In an assessment of 29 global climate models (GCMs), Part I of this study identified biases in boreal winter MJO teleconnections in anomalous 500-hPa geopotential height over the Pacific–North America (PNA) region that are common to many models: an eastward shift, a longer persistence, and a larger amplitude. In Part II, we explore the relationships of the teleconnection metrics developed in Part I with several existing and newly developed MJO and basic state (the mean subtropical westerly jet) metrics. The MJO and basic state diagnostics indicate that the MJO is generally weaker and less coherent and propagates faster in models compared to observations. The mean subtropical jet also exhibits notable biases such as too strong amplitude, excessive eastward extension, or southward shift. The following relationships are found to be robust among the models: 1) models with a faster MJO propagation tend to produce weaker teleconnections; 2) models with a less coherent eastward MJO propagation tend to simulate more persistent MJO teleconnections; 3) models with a stronger westerly jet produce stronger and eastward shifted MJO teleconnections; 4) models with an eastward extended jet produce an eastward shift in MJO teleconnections; and 5) models with a southward shifted jet produce stronger MJO teleconnections. The results are supported by linear baroclinic model experiments. Our results suggest that the larger amplitude and eastward shift biases in GCM MJO teleconnections can be attributed to the biases in the westerly jet, and that the longer persistence bias is likely due to the lack of coherent eastward MJO propagation.
Global Climate Models (GCMs) exhibit substantial biases in their simulation of tropical climate. One particularly problematic bias exists in GCMs' simulation of the tropical rainband known as the Intertropical Convergence Zone (ITCZ). Much of the precipitation on Earth falls within the ITCZ, which plays a key role in setting Earth's temperature by affecting global energy transports, and partially dictates dynamics of the largest interannual mode of climate variability: The El Niño‐Southern Oscillation (ENSO). Most GCMs fail to simulate the mean state of the ITCZ correctly, often exhibiting a “double ITCZ bias,” with rainbands both north and south rather than just north of the equator. These tropical mean state biases limit confidence in climate models' simulation of projected future and paleoclimate states, and reduce the utility of these models for understanding present climate dynamics. Adjusting GCM parameterizations of cloud processes and atmospheric convection can reduce tropical biases, as can artificially correcting sea surface temperatures through modifications to air‐sea fluxes (i.e., “flux adjustment”). Here, we argue that a significant portion of these rainfall and circulation biases are rooted in orographic height being biased low due to assumptions made in fitting observed orography onto GCM grids. We demonstrate that making different, and physically defensible, assumptions that raise the orographic height significantly improves model simulation of climatological features such as the ITCZ and North American rainfall as well as the simulation of ENSO. These findings suggest a simple, physically based, and computationally inexpensive method that can improve climate models and projections of future climate.
more » « less- NSF-PAR ID:
- 10360166
- Publisher / Repository:
- DOI PREFIX: 10.1029
- Date Published:
- Journal Name:
- AGU Advances
- Volume:
- 2
- Issue:
- 2
- ISSN:
- 2576-604X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
-
more » « less
Abstract The influence of coupled model sea surface temperature (SST) climatological biases and SST projections on daily convection over the Intra‐American Seas (IAS) during the May–November rainy season are examined by clustering (
k − means) daily OLR anomalies in ECHAM5 atmospheric global climate model (AGCM) experiments. The AGCM is first forced by 1980–2005 observed SSTs (GOGA), then with climatological, multi‐model mean monthly climatological SST bias from 31 CMIP5 coupled models (HIST) and projected SST changes for 2040–2059 (PROJ) and 2080–2099 (PROJ2) imposed on top of observed values. A typology of seven recurrent convection regimes is identified and consists of three dry and four wet regimes, including three regimes characterized by tropical‐midlatitude interactions between surface convection cells across the IAS and Rossby wave in the upper‐troposphere, and a regime of broad wettening typical of the ITCZ. Compared to an earlier observational study, all seven regimes are reasonably well reproduced in the HIST runs. However, the latter exhibit drier dry regimes, a less wet ITCZ‐like wet regime and a southeastward shift of convective anomalies developing across the IAS in the three other regimes, all result in a drier simulated IAS climate compared to GOGA. ECHAM5 projection runs for PROJ and PROJ2 are both characterized by the inhibition of the broad ITCZ‐like wet regime, indicating a significant trend towards more frequent dry weather. Meanwhile, convection anomalies related to tropical‐midlatitude interactions are shifted further east of the Caribbean as lead increases. These results suggest more frequent and intense extreme rainfall over the tropical Atlantic and the southeast US, while parts of the Caribbean are projected to experience drier climate. The projected drying, however, is of the same order of magnitude as results from historical SST biases, suggesting that the latter need to be considered in model projections, which might underestimate future IAS drying. -
more » « less
Abstract With the recent advances in data science, machine learning has been increasingly applied to convection and cloud parameterizations in global climate models (GCMs). This study extends the work of Han et al. (2020,
https://doi.org/10.1029/2020MS002076 ) and uses an ensemble of 32‐layer deep convolutional residual neural networks, referred to as ResCu‐en, to emulate convection and cloud processes simulated by a superparameterized GCM, SPCAM. ResCu‐en predicts GCM grid‐scale temperature and moisture tendencies, and cloud liquid and ice water contents from moist physics processes. The surface rainfall is derived from the column‐integrated moisture tendency. The prediction uncertainty inherent in deep learning algorithms in emulating the moist physics is reduced by ensemble averaging. Results in 1‐year independent offline validation show that ResCu‐en has high prediction accuracy for all output variables, both in the current climate and in a warmer climate with +4K sea surface temperature. The analysis of different neural net configurations shows that the success to generalize in a warmer climate is attributed to convective memory and the 1‐dimensional convolution layers incorporated into ResCu‐en. We further implement a member of ResCu‐en into CAM5 with real world geography and run the neural‐network‐enabled CAM5 (NCAM) for 5 years without encountering any numerical integration instability. The simulation generally captures the global distribution of the mean precipitation, with a better simulation of precipitation intensity and diurnal cycle. However, there are large biases in temperature and moisture in high latitudes. These results highlight the importance of convective memory and demonstrate the potential for machine learning to enhance climate modeling. -
null (Ed.)Abstract Climate models in phase 5 of the Coupled Model Intercomparison Project (CMIP5) vary significantly in their ability to simulate the phase and amplitude of atmospheric stationary waves in the midlatitude Southern Hemisphere. These models also suffer from a double intertropical convergence zone (ITCZ), with excessive precipitation in the tropical eastern South Pacific, and many also suffer from a biased simulation of the dynamics of the Agulhas Current around the tip of South Africa. The intermodel spread in the strength and phasing of SH midlatitude stationary waves in the CMIP archive is shown to be significantly correlated with the double-ITCZ bias and biases in the Agulhas Return Current. An idealized general circulation model (GCM) is used to demonstrate the causality of these links by prescribing an oceanic heat flux out of the tropical east Pacific and near the Agulhas Current. A warm bias in tropical east Pacific SSTs associated with an erroneous double ITCZ leads to a biased representation of midlatitude stationary waves in the austral hemisphere, capturing the response evident in CMIP models. Similarly, an overly diffuse sea surface temperature gradient associated with a weak Agulhas Return Current leads to an equatorward shift of the Southern Hemisphere jet by more than 3° and weak stationary wave activity in the austral hemisphere. Hence, rectification of the double-ITCZ bias and a better representation of the Agulhas Current should be expected to lead to an improved model representation of the austral hemisphere.more » « less
-
more » « less
Abstract With widespread influence on global climate and weather extremes, the Madden-Julian Oscillation (MJO) plays a crucial role in subseasonal prediction. Our latest global climate models (GCMs), however, have great difficulty in realistically simulating the MJO. This model inability is largely due to problems in representation of MJO’s cumulus organization. This study, based on a series of idealized aqua-planet model experiments using an atmospheric-only GCM, clearly demonstrates that MJO propagation is strongly modulated by the large-scale background state in which the lower-tropospheric mean moisture gradient and zonal winds are critical. Therefore, when tuning climate models to achieve improved MJO simulations, particular attention needs to be placed on the model large-scale mean state that is also significantly affected by cumulus parameterizations. This study indicates that model biases in representing MJO propagation may be related to the widely reported double-ITCZ (intertropical convergence zone) problem in climate models.
