Decadal Strengthening of Interior Flow of North Atlantic Deep Water Observed by GRACE Satellites

Abstract

The Gravity Recovery and Climate Experiment (GRACE) satellite mission provides information on changes to the Earth's gravity field, including ocean mass. Long-term trends in GRACE data are often considered unreliable due to uncertainties in the corrections made to calculate ocean mass from the raw measurements. Here, we use an independent estimate of ocean mass from satellite altimetry and in situ density data from five mooring sites and repeat hydrography to validate trends in GRACE over the North Atlantic, finding substantial agreement between the methods. The root-mean-square difference between ocean mass changes calculated with this method from the mooring data and those measured by GRACE is 3.5 mm per decade, much lower than the mean signal of 15.6 ± 1.8 mm per decade for GRACE and 17.8 ± 5.2 mm per decade for the altimetry-mooring estimate. The GRACE ocean mass data are then used to study the change in the deep circulation of the North Atlantic between the 1 April 2002 to 31 March 2009 and 1 April 2010 to 31 March 2017 periods, revealing a large-scale anticyclonic circulation anomaly off the North American coast. The change is associated with an increase of 13.9 ± 3.3 Sv (1 Sv = 10 6 m 3 s −1) of southward North Atlantic Deep Water flow in the interior between 30 • N and 40 • N, largely balanced by a northward anomaly of 10.7 ± 3.3 Sv for the boundary circulation. This implies an increased importance of interior pathways compared to the Deep Western Boundary Current for the spreading of North Atlantic Deep Water, which constitutes the lower limb of the Atlantic Meridional Overturning Circulation. Plain Language Summary The Gravity Recovery and Climate Experiment (GRACE) satellites measure the Earth's gravity field, which can be used to study ocean currents. In this study, we validate long-term trends in the data using independent ocean measurements to confirm that they reflect real changes in the ocean. In the Atlantic Ocean, North Atlantic Deep Water flows southward from its formation region near Greenland, spreading throughout the ocean basin. The flow of North Atlantic Deep Water is a crucial part of Earth's climate system, making it an important process to understand. Most of the water spreads southward in the so-called Deep Western Boundary Current along the continental shelves of North America, but there is also a significant flow further offshore in the interior of the basin. We use the GRACE data to measure changes to these two pathways between the 2002-2009 and 2010-2017 periods. The results show that southward flow in the interior has strengthened during this time, with opposite changes near the continent. This marks a significant change to how this water spreads from the formation region.

Weakening Trend of Luzon Strait Overflow Transport in the Past Two Decades

“…With the time mean removed, the OBP changes manifest uptrend variations with a linear trend of 3.29 ± 0.10, 3.39 ± 0.10, and 3.03 ± 0.09 mm yr −1 at 19, 20, and 21°N, respectively, indicative of a significant rising trend of the sea surface height (SSH) induced by ocean mass changes over the past decades (Figure 1c). These rising rates are greater than the trend of 2.38 mm yr −1 in mid‐latitude North Atlantic between 2002 and 2017 (Koelling et al., 2020), and 1.80 ± 0.47 mm yr −1 in global mean ocean mass contribution between 2005 and 2011 (Chen et al., 2013), but smaller than the trend of 6.1 ± 0.7 mm yr −1 averaged in the northwestern Pacific between 2003 and 2011 (Cheng et al., 2013). The OBP rise over the past two decades reflects inhomogenous spatial responses of the OBP to ocean mass redistribution, corresponding with changes in geostrophic transport.…”

“…The main reason for us to use the ECCO OBP data rather than the Jet Propulsion Laboratory (JPL) GRACE OBP data is that the JPL GRACE OBP data is purely derived from the GRACE satellite mission with very coarse horizontal resolution and smoothed spatially with a 500 km Gaussian smoothing, thus not appropriate to resolve the Luzon Strait overflow. In addition, the JPL GRACE raw data may have possible uncertainty in the long‐term OBP trends (Watkins et al., 2015), which has to be solved by either detrended the data (Landerer et al., 2015; Worthington et al., 2019) or validated the trends (Koelling et al., 2020). As the GRACE satellite mission only provides OBP anomaly, that is, gravity field variation relative to the mean, the OBP data is not applicable for seeking an accurate time‐mean transport, but suitable for investigating changes relative to the mean transport.…”

Decadal Weakening of Abyssal South China Sea Circulation

“…Considering large-scale geostrophic flow in the deep layer, the change in deep flow at depth z is determined by geostrophic balance (e.g., Koelling et al, 2020), where p is pressure, 𝐴𝐴 𝐴𝐴0 is the typical density in the SCS, f is the Coriolis parameter, and primes represent changes relative to the mean. The pressure difference in the above equations can be approximately calculated by horizontal OBP difference as examined by Hughes (2008, 2009).…”

Revisiting AMOC Transport Estimates From Observations and Models

“…There is an urgent need to address these issues so that the observational methods can be used not only to evaluate latitudinal consistency of AMOC characteristics in observations but also in the analysis of model simulations of future AMOC changes. Recent efforts combining the GRACE satellite data sets with in situ data from moorings and repeat hydrography can provide much-needed reliable estimates of deep ocean transport variability via bottom pressure estimates (Koelling et al, 2020; see also Text S2). Our results also suggest that altimetry could potentially provide another solution for estimating time-varying reference velocities required by the end-point geostrophic arrays.…”

“…For the MOVE site, the assumption of monitoring only the southward DWBC–associated with the North Atlantic Deep Water (NADW)—flow to infer AMOC variability is supported by numerical simulations (Kanzow et al., 2008). In addition, the assumption that there is a level‐of‐no‐motion at depth on long timescales (Send et al., 2011) seems to be supported, at least for the period covering the last two decades, by a recent study based on data from the Gravity Recovery and Climate Experiment mission (GRACE; Koelling et al., 2020).…”