Browse

You are looking at 81 - 90 of 117,435 items for

  • All content x
Clear All
Robin Waldman, Joël Hirschi, Aurore Voldoire, Christophe Cassou, and Rym Msadek

Abstract

This work aims to clarify the relation between the Atlantic meridional overturning circulation (AMOC) and the thermal wind. We derive a new and generic dynamical AMOC decomposition that expresses the thermal wind transport as a simple vertical integral function of eastern minus western boundary densities. This allows us to express density anomalies at any depth as a geostrophic transport in Sverdrups (1 Sv ≡ 106 m3 s−1) per meter and to predict that density anomalies around the depth of maximum overturning induce most AMOC transport. We then apply this formalism to identify the dynamical drivers of the centennial AMOC variability in the CNRM-CM6 climate model. The dynamical reconstruction and specifically the thermal wind component explain over 80% of the low-frequency AMOC variance at all latitudes, which is therefore almost exclusively driven by density anomalies at both zonal boundaries. This transport variability is dominated by density anomalies between depths of 500 and 1500 m, in agreement with theoretical predictions. At those depths, southward-propagating western boundary temperature anomalies induce the centennial geostrophic AMOC transport variability in the North Atlantic. They are originated along the western boundary of the subpolar gyre through the Labrador Sea deep convection and the Davis Strait overflow.

Open access
Adam Vaccaro, Julien Emile-Geay, Dominque Guillot, Resherle Verna, Colin Morice, John Kennedy, and Bala Rajaratnam

Abstract

Surface temperature is a vital metric of Earth’s climate state, but is incompletely observed in both space and time: over half of monthly values are missing from the widely used HadCRUT4.6 global surface temperature dataset. Here we apply GraphEM, a recently developed imputation method, to construct a spatially complete estimate of HadCRUT4.6 temperatures. GraphEM leverages Gaussian Markov random fields (aka Gaussian graphical models) to better estimate covariance relationships within a climate field, detecting anisotropic features such as land/ocean contrasts, orography, ocean currents and wave-propagation pathways. This detection leads to improved estimates of missing values compared to methods (such as kriging) that assume isotropic covariance relationships, as we show with real and synthetic data.

This interpolated analysis of HadCRUT4.6 data is available as a 100-member ensemble, propagating information about sampling variability available from the original HadCRUT4.6 dataset. A comparison of NINO3.4 and global mean monthly temperature series with published datasets reveals similarities and differences due in part to the spatial interpolation method. Notably, the GraphEM-completed HadCRUT4.6 global temperature displays a stronger early twenty-first century warming trend than its uninterpolated counterpart, consistent with recent analyses using other datasets. Known events like the 1877/1878 El Niño are recovered with greater fidelity than with kriging, and result in different assessments of changes in ENSO variability through time. Gaussian Markov random fields provide a more geophysically-motivated way to impute missing values in climate fields, and the associated graph provides a powerful tool to analyze the structure of teleconnection patterns. We close with a discussion of wider applications of Markov random fields in climate science.

Restricted access
Laurence C. Breaker and Dustin Carroll

Abstract

The purpose of this study is to extract more information about the scaling exponents we obtain from sea surface temperature (SST) because their information content is limited to a single value. We examine the application of Empirical Mode Decomposition (EMD) to power law scaling using sea surface temperature (SST) from Scripps Pier, California. The daily observations we employ extend from 1920 to 2009, a period of 90 years. The annual cycle and the long-term trend were first removed. The decomposition produced a total of 15 modes. The scaling exponents were then calculated separately for each mode from the EMD decomposition. We have examined the distribution of scaling exponents with respect to the ensemble, and then with respect to the individual modes for the oceanic processes that we may infer from them. The first three modes are anti-persisitent and contain about a quarter of the total variance. The pattern of modes that was obtained is continuous and relatively smooth beyond mode 3 with increasing values up to mode 8 and generally decreasing values thereafter. The pattern exhibits intramodal correlation, as expected, and intermodal correlation as well. Intermodal correlation is likely due, for the most part, to Long-range Persistence (LRP). Finally, the annual cycle in SST at Scripps Pier is a dominant feature in the record and contains almost 70% of the variance. A method for removing the annual cycle that is not based on removing the mean value is introduced and recommended for future use.

Restricted access
Song Yang, Vincent Lao, Richard Bankert, Timothy R. Whitcomb, and Joshua Cossuth

Abstract

Accurate precipitation climatology is presented for tropical depression (TD), tropical storm (TS), and tropical cyclone (TC) over oceans using the recently-released, consistent and high quality precipitation datasets from all passive microwave sensors covering 1998-2012 along with the Automated Rotational Center Hurricane Eye Retrieval (ARCHER)-based TC center positions. Impacts with respect to the direction of both TC movement and the 200-850 hPa wind shear on the spatial distributions of TC precipitation are analyzed. The TC eyewall contraction process during its intensification is noted by a decrease in the radius of maximum rainrate with an increase in TC intensity. For global TCs, the maximum rainrate with respect to the direction of TC movement is located in the down-motion quadrants for TD, TS, and Cat 1-3 TCs, and in a concentric pattern for Cat 4-5 TCs. A consistent maximum TC precipitation with respect to the direction of the 200-850 hPa wind shear is shown in the down shear left quadrant (DSLQ). With respect to direction of TC movement, spatial patterns of TC precipitation vary with basins and show different features for weak and strong storms. The maximum rainrate is always located in DSLQ for all TC categories and basins, except the Southern Hemisphere basin where it is in the down shear right quadrant (DSRQ). This study not only confirms previously published results on TC precipitation distributions relative to vertical wind shear direction, but also provides a detailed distribution for each TC category and TS, while TD storms display an enhanced rainfall rate ahead of the down shear quadrants.

Restricted access
Audrey Delpech, Claire Ménesguen, Yves Morel, Leif N. Thomas, Frédéric Marin, Sophie Cravatte, and Sylvie Le Gentil

Abstract

At low latitudes in the ocean, the deep currents are shaped into narrow jets flowing eastward and westward, reversing periodically with latitude between 15°S and 15°N. These jets are present from the thermocline to the bottom. The energy sources and the physical mechanisms responsible for their formation are still debated and poorly understood. This study explores the role of the destabilization of intra-annual equatorial waves in the jets’ formation process, as these waves are known to be an important energy source at low latitudes. The study focuses particularly on the role of barotropic Rossby waves as a first step toward understanding the relevant physical mechanisms. It is shown from a set of idealized numerical simulations and analytical solutions that nonlinear triad interactions (NLTIs) play a crucial role in the transfer of energy toward jet-like structures (long waves with short meridional wavelengths) that induce a zonal residual mean circulation. The sensitivity of the instability emergence and the scale selection of the jet-like secondary wave to the forced primary wave are analyzed. For realistic amplitudes around 5–20 cm s−1, the primary waves that produce the most realistic jet-like structures are zonally propagating intra-annual waves with periods between 60 and 130 days and wavelengths between 200 and 300 km. The NLTI mechanism is a first step toward the generation of a permanent jet-structured circulation and is discussed in the context of turbulent cascade theories.

Restricted access
Daosheng Wang, Haidong Pan, Lin Mu, Xianqing Lv, Bing Yan, and Hua Yang

Abstract

The coastal ocean sea level (SL) variations result from multiscale processes and are dominated by SL changes due to meteorological forcing. In this study, a new methodology, which combines inverted barometer correction and regression analysis (IBR), is developed to estimate the coastal ocean response to meteorological forcing in shallow water. The response is taken as the combination of the static ocean response calculated using the inverted barometer formula and the dynamic ocean response estimated using the multivariable linear regression involving atmospheric pressure and the wind component in the dominant wind orientation. IBR was implemented to estimate the coastal ocean response at two stations, E1 and E2, in Bohai Bay, China. The analyzed results indicate that at both stations, the adjusted SLs are related more to the regional wind, which is the averaged value of ERA-Interim data in Bohai Bay, than to the local wind. The estimated response using IBR with the regional meteorological forcing is much closer to the observed values than other methods, including the classical inverted barometer correction, the dynamic atmospheric correction, the multivariable linear regression, and the IBR with local forcing. The deviations between the observed values and the estimated values using IBR with regional meteorological forcing can be primarily attributed to remote wind. This case study indicates that IBR is a feasible and relatively effective method to estimate the coastal ocean response to meteorological forcing in shallow water.

Restricted access
Mu Xiao, Sarith P. Mahanama, Yongkang Xue, Fei Chen, and Dennis P. Lettenmaier

Abstract

When compared with differences in snow accumulation predicted by widely used hydrological models, there is a much greater divergence among otherwise “good” models in their simulation of the snow ablation process. Here, we explore differences in the performance of the Variable Infiltration Capacity model (VIC), Noah land surface model with multiparameterization options (Noah-MP), the Catchment model, and the third-generation Simplified Simple Biosphere model (SiB3) in their ability to reproduce observed snow water equivalent (SWE) during the ablation season at 10 Snowpack Telemetry (SNOTEL) stations over 1992–2012. During the ablation period, net radiation generally has stronger correlations with observed melt rates than does air temperature. Average ablation rates tend to be higher (in both model predictions and observations) at stations with a large accumulation of SWE. The differences in the dates of last snow between models and observations range from several days to approximately a month (on average 5.1 days earlier than in observations). If the surface cover in the models is changed from observed vegetation to bare soil in all of the models, only the melt rate of the VIC model increases. The differences in responses of models to canopy removal are directly related to snowpack energy inputs, which are further affected by different algorithms for surface albedo and energy allocation across the models. We also find that the melt rates become higher in VIC and lower in Noah-MP if the shrub/grass present at the observation sites is switched to trees.

Restricted access
Simon R. Osborne and Graham P. Weedon

Abstract

A meteorological drought in 2018 led to senescence of the C3 grass at Cardington, Bedfordshire, United Kingdom. Observations of near-surface atmospheric variables and soil moisture are compared to simulations by the JULES land surface model (LSM) as used for Met Office forecasts. In years without drought, JULES provides better standalone simulations of evapotranspiration (ET) and soil moisture when the canopy height and rooting depth are reduced to match local conditions. During drought with the adjusted configuration, JULES correctly estimates total ET, but the components are in the wrong proportions. Several factors affect the estimation of ET including modeled skin temperatures, dewfall, and bare-soil evaporation. A diurnal range of skin temperatures close to observed is produced via the adjusted configuration and doubling the optical extinction coefficient. Although modeled ET during drought matches observed ET, this includes simulation of transpiration but in reality the grass was senescent. Excluding transpiration, the modeled bare-soil evaporation underestimates the observed midday latent heat flux. Part of the missing latent heat may relate to inappropriate parameterization of hydraulic properties of dry soils and part may be due to insufficient evaporation of dew. Dew meters indicate dewfall of up to 20 W m−2 during drought when the surface is cooling radiatively and turbulence is minimal. These data demonstrate that eddy-covariance techniques fail to reliably record the times, intensity, and variations in negative latent heat flux. Furthermore, the parameterization of atmospheric turbulence as used in LSMs fails to represent accurately dewfall during calm conditions when the surface is radiatively cooled.

Restricted access
Simon Veldkamp, Kirien Whan, Sjoerd Dirksen, and Maurice Schmeits

Abstract

Current statistical post-processing methods for probabilistic weather forecasting are not capable of using full spatial patterns from the numerical weather prediction (NWP) model. In this paper we incorporate spatial wind speed information by using convolutional neural networks (CNNs) and obtain probabilistic wind speed forecasts in the Netherlands for 48 hours ahead, based on KNMI’s deterministic Harmonie-Arome NWP model. The probabilistic forecasts from the CNNs are shown to have higher Brier skill scores for medium to higher wind speeds, as well as a better continuous ranked probability score (CRPS) and logarithmic score, than the forecasts from fully connected neural networks and quantile regression forests. As a secondary result, we have compared the CNNs using 3 different density estimation methods (quantized softmax (QS), kernel mixture networks, and fitting a truncated normal distribution), and found the probabilistic forecasts based on the QS method to be best.

Restricted access
Nicholas P. Klingaman, Matthew Young, Amulya Chevuturi, Bruno Guimaraes, Liang Guo, Steven J. Woolnough, Caio A. S. Coelho, Paulo Y. Kubota, and Christopher E. Holloway

Abstract

Skillful and reliable predictions of week-to-week rainfall variations in South America, two to three weeks ahead, are essential to protect lives, livelihoods, and ecosystems. We evaluate forecast performance for weekly rainfall in extended austral summer (November–March) in four contemporary subseasonal systems, including a new Brazilian model, at 1–5-week leads for 1999–2010. We measure performance by the correlation coefficient (in time) between predicted and observed rainfall; we measure skill by the Brier skill score for rainfall terciles against a climatological reference forecast. We assess unconditional performance (i.e., regardless of initial condition) and conditional performance based on the initial phase of the Madden–Julian oscillation (MJO) and El Niño–Southern Oscillation (ENSO). All models display substantial mean rainfall biases, including dry biases in Amazonia and wet biases near the Andes, which are established by week 1 and vary little thereafter. Unconditional performance extends to week 2 in all regions except for Amazonia and the Andes, but to week 3 only over northern, northeastern, and southeastern South America. Skill for upper- and lower-tercile rainfall extends only to week 1. Conditional performance is not systematically or significantly higher than unconditional performance; ENSO and MJO events provide limited “windows of opportunity” for improved S2S predictions that are region and model dependent. Conditional performance may be degraded by errors in predicted ENSO and MJO teleconnections to regional rainfall, even at short lead times.

Open access