Session: B.4 Oceanography
(Convener: )

Title: Sea Level Rise via GRACE and Altimetry: Analysis Results and Variability within the Water Column
Presenter: Dickey, Jean O.
Co-Authors: J.K. Willis, O. de Viron, W. Llovel

Abstract: This work has two major thrusts: 1) Determination of sea level rise (SLR) from 700 to 2000m, and 2) the investigation of SLR variability as a function of depth using SSA and M-SSA techniques. We use the analysis of (Llovel, 2014) to study three different slices of the ocean as a function of depths, 0-700m, 0-2000, and 700-2000m. Llovel et al. (2014) found that the contribution below 2000m was consistent with zero, when the uncertainties were considered.

The contribution from 700-2000m was 0.38 +/- 0.05 mm/yr, a significant amount explaining in part the current hiatus. Wavelet analysis and singular spectrum analysis (SSA) were used to study the time dependence; values of the first three periods agree considering the uncertainties in these cases. In both the SSA and multi-channel SSA (M-SSA), the first 8 modes were calculated. Internal variability of the ocean is investigated as a function of depth using Multi-Channel Singular Spectrum Analysis (M-SSA) and Wavelet analysis. Here we consider the ocean to have 20 layers, each 100m in depth, together covering the top 2000 m of the ocean. Each one of the ocean layers is considered a different channel (20 channels in total), which is analyzed as a function of time. The first mode was the largest and had the most robust signal active throughout the full column (0-2000m) with a period that is larger than the data series span considered. Modes 2-8 are concentrated near the surface. They appears have two ribbons of variability. The first signal is at sea level down to about 120-170 m; below the first ribbon, the second ribbon exists from 120-170 m down to 300-500 m. These signals are the fingerprint of the oceans internally rearranging heat on timescales of these periods. The results of these analyses will be presented with their implications explained.

Back to top

Improved estimates to global sea level from Ice Sheets, glaciers and land water storage from GRACE
Presenter: Velicogna, Isabella
Co-Authors: C. Hsu, E. Ciraci’, T. Sutterley

Abstract: We use observations of time variable gravity from GRACE to estimate mass changes for the Antarctic and Greenland Ice Sheets, the Glaciers and Ice Caps (GIC) and land water storage for the time period 2002-2015 and evaluate their total contribution to sea level. We calculate regional sea level changes from these mass fluxes using an improved scaling factor for the GRACE data that accounts for the spatial and temporal variability of the observed signal. To estimate the contribution of the GIC, we use a least square mascon approach and re-analyze recent inventories to optimize the distribution of mascon and recover the GRACE signal more accurately. We find that Greenland controls 43% of the global trend in eustatic sea level rise, versus 16% for Antarctica and 29% for the GIC. The contribution from the GIC is dominated by the mass loss from the Canadian Arctic Archipelago, followed by Alaska, Patagonia and the High Mountains of Asia. In Greenland, following the record 2012 summer melt, 2013 and 2014 experienced less mass loss. In Antarctica, the mass loss is on the rise with increased contributions from the Amundsen Sea sector and the Wilkes Land sector of East Antarctica, whereas the Queen Maud Land sector experienced a zero mass gain since 2013 following a large snowfall in 2009-2010. We compare sea level changes from the GRACE-derived mass fluxes with sea level change from satellite radar altimetry (AVISO) corrected for steric signal of the ocean using Argo measurements. We find an excellent agreement in amplitude, phase and trend between these estimates. This work was conducted at UC Irvine and at Caltech's Jet Propulsion Laboratory under a contract with NASA's Cryospheric Science Program.

Here we demonstrate a plausible approach to derive the Indonesian Throughflow (ITF) transport proxy using satellite altimetry SSH, GRACE OBP data, in situ measurements from the Makassar Strait from 1996-1998 and 2004-2009, and a theoretical formulation. We first identified the optimal locations of the correlation between the observed ITF transport through the Makassar Strait and the pressure gradients, represented by the SSH and OBP differences between the Pacific and Indian Oceans at a 1°x1° horizontal resolution. The optimal locations were found centred at 162°E and 11°N in the Pacific Ocean and 80°E and 0° in the Indian Ocean, then were used in the theoretical formulation to estimate the throughflow. The proxy time series follow the observation time series quite well, with the 1993-2011 mean proxy transport of 11.6±3.2Sv southward, varying from 5.6Sv during the strong 1997 El Niño to 16.9Sv during the 2007 La Nina period, which are consistent with previous estimates. The observed Makassar mean transport is 13.3±3.6Sv southward over 2004-2009, while the proxy gives an ITF mean transport of 13.9±2.5Sv using SSH data alone and 15.9±3.6Sv using a combination of SSH and OBP data for the same period.

Back to top

Title: Variability in deep ocean circulation from GRACE
Presenter: Boening, Carmen
Co-Authors: M.M. Watkins

Abstract: Although nearly impossible to observe on a global scale, total water mass transport and inter-basin exchange are central to understanding long-term changes in ocean circulation. Of particular interest are changes in the Meridional Overturning Circulation (MOC) as they pose potential impacts in continental climtae. However, in-situ observations are limited in space and time preventing a holistic view of current variability. The representation of long-term transports in ocean models is highly dependent on the atmospheric forcing fields, which may misrepresent real interannual variability. The bottom pressure observations from the Gravity Recovery And Climate Experiment (GRACE) provide for the first time the ability to observe this global water mass transport. Here, we present the first near-global maps of variability in the depth-independent ocean circulation derived from advanced analysis of GRACE data. We find that significant variability on annual to decadal time scales exists in the deep large-scale circulation, some of which are related to the Southern Annular Mode forcing dominating Southern Ocean variability.

Back to top

Inter-Ocean Transports from GRACE: A Case Study fron Indonesian Throughflow
Presenter: Song, Tony
Co-Authors: D. Susanto

Abstract: Inter-ocean transports are of fundamental interest to physical oceanography and ocean climate considerations, but are poorly understood and difficult to measure because long-term direct measurements of strait circulations are an expensive alternative and their implementation remains logistically challenging. We hypothesize that the magnitude and variability of inter-ocean transports vary with sea-surface height (SSH) and ocean bottom pressure (OBP) gradients between two inter-connected oceans.

Here we demonstrate a plausible approach to derive the Indonesian Throughflow (ITF) transport proxy using satellite altimetry SSH, GRACE OBP data, in situ measurements from the Makassar Strait from 1996-1998 and 2004-2009, and a theoretical formulation. We first identified the optimal locations of the correlation between the observed ITF transport through the Makassar Strait and the pressure gradients, represented by the SSH and OBP differences between the Pacific and Indian Oceans at a 1° x1° horizontal resolution. The optimal locations were found centred at 162° E and 11° N in the Pacific Ocean and 80° E and 0° in the Indian Ocean, then were used in the theoretical formulation to estimate the throughflow. The proxy time series follow the observation time series quite well, with the 1993-2011 mean proxy transport of 11.6±3.2Sv southward, varying from 5.6Sv during the strong 1997 El Niño to 16.9Sv during the 2007 La Nina period, which are consistent with previous estimates. The observed Makassar mean transport is 13.3±3.6Sv southward over 2004-2009, while the proxy gives an ITF mean transport of 13.9±2.5Sv using SSH data alone and 15.9±3.6Sv using a combination of SSH and OBP data for the same period.

Title: North Atlantic meridional overturning circulation variations from GRACE ocean bottom pressure anomalies
Presenter: Landerr, Felix
Co-Authors: D. Wiese, K. Bentel, M. Watkins, C. Boening

Abstract: Concerns about North-Atlantic Meridonal Overturning Circulation (AMOC) changes imply the need for a continuous, large-scale observation capability to detect changes on interannual to decadal time scales. Here, we present the first measurements of lower North-Atlantic-Deep-Water (LNADW) transport changes using only time-variable gravity observations from Gravity Recovery and Climate Experiment (GRACE) satellites from 2003 until now. Improved monthly gravity field retrievals allow the detection of North Atlantic interannual bottom pressure anomalies and LNADW transport estimates that are in good agreement with those from the ocean RAPID-MOCA array. Concurrent with the observed AMOC transport anomalies from late-2009 through early-2010, GRACE measured ocean bottom pressures changes in the 3000-5000 m deep western North Atlantic on the order of 20 mm-H2O, implying a southward volume transport anomaly in that layer of approximately -5.5 Sv. Our results highlight the efficacy of space-gravimetry for observing AMOC variations to evaluate latitudinal coherency and long-term variability.

Back to top

Title: Evaluating Atlantic MOC variability and coherence with GRACE ocean bottom pressure
Presenter: Bentel, Katrin
Co-Authors: F.W. Landerer; C. Boening; D.N. Wiese; M.M. Watkins

Abstract: The Atlantic Meridional Overturning Circulation (AMOC) is a key mechanism in basin-scale heat transport to high latitudes. It has significant impact in particular in the Northern Hemisphere and on Northwestern Europe’s climate. The dynamics of the AMOC especially in the North Atlantic have been observed and described in recent model and observational studies. However, in-situ observations are limited to a few latitudes where observational arrays are deployed. Also, the physical relationship between ocean bottom pressure (OBP) and the AMOC has been characterized in recent literature.

Here, we employ GRACE-derived OBP (from the JPL-RL05M mascons solution) to derive AMOC anomalies through physical relations. We validate our results by comparison to the few in-situ observations available. Furthermore, in contrast to in-situ observations, GRACE observations provide the unique opportunity to derive AMOC anomalies continuously (for the GRACE time period from 2003 until present) across all latitudes of the basin, and evaluate spatial and temporal coherence.

Back to top

Title: Using GRACE Ocean Bottom Pressure to Observe Transport Variability in the Southern Ocean
Presenter: Makowski, Jessica
Co-Authors: D.P. Chambers; J.A. Bonin

Abstract: Previous studies have shown that ocean bottom pressure (OBP), from bottom pressure recorders, models, and other OBP data sets, can be used to estimate the transport variability of the Antarctic Circumpolar Current. More recent studies have shown that OBP data from the Gravity Recovery and Climate Experiment (GRACE) is able to estimate the transport variability south of Australia within reasonable error bounds (Makowski, Chambers, and Bonin, 2015). We further investigate the transport variability of the Antarctic Circumpolar Current using the ocean bottom pressure data from CSR RL05 data, a higher resolution general circulation model (ECCO 2), and the JPL MASCONs. We will expand our study area over other regions of the Southern Ocean via a sliding window, in an effort to achieve a higher understanding of the general dynamics of the Southern Ocean circulation system.

Back to top

Title: Low-frequency transport variability in the Southern Ocean: the importance of regional variations
Presenter: Chambers, Don
Co-Authors: M. Kosempa, J. Makowski

Abstract: We use satellite altimetry, temperature/salinity from Argo, ocean bottom pressure from GRACE and the ECCO2 state estimate, as well as satellite vector winds to quantify and understand zonal geostrophic transport variability in the Southern Ocean. Altimetry and Argo data are used to estimate the transport variability above 2000 dbar, while GRACE and ECCO2 are used to measure the full-depth transport associated with the bottom current (i.e., the barotropic component). We find that for interannual periods, the transport variations are dominated by the Southern Annular Mode (SAM) variability, but that there are significant differences in the two estimates. The barotropic component is more highly correlated with SAM, suggesting either issues in the altimetry/Argo estimate or significant baroclinic differences. More importantly, we observe a significant difference in decadal trends between the Southern Indian Ocean and South Pacific, with different signs. This is found in both the GRACE and ECCO2 estimates, and is shown to be related to regional wind differences - the winds in the Atlantic/Indian Ocean sectors have been decreasing over the last decade while they have been increasing west of the Drake Passage, although zonally averaged winds show little change. These results have important implications for studies trying to estimate changes of the Antarctic Circumpolar Current transport at the Drake Passage.

Back to top

Title: A new GRACE-derived runoff time series for Arctic river basins as input for numerical ocean simulation
Presenter: Eicker, Annette
Co-Authors: C. Boening, D. Menemenlis, J.T. Reager, D. Wiese, M. Watkins

Abstract: Freshwater flux from rivers into the Arctic Ocean influences ocean circulation, sea ice distribution, and biogeochemistry, as it affects driving quantities such as salinity, temperature, and nutrients. Long term variability in these fluxes, therefore, impacts the Arctic environment and has attracted increasing attention from scientific community in recent years due to rapidly changing Arctic climate. In this study, we examine the influence of GRACE-derived river runoff used as freshwater input for numerical simulations of ocean and sea-ice.

Most numerical Arctic Ocean simulations introduce runoff as a climatology, which is derived using in-situ observations at gauging stations sampled over the past decades. Inter-annual runoff variability is neglected. The goal of our study is the computation of a continuous runoff time series with monthly resolution for the 6 largest Arctic river basins during the GRACE time span and the evaluation of its impact on regional Arctic Ocean simulations.

Introducing the observed runoff directly into the ocean model is difficult, as the availability of in-situ runoff observations varies strongly among the different basins. In addition, in-situ data coverage is very limited for the most recent years. To create a continuous time series, the terrestrial water balance equation (dS/dt = P-E-R) is used to compute runoff estimates (R) from GRACE storage changes (dS/dt) and from precipitation (P) and evaporation (E) provided by atmospheric reanalyses such as ERA-Interim. We show that after applying a simple calibration procedure using in-situ runoff observations in overlapping time spans, GRACE provides realistic monthly runoff estimates.

In this presentation, we will describe the computation of the GRACE-derived runoff time series and we will show first results of impacts on regional Arctic Ocean simulations.

Back to top

Title: Ocean tide solutions from GRACE range-rate data
Presenter: Ray, Richard
Co-Authors: B. Loomis, S. Luthcke

Abstract: A series of global tidal solutions have been computed from GRACE range-rate data for the major short-period constituents M2, S2, O1, K1, and the long-period constituent Mf. Preliminary solutions show that the GRACE time series is now sufficiently long that there is little "cross-talk" between constituents, so we are now computing the constituents individually in separate inversions, generally to degree/order 90. It appears necessary that non-tidal mass variability be separately modeled and removed from the GRACE residuals prior to attempting tidal solutions. We have solutions based on (1) using forward models of atmospheric, oceanic, and terrestrial water hydrology from ECMWF, OMCT, and GLDAS, and (2) using similar forward models augmented by prior GRACE global mascon solutions, which thus include cryospheric mass variability. Even though our prior mascon solutions included estimates of various GRACE state and accelerometer parameters, it appears essential that such arc parameters be re-estimated during the tidal inversions.

Global solutions for M2, O1, and K1 are physically reasonable at long wavelengths. In contrast, S2 is noisy and evidently contaminated by non-tidal 161-day variability. While all constituents are dominated by signals from poorly modeled polar seas, the M2 solution is sufficiently clean that it appears we can now also detect errors caused by the (inadequate) assumption of constant seawater density. The Mf solution, computed to degree/order 30, is dominated by two high-amplitude spots in the Gulf of Carpentaria and in the Gulf of Thailand extending to Karimata Strait; these anomalies are surely caused by nonlinear interactions between K1 and O1 which can generate energy at the frequency of Mf. Some very sparse tide-gauge data tends to confirm this.

Aside from S2, these tidal solutions can be used by other groups for processing GRACE data. However, they cannot be used in other applications until they are properly assimilated into a high-resolution numerical ocean tide model. Work along those lines is in progress.

Back to top

Title: Improving estimates of the ocean circulation and climate using GRACE-derived variable ocean bottom pressure fields
Presenter: Ponte, Rui
Co-Authors: K.J. Quinn; I. Fukumori; P. Heimbach; D. Wiese

Abstract: Global ocean bottom pressure (OBP) measurements made from space have been accumulating since the launch of GRACE in 2002. Insights into the utility of the novel dataset can be gained by combining it with other more traditional observations, such as altimetry and in situ hydrography, and using it to constrain models of the ocean circulation and climate. Such efforts have been a recent focus of the project on “Estimating the Circulation and Climate of the Ocean” (ECCO), which provides global ocean state estimates using advanced optimization and modeling tools. Comparisons of monthly OBP fields derived from the latest ECCO solution release, done without using any OBP constraints, and from GRACE spherical harmonics reveal a number of issues with both GRACE and ECCO (e.g., noisy data near some coastal regions prone to land leakage issues and apparent excess energy in ECCO solution in the Arctic). These comparisons are used to estimate GRACE uncertainties and derive weights for properly adding OBP data as a new constraint for the ECCO solutions. Preliminary results from the ongoing optimization with OBP constraints indicate a significant impact of the GRACE fields on estimates of OBP in several mid and high latitude regions. Further possible improvements are suggested by assessing results against newly released JPL mascon OBP fields, which in several regions attain better signal-to-noise ratios than OBP fields based on GRACE spherical harmonics.

Back to top

Title: Assessing GRACE Mean Gravity Models using Altimetry and Drifter data
Presenter: Knudsen, Per
Co-Authors: O. Andersen; N. Maximenko

Abstract: A series of newer mean gravity models including data from GRACE are combined with the DTU13MSS mean sea surface to derive models for the Mean Dynamic Topography (MDT). The series of GRACE based MDT models are compared in regional analyses to identify differences and to quantify quality measures associated with the models. By using Fourier techniques the spectral characteristics are obtained as well as their anisotropic patterns.

Then, regional analyses are carried out using in-situ observations of the geostrophic surface currents. This is done to analyse correlations and to derive resolution capacities of the MDT models. Also this information is used as quantified quality measures associated with the GRACE only and combination gravity models.

Back to top

Assessing GRACE Mean Gravity Models using Altimetry and Drifter data">Title: Assessing GRACE Mean Gravity Models using Altimetry and Drifter data
Presenter: Piecuch, Christopher G.
Co-Authors: R.M. Ponte

Abstract: A wind-driven, spatially coherent mode of nonseasonal, depth-independent variability in the Canadian inland seas (Hudson Bay, James Bay, and Foxe Basin) is identified based on GRACE as well as a tide gauge record and a barotropic model. The mode is related to the North Atlantic Oscillation and associated with net flows into and out of the Canadian inland seas. The anomalous inflows and outflows, which are reflected in mean sea level and bottom pressure changes, are driven by wind stress over Hudson Strait, possibly related to wind setup, and over the northern North Atlantic Ocean, potentially mediated by various wave mechanisms. This mode is also associated with internal mass redistribution within the Canadian inland seas, reflecting the linear response to local wind stress under the combined influences of rotation, gravity, and topography. Results exemplify the utility of GRACE for studying circulation and climate in a boreal marginal sea.

Back to top