Exploring the Relationship Between Upper Ocean States and the Falling Ice Radiative Effects using ECCO Product and Global Climate Models

Abstract This study aims to investigate the relationship between upper ocean current (UOC) anomaly (above 200 meters) and surface wind stress (TAU) regarding the impact of falling ice (snow) radiative effects (FIREs) over the tropical and subtropical Pacific. To achieve this, we conducted sensitivity experiments using the CESM1-CAM5 model with FIREs turned off (NOS) and on (SON) using the Coupled Model Intercomparison Project phase 5 (CMIP5) historical run setting. The monthly ocean current reanalyses from NASA JPL Estimating the Circulation and Climate of the Ocean (ECCO) ocean reanalysis, derived from satellite measurements, serve as a reference for this study. The spatial patterns of the difference in horizontal UOC anomaly (UOCA) between the experiments with NOS and those with SON strongly correlate with the TAU patterns over the studied domain. When compared to the experiments with NOS, the experiments with SON demonstrate an enhancement in the annual mean UOC. The improvement in UOC can be attributed to the enhancements in TAU, specifically in the trade-wind regions. The enhancements in TAU play a significant role in influencing the UOCA patterns and contribute to the overall improvement observed in the experiments with SON. In SON, the average absolute bias of simulated UOCA over the study area is reduced by up to 30% compared to NOS against ECCO. Despite the biases observed in UOC over the south and north flanks of the equator in SON, the annual mean biases in improved ocean currents are closely linked to the improvements in TAU resulting from the presence of FIREs. Specifically, stronger ocean currents magnitudes are associated with stronger TAU changes through Coriolis forces. When we examine the ensemble mean absolute biases of UOC from the CMIP5 models, similarities to NOS are primarily observed over the South Pacific region. Keywords: Upper ocean currents, falling ice radiative effects (FIREs), surface wind stress (TAU)   1. Introduction Ocean circulation plays an important role in balancing global water mass distribution. As proposed by Yu and McPhaden (1999) and Knox (1976), the momentum of surface and subsurface ocean currents in the equatorial ocean under the mean state corresponded with the local wind pattern, suggesting that the observed currents were mainly in response to the behavior of the wind. Since ocean circulation is strongly associated with wind field variations, it is of great essentiality to verify the relationship between ocean current field and wind intensity. Besides, the prevailing wind exerts wind stress on the ocean and thus leads to the heating or cooling effect, as noted by Sverdrup (1947). Such an effect serves as a crucial mechanism for supplying the energy for the maintenance of permanent currents and the water mass distribution. Additionally, ocean circulation is responsible for transporting heat; therefore, its variations modify the climate system. The ocean current intensity is related to the strength and position of the prevailing wind. For example, the intensified North Equatorial Current was found in the western Pacific in the long run, which could be related to the Sverdrup dynamics and the wind stress curl variation (Hsin, 2016). On the other hand, the intensified eastward current anomaly was recorded during the 1996–1998 El Niño–Southern Oscillation (ENSO) event due to the burst of westerly anomaly (Grodsky et al., 2001). The equatorial zonal currents velocities reflect largely on the location of maximum sea surface temperature (SST) anomalies during the development of El Niño (Wang & Wu, 2013). The direction of equatorial upper layer currents also serves as a predictor of the super El Niño events as the climatological zonal currents reverse eastward during the developing phase (Kim & Cai, 2014). Furthermore, in terms of thermocline feedback (TH) and zonal advective feedback (ZA), both the zonal and vertical scales of ocean currents act as the dominant feedbacks in the growth and phase transition of ENSO (Ren & Jin, 2013). Hence, the realistic simulation of tropical ocean currents in coupled climate models is essential given its importance in the realistic simulation of tropical air-sea dynamics. Observational studies have shown that the spatial distributions of sea surface height (SSH), surface wind stress (TAU), and SST are strongly interconnected, particularly in tropical regions (Zhang et al., 1997; Casey & Adamec, 2002; Text 1 in Supplementary Information SI), highlighting the crucial role of their interactions in comprehending ocean dynamics and processes, including upwelling and downwelling phenomena. The prevailing trade winds exert a significant influence on the near-surface ocean water, causing it to move westward and accumulate in the western Pacific warm pool region, resulting in higher SSH in that area and lower SSH over the eastern Pacific. These changes in SSH, primarily driven by variations in the prevailing trade winds, subsequently induce alterations in SST (Casey & Adamec, 2002; Flato et al., 2013; Yang et al., 2022). The interplay between SSH, TAU, and SST highlights their collective influence on oceanic patterns and the broader climate system. Therefore, the accuracy of future SSH projections hinges on the ability of general circulation models (GCMs) to accurately simulate the present-day mean state of SSH. Additionally, the ability of these models to replicate realistic ocean currents is crucial for improving simulations of projected climate variability. However, there are notable discrepancies among GCMs in their simulations of present-day ocean currents mean states related to biases in TAU (Flato et al., 2013). For instance, Landerer et al. (2014) and Li et al. (2022b) demonstrated that CMIP5 models (Taylor et al., 2012) exhibited noteworthy biases in simulating SSH anomalies (SSHA) compared to observations, particularly in tropical regions. These biases were generally associated with TAU biases. The aforementioned findings emphasize the significance of precisely capturing the complex interplay between SSH, TAU, and SST within GCMs but not yet on the interplay between TAU and upper ocean currents (UOC). Moreover, in the Atlantic, Xu et al. (2014) mentioned that the severe SST biases along the coast of southern Africa are associated with deficient simulated Benguela upwelling. In the western equatorial Indian Ocean region, Fathrio et al. (2016) pointed out that the biases in the ocean mixed layer and ocean circulations contribute to the SST bias. By ensuring the faithful representation of these interactions, we can improve our comprehension and forecasting of future climate dynamics. The accurate simulation of SSH, TAU, and SST in GCMs is essential for achieving more reliable projections and a deeper understanding of the intricate mechanisms driving climate change. Understanding the impact of missing atmospheric processes on modeled TAU is crucial, as SSH, ocean currents and SST are influenced by TAU. Li et al. (2014, 2015) discovered that a significant number of CMIP5 GCMs do not account for the radiative effects of falling ice (snow), known as the falling ice radiative effects (FIREs). This omission leads to weakened TAU and overly warm SSTs over subtropical and tropical oceans (Li et al., 2012, 2013), which can be explained by a mechanism described below. The inclusion of FIREs in the models improves the radiation fields, TAU, and SST simulations in the tropical climate states of CMIP5 (Li et al., 2014, 2015; Text 2 in Supplementary Information SI). A brief summary is the following: Li et al. (2015) and Supplementary Information (SI) highlighted that CMIP5 models tend to exhibit excessively strong convection, resulting in anomalous low-level outflows over regions such as the Intertropical Convergence Zone (ITCZ), South Pacific Convergence Zone (SPCZ), and Maritime Continent. This phenomenon leads to a weakening of TAU and upper ocean mixing, ultimately causing warmer SSTs in the trade-wind regions (Li et al., 2014, 2016, 2018). These findings underscore the importance of considering the impact of missing physical processes in the atmosphere on TAU, as it directly influences both SSH (Figures 1 and 2 in Li et al., 2022b) and SST and has significant implications for accurately modeling and understanding climate dynamics. This study aims to investigate the relationship between the response of UOC to the absence of FIREs and the resulting weaker TAU. Specifically, the study will analyze the impacts of changes in TAU, influenced by FIREs, on local UOC. To address this question, the study will examine the annual-mean spatial patterns of mean biases (MBs) and mean absolute biases (MABs) within the study domain (240°W – 0°, 40°S – 40°N), with a particular focus on the South Pacific trade-wind regions (160°W – 120°W, 30°S – 0°S). The analysis will explore the biases of UOC and their connection to differences in TAU between NOS and SON. We will not only examine the TAU bias and its potential connection to UOC and simulated SSTs, but also evaluate cross sections to examine vertical stratification for select regions based on their high correlation between TAU and ocean currents against NASA JPL Estimating the Circulation and Climate of the Ocean (ECCO) ocean reanalysis. Through an analysis of these three factors, the degree to which the absence of FIREs influences the responses of UOC, mediated by changes in TAU through FIREs, will be determined. In Section 2, a brief overview is provided about the data, methodology, and model simulations utilized in the study. Section 3 focuses on the simulation assessment and explores the impact of FIREs on ocean currents and SST performance. It also examines the relationship between FIREs, changes in SST, and TAU. Finally, Section 4 summarizes and discusses the major findings that were obtained from the study.