Doctor of Philosophy, The Ohio State University, 2024, Earth Sciences

Seasonal snow plays a large role in the water cycle and local ecosystem dynamics in snow dominated regions. However, two characteristics of the snowpack, the snow water equivalent (SWE) and density, are challenging to measure at scale. Modeling and remote sensing allow for the estimation of these characteristics at wide spatial scales, but practical limitations remain on our ability to estimate at a fine spatial fidelity, wide spatial extent, and daily temporal resolution. Regional Climate Models (RCMs) have been shown to successfully estimate SWE at basin-wide scales but remain too computationally expensive to run at sub-kilometer resolutions over large domains. In this thesis, I present two alternative methods to estimate daily SWE at a high spatial and temporal resolution on a basin-wide scale. The first, Blender, presented in Chapter 2, merges 9 km RCM estimates of SWE, precipitation, and top of the snowpack energy balance from the Weather and Research Forecasting (WRF) model with remotely sensed snow cover fraction (SCF) measurements to produce 500 m estimates of SWE timeseries. Blender re-solves the mass and energy balance of the snowpack with a constrained non-linear optimization, forced by the timing of the snow on and off dates from the SCF data. Compared against 50 m LiDAR estimates of SWE from 18 Airborne Space Observatory (ASO) flights, Blender has an average spatial RMSE of 11.5% of maximum SWE, while the prior from WRF has an average spatial RMSE of 17% of maximum SWE. The mean absolute bias of the total basin snow water storage (SWS) for the Blender estimates is 7.3% in the winter, and 31.6% for the WRF prior. This method, Blender, requires ~ 20% extra computing time on top of the original WRF runs, and improves both the spatial RMSE and basin SWS absolute bias, all while better matching the melt timing to the remotely sensed SCF. In Chapter 3 we present the second method, Linear Blender, is a linearized version of Blender, Chapter 2. This meth (open full item for complete abstract)

Committee: Michael Durand (Advisor); Ian Howat (Committee Member); Jim Stagge (Committee Member); Demián Gómez (Committee Member) Subjects: Earth; Environmental Science; Geography

Doctor of Philosophy, The Ohio State University, 2022, Earth Sciences

Ice sheet internal temperature is a first-order control on glacier dynamics. However, measurements of ice sheet internal temperature, both from direct in situ and indirect remote sensing techniques, are lacking. The ultra-wideband software-defined microwave radiometer (UWBRAD) was designed to estimate internal ice sheet temperature (𝑇i ) by measuring microwave brightness temperatures (𝑇 b) from 0.5 GHz to 2 GHz. 𝑇i was then retrieved from 𝑇b . In this paper, we first presented a retrieval algorithm and a simulation study to assess its feasibility. Then we apply the algorithm to real instrument measurements over a Greenland flight path and produce temperature profiles of the corresponding places. The estimated temperature profiles were assessed by the abilities to realize three science goals: the retrieval of a) 𝑇i at 10 m depth to within 1 K; b) vertically- averaged 𝑇 to within 1 K; and c) the vertical 𝑇 profile to within 1 K RMSE. First, a “Virtual Experiment” performed using synthetic UWBRAD observations indicates that the science goals are achievable with the caveats that ice thickness and UWBRAD 𝑇b precision impact performance. Assuming a UWBRAD 𝑇b precision of 0.5 K, and for places where ice sheet thickness is less than 3 km, all science goals can be achieved. Second, the application using real UWBRAD measurements similarly demonstrated that the goal of 1 K precision is able to be achieved at most places where ice thickness is less than 3 km. These real data experiments focused on the role of so-called “nuisance parameters”, factors related to the density profile that also affect the UWBRAD measures. A novel approach is presented to estimate the nuisance parameters at the boreholes, and develop prior information for estimates away from boreholes. Experiments demonstrated that UWBRAD measurements combined with prior information improve the uncertainties in ice sheet internal temperature estimates compared an estimate based entirely on prior information. Finall (open full item for complete abstract)

Committee: Michael Durand (Advisor); Joel Johnson (Committee Member); C.K Shum (Committee Member); Kenneth Jezek (Committee Member); Ian Howat (Committee Member) Subjects: Earth

Doctor of Philosophy, The Ohio State University, 2021, Electrical and Computer Engineering

Sea ice is one of the most important components of Earth's cryosphere that regulates heat flow between the ocean and atmosphere, impacts water cycles and oceanic currents due to salt transport during the melt and freeze seasons, provides a natural habitat for many life forms, and affects global transportation. Remote sensing of sea ice has become more important than ever given the substantial reduction in sea ice extent in the Arctic in recent decades. A major advantage of remote sensing for sea ice studies is the continuous monitoring available from space that eliminates the need for expensive and difficult in-situ measurements. Although satellite-borne passive microwave radiometer observations of sea ice have been performed since the 1970s, significant challenges remain. Radiometer measurements at frequencies greater than approximately 19 GHz can detect sea ice concentration and extent but are impacted by clouds. L-band radiometers operating at 1.4 GHz have been shown to successfully estimate sea ice thicknesses up to ~1.5 m; however, thickness sensing performance is limited by the penetration depth in sea ice at this signal frequency. Radiometers operating at lower frequencies have not been used on satellites to date due to the strong man-made radio frequency interference (RFI) present at frequencies less than 1.4 GHz. The Ultra-Wideband Software-defined Microwave Radiometer (UWBRAD) was developed at The Ohio State University to observe thermal emissions in the presence of RFI. The instrument performs RFI detection and mitigation algorithms at multiple frequency channels from 0.5 – 2.0 GHz. As a result, UWBRAD is capable of retrieving sea ice thickness for ice thicknesses greater than 1.5 m due to its lower operating frequencies. In addition, the measurement of sea ice thermal emissions over the 0.5-2 GHz spectrum reveals additional information on ice characteristics and can allow the simultaneous retrieval of sea ice thickness and salinity. This dissertati (open full item for complete abstract)

Committee: Joel T. Johnson (Advisor); Robert Lee (Committee Member); Caglar Yardim (Committee Member); Kenneth C. Jezek (Committee Member) Subjects: Electrical Engineering; Electromagnetics; Remote Sensing

Doctor of Philosophy, The Ohio State University, 2020, Earth Sciences

The Greenland Ice Sheet (GrIS) is losing mass at accelerated rates in the 21st century, due in part to faster flow at large outlet glaciers. Chapter 2 presents work published in The Cryosphere (King et al., 2018). Here, we sample rapid changes in thickness and velocity at all large outlet glaciers to derive the first continuous, GrIS-wide record of total ice sheet discharge, or the volume of ice glaciers export, for the 2000-2016 period. We resolve a distinct pattern of seasonal variability with an amplitude of 6%, and analyze how seasonal to annual variability in the discharge time series relates to both meltwater runoff and glacier front position changes over the same period. We find that the annual magnitude of discharge is closely related to cumulative front position change (r2 = 0.79), averaging over 2 km of retreat since 2000. We find that larger seasonal quantities of runoff do not relate to increased annual discharge, although seasonal acceleration of ice discharge does closely coincide with the onset of the melt season. These results suggest that changes in glacier front position drive secular trends in discharge, whereas the impact of runoff is likely limited to the summer months when observed seasonal variations are substantially controlled by the timing of meltwater input. In Chapter 3, we extend our 2000-2016 discharge time series to the period 1985-2018, combining more than three decades of GrIS-wide observational products of outlet glacier velocity, elevation, and front position changes, and compare decadal variability in discharge with calving front position. We find that the close relationship between frontal change and ice discharge identified over the 2000-2016 record holds true for the 34-year record, and that increased glacier discharge can be attributed almost entirely to the retreat of glacier fronts, rather than inland ice sheet processes, such as changes in meltwater runoff. Discharge sensitivity to retreat is remarkably consistent across (open full item for complete abstract)

Committee: Ian Howat (Advisor); Lonnie Thompson (Committee Member); Michael Durand (Committee Member); Bryan Mark (Committee Member) Subjects: Climate Change; Earth; Environmental Studies; Geological; Geophysical; Geophysics

MA, University of Cincinnati, 2007, Arts and Sciences : Geography

In April 2005 near Barrow, Alaska, a Differential Global Positioning Systems (DGPS) survey was conducted along a snowdrift formed by a 2.2 km long snow fence. A snow machine pulled a sledge equipped with the DGPS, recording geographic location and elevation along transects parallel to the snow fence. Empirical measurements of snow depth were collected with a calibrated probe along transects perpendicular to the snow fence, and were used for an accuracy assessment. Five GIS models were established each using different interpolation methods to estimate snow depth. The base map was a Digital Surface Model of the snow-free surface derived from Interferometric Synthetic Aperture Radar (IFSAR) data. A natural neighbor interpolation algorithm provided the most accurate snow depth estimates with an RMSE of 21.29 cm. The snow drifts cover an area of 227,150 m 2; the mean snow depth is 1.87 m and occupies a volume of 425,974 m 3.

Committee: Dr. Kenneth Hinkel (Advisor) Subjects: Geography

Doctor of Philosophy, The Ohio State University, 2010, Geological Sciences

Continuous monitoring of changes in the Greenland ice sheet from both space and air borne sensors has been conducted since the early 1970's. Since the mid-1990's dramatic changes occurring on the Greenland ice sheet have been observed both from space borne sensors and field work. These changes, primarily mass loss from the ice sheet, are related to the observed trend of earth's warmer climate in recent decades both in peer reviewed journals and in popular media. This dissertation addresses two parameters that contribute to Greenland ice sheet mass balance estimates. The first factor is characterization of surface melting of the Greenland ice sheet from satellite-based passive and active microwave sensors. We use a wavelet based edge detection technique to delineate surface melt from brightness temperature measured by passive microwave sensors. Along with brightness temperature data, we also use normalized backscatter data from the Quick Scatterometer (QuikSCAT) as an independent sensor for comparison with the radiometer derived results. We use a semi-empirical threshold based method for surface melt detection from QuikSCAT. Our results show a step-like, consistent increase in melt area of the Greenland ice sheet since 1995. This step-like increase is also observed in the mean summer air temperature along portions of the Greenland coast. The 1995 step-like increase of melt area (and melt index, a measure of melt intensity) is correlated with a distinct change of the North Atlantic Oscillation (NAO) index (from positive to negative) after 1995. The second factor is mass accumulation in the upper reaches of the ice-sheet. We use an empirical model that correlates mean annual brightness temperature to annual accumulation rate. We apply a microwave emission model for the dry snow region of Greenland to show that 37 GHz vertically polarized brightness temperature data are better suited to capture the inter-annual variability of snow accumulation. Using our model we esti (open full item for complete abstract)

Committee: Kenneth Jezek (Advisor); Joel Johnson (Committee Member); Ralph von Frese (Committee Member) Subjects: Geophysics

Master of Science, The Ohio State University, 2010, Atmospheric Sciences

A polynya is an area of open water or less thick ice compared to its surroundings that forms in polar regions during weather conditions that would normally produce thicker sea ice, and releases the most sensible and latent heat flux to the atmosphere. This thesis incorporates satellite imagery from optical, thermal, and microwave sensors to investigate the climate and cryosphere of the North Water polynya, the largest, reoccurring polynya in the Arctic. Ecologically, it is known to be of significance to marine animals, e.g. narwhals and beluga whales. The polynya normally results from the formation of an ice arch at the southern limit of Nares Strait that prevents multi-year Arctic sea ice drift through the strait. The ice arch usually collapses by early to mid-summer, thus ending the polynya. In a dramatic display of climate-cryosphere dynamics, the winter 2008-2009 sea ice arch consolidated, 500 km north of its normal position. Summer 2009, sea surface temperatures were 1.5 °C above normal in central Nares Strait. In June 2009, 25 km resolution passive microwave sea ice concentrations were 33% below normal. 12.5 km resolution passive microwave sea ice concentrations were 20% below the 2002-2009 average north of Smith Sound in southern Nares Strait, and 60% below normal in northern Baffin Bay. Optical satellite imagery are used to measure area and length changes of the Petermann and Humboldt glaciers that empty into Nares St., where area losses of 215.4 km2 and 175.2 km2, respectively, are observed since end of summer 2000; an area 4.5 times that of Manhattan Island, New York, USA (87.5 km2), ranking them among the top-area losses among Arctic glaciers. Polar MM5 simulations have been made for each season to calculate the net snow accumulation, melt water runoff, and surface mass balance for 2000-2009. Snow line detection data has been developed to approximate the equilibrium line altitude. Results at Humboldt glacier indicate a change in the snow line associated (open full item for complete abstract)

Committee: Jason Box Dr. (Advisor); Jay S. Hobgood Dr. (Committee Member); Jeffery C. Rogers Dr. (Committee Member) Subjects: Atmosphere; Oceanography; Remote Sensing

Master of Science, The Ohio State University, 2009, Atmospheric Sciences

Supra-glacial melt lakes form in the Greenland ice sheet ablation region in response to surface melt. The Jakobshavn Ablation Region (JAR) in west – central Greenland (68.2 – 68.8°N) is an area with a high areal concentration of melt lakes, providing an ideal region to study melt lake development. Moderate Resolution Imaging Spectroradiometer (MODIS) imagery are acquired for the 2000 – 2008 melt seasons (days 150 – 274) to observe the spatial and temporal melt lake characteristics. Knowing that melt rates vary with elevation, JAR is divided into five elevation zones of 250 m intervals, between 585 – 1835 m above sea level. An empirically-derived depth function, based on MODIS optical reflectance, is applied to classified melt lake pixels at JAR, yielding depth, area and volume statistics. There is a strong correlation between melt lake area and volume quantities, regardless of elevation. Peak zonal fractional melt area, volume maxima and peak mean melt lake depth are reached at the mid-ablation zone (1035 – 1334 m). Melt intensity is determined from a Positive Degree Day (PDD) model. A correlation is found between melt lake area and volume anomalies and PDD anomalies that decrease with elevation. The melt season at the uppermost elevation (1585 – 1834 m) begins five weeks after the onset at the lowest elevation (585 – 834 m). The date of maximum area and volume also increase with time, with a difference of 50 - 60 days. Average melt season at JAR lasts 70 – 85 days below 1584 m and decreases to 30 days at the uppermost zone (1585 – 1834 m). To verify MODIS-derived lake area accuracy, three IKONOS 1 m resolution images are compared for a single lake, Lake Disco (67.23°N, 48.61°W). Uncertainty in MODIS estimates of area are 20±2%. Such differences confirm the difficulty of identifying depth values between 0 and ~2.5 m from the Box and Ski (2007) lake depth-retrieval classification. Shallow depths prevent the MODIS sensor's coarse resolution from identifying lake per (open full item for complete abstract)

Committee: Jason Box (Advisor); Ian Howat (Committee Member); Bryan Mark (Committee Member) Subjects: Earth; Geography; Remote Sensing