observations reveal the recent acceleration in mass loss of the ice sheets through measured change in elevation, gravity, and ice-flow variation. It remains difficult, however, to use these satellite observations to predict future behavior, as they are the surface expression of non-unique subsurface processes. Airborne radar sounder measurements offer the potential to constrain non-unique ice-sheet surface dynamics because they can map out subsurface parameters (ice temperature/rheology, crystal-orientation fabric, englacial velocity, bed roughness, bed thermal state, and subglacial hydrology) on an ice-sheet-wide scale. These more advanced applications of radar data require joint interpretation of geometric and radiometric properties of radar data on a large scale.  However, five decades of radar data collection by multiple organizations deploying a range of systems with data distributed under inconsistent data policies and processing methods leads to siloed research resulting in lower efficiency and increased time to science. Open Polar Radar (OPoRa) brings together many of the data providers with the largest datasets and leverages an expert science team of collaborators in applied mathematics, radar engineering, glaciology, and artificial intelligence (AI), to produce standardized AI-ready data products, search services, and user tools firmly based on Findable Accessible Interoperable and Reusable (FAIR) principles to improve accuracy and reduce uncertainty in sea-level projection.

The OPoRa team covers 83% of Antarctica data and nearly complete Greenland and polar sea-ice coverage. For the first time, these datasets will be placed in common formats and made available through a common interface with a common set of tools for scientists via an end-user driven process.