USAP-DC

This is a project jointly funded by the National Science Foundation’s Directorate for Geosciences (NSF/GEO) and the National Environment Research Council (NERC) of the United Kingdom (UK) via the NSF/GEO-NERC Lead Agency Opportunity. This Lead Agency Opportunity allows a single joint US/UK proposal to be submitted and peer-reviewed by the Agency whose investigator has the largest proportion of the budget. Upon successful joint determination of an award recommendation, each Agency funds the proportion of the budget that supports scientists at institutions in their respective countries. Sea ice plays a critical role in the exchange of heat and gases between the ocean and atmosphere and provides essential habitat for globally important wildlife. Antarctic sea ice has become a new focus of scientific interest due record-breaking low ice extents in recent years, which has only been possible to determine from the 50-year record of sea ice concentration (SIC) derived from satellite remote sensing. Passive microwave (PM) derived datasets of sea ice provide unprecedented insight into decadal scale trends in sea ice conditions, but these products are inadequate because they do not provide uncertainty associated with this record. For example, when PM data calculates that a locality has a SIC of 50% (corresponding to 50% of the ocean’s surface being covered by ice), the uncertainties and error associated with this measurement remain undefined. In this proposal, a comprehensive assessment of the spatial and temporal uncertainties associated with the SIC values derived from PM data will be conducted. To quantify these uncertainties, PM derived SIC values will be compared to finer-scale, high resolution SIC products that have been derived by applying cutting edge machine learning tools to other satellite imagery. A comprehensive understanding of differences in accuracy and uncertainty of the PM derived SIC values over space and time will be developed. Passive microwave derived sea ice concentration data are possibly the most utilized satellite product for the polar regions; however, there is a lack of rigor in their assessment of uncertainty on both temporal and spatial scales, and a diverse range of algorithms exist that can be used to derive SIC value from PM measurements. The fundamental difficulty in calculating uncertainty in these products results from the relatively coarse size of passive microwave pixels, which precludes the use of ground-truthing methods as viable routes to assess their accuracy over space and time. Two recent developments will allow the team to overcome these challenges. This project will leverage the step-change increase in the rate of satellite image acquisition with petabytes of finer resolution satellite imagery pertaining to sea ice conditions now being available to validate the accuracy of the PM SIC pixels. Advances in the capabilities of digital technologies will enable automation of the storage, processing, and analysis of these datasets to validate PM SIC data at unprecedented temporal and spatial scales. Digital technologies and machine learning (ML) methods will be applied to remote sensing imagery to quantify the spatial and temporal uncertainties associated with the calculation of Antarctic SIC derived from PM data at a decadal, pan-Antarctic scale. ML techniques will also be developed to automatically generate fine-resolution sea ice concentration charts from multispectral and radar satellite imagery. These high-resolution products will be used to quantify the uncertainty in PM derived SIC charts and conduct sensitivity analyses to diagnose the primary cause of uncertainty within the PM products. This award reflects NSF''s statutory mission and has been deemed worthy of support through evaluation using the Foundation''s intellectual merit and broader impacts review criteria.

Person	Role
Stroeve, Julienne	Investigator and contact
Meier, Walter N.	Investigator

Unknown Program

Award # 2533209

USAP-2533209_1

None in the Database