Search for: All records

Abstract A neural network (NN) surrogate of the NASA GISS ModelE atmosphere (version E3) is trained on a perturbed parameter ensemble (PPE) spanning 45 physics parameters and 36 outputs. The NN is leveraged in a Markov Chain Monte Carlo (MCMC) Bayesian parameter inference framework to generate a secondposteriorconstrained ensemble coined a “calibrated physics ensemble,” or CPE. The CPE members are characterized by diverse parameter combinations and are, by definition, close to top‐of‐atmosphere radiative balance, and must broadly agree with numerous hydrologic, energy cycle and radiative forcing metrics simultaneously. Global observations of numerous cloud, environment, and radiation properties (provided by global satellite products) are crucial for CPE generation. The inference framework explicitly accounts for discrepancies (or biases) in satellite products during CPE generation. We demonstrate that product discrepancies strongly impact calibration of important model parameter settings (e.g., convective plume entrainment rates; fall speed for cloud ice). Structural improvements new to E3 are retained across CPE members (e.g., stratocumulus simulation). Notably, the framework improved the simulation of shallow cumulus and Amazon rainfall while not degrading radiation fields, an upgrade that neither default parameters nor Latin Hypercube parameter searching achieved. Analyses of the initial PPE suggested several parameters were unimportant for output variation. However, many “unimportant” parameters were needed for CPE generation, a result that brings to the forefront how parameter importance should be determined in PPEs. From the CPE, two diverse 45‐dimensional parameter configurations are retained to generate radiatively‐balanced, auto‐tuned atmospheres that were used in two E3 submissions to CMIP6. 

Sea surface temperature (SST) gradients are a primary driver of low‐level wind convergence in the east Pacific Inter‐Tropical Convergence Zone (ITCZ) through their hydrostatic relationship to the surface pressure gradient force (PGF). However, the surface PGF may not always align with SST gradients due to variations in boundary layer temperature gradients with height, that is, the boundary layer contribution to the surface PGF. In this study, we investigate the observed northern hemisphere position of the east Pacific ITCZ using a slab boundary layer model (SBLM) driven by different approximations of the boundary layer virtual temperature field. SBLM simulations using the entire boundary layer virtual temperature profile produce a realistic northern hemisphere ITCZ. However, SST‐only simulations produce excessive equatorial divergence and southern hemisphere convergence, resulting in a latitudinally confined double ITCZ‐like structure. Observed virtual temperature gradients highlight the importance of northward temperature gradients strengthening with height from the equator to 15°S below the trade wind inversion (TWI). Our interpretation is that the equatorial cold tongue induces relatively weak high surface pressure and double ITCZ‐like convergence because the resulting layer of cold air is shallow. Concurrently, relatively strong high surface pressure spreads out in the southern hemisphere due to interactions between stratocumulus clouds and the ocean surface. Together, the equatorial cold tongue and the TWI/stratocumulus clouds enable a more northern hemisphere dominant ITCZ. Thus, we provide evidence of a dynamical link between the equatorial cold tongue, low clouds, and double ITCZs, which continue to be problematic in Earth system models.