Search for: All records

Abstract The western Pacific warm pool (WPWP) is the heat engine of the global climate system delivering vast amounts of heat and moisture to the atmosphere. Controls on regional convection, however, are numerous, making it difficult to simulate past and future changes in WPWP hydroclimate with confidence. Here, we synthesize new and previously available precipitation sensitive records from the WPWP spanning the last and present interglacial periods. We find two primary modes of rainfall variability, both driven by precession forcing, that are common to both interglacial periods: (a) a contraction of the tropical rain band across the interglacial and (b) a mid‐interglacial strengthening of the Pacific Walker Circulation (PWC). We further demonstrate that while the amplitude of the change in seasonal insolation across the Holocene is far lower than during the LIG due to the low eccentricity state of Earth's orbit, the response of regional rainfall is comparable during both interglacials, indicating a nonlinear response to the insolation forcing. Finally, we suggest an enhanced sensitivity of the PWC to non‐insolation climate forcing, including greenhouse gases and sea level change, under strongly reduced boreal fall insolation as observed during the late Holocene and late LIG. 

ABSTRACT This is the second in a series of papers in which we use JWST Mid Infrared Instrument multiband imaging to measure the warm dust emission in a sample of 31 multiply imaged quasars, to be used as a probe of the particle nature of dark matter. We present measurements of the relative magnifications of the strongly lensed warm dust emission in a sample of nine systems. The warm dust region is compact and sensitive to perturbations by populations of haloes down to masses $$\sim 10^6$$ M$$_{\odot }$$. Using these warm dust flux-ratio measurements in combination with five previous narrow-line flux-ratio measurements, we constrain the halo mass function. In our model, we allow for complex deflector macromodels with flexible third- and fourth-order multipole deviations from ellipticity, and we introduce an improved model of the tidal evolution of subhaloes. We constrain a WDM model and find an upper limit on the half-mode mass of $$10^{7.6}\, {\rm M}_\odot$$ at posterior odds of 10:1. This corresponds to a lower limit on a thermally produced dark matter particle mass of 6.1 keV. This is the strongest gravitational lensing constraint to date, and comparable to those from independent probes such as the Ly $$\alpha$$ forest and Milky Way satellite galaxies. 

Oxygen isotopes (δ18O) are the most commonly utilized speleothem proxy and have provided many foundational records of paleoclimate. Thus, understanding processes affecting speleothem δ18O is crucial. Yet, prior calcite precipitation (PCP), a process driven by local hydrology, is a widely ignored control of speleothem δ18O. Here we investigate the effects of PCP on a stalagmite δ18O record from central Vietnam, spanning 45 – 4 ka. We employ a geochemical model that utilizes speleothem Mg/Ca and cave monitoring data to correct the δ18O record for PCP effects. The resulting record exhibits improved agreement with regional speleothem δ18O records and climate model simulations, suggesting that the corrected record more accurately reflects precipitation δ18O (δ18Op). Without considering PCP, our interpretations of the δ18O record would have been misleading. To avoid misinterpretations of speleothem δ18O, our results emphasize the necessity of considering PCP as a significant driver of speleothem δ18O. 

Computational fluid dynamics (CFD) and its uncertainty quantification are computationally expensive. We use Gaussian Process (GP) methods to demonstrate that machine learning can build efficient and accurate surrogate models to replace CFD simulations with significantly reduced computational cost without compromising the physical accuracy. We also demonstrate that both epistemic uncertainty (machine learning model uncertainty) and aleatory uncertainty (randomness in the inputs of CFD) can be accommodated when the machine learning model is used to reveal fluid dynamics. The demonstration is performed by applying simulation of Hagen-Poiseuille and Womersley flows that involve spatial and spatial-tempo responses, respectively. Training points are generated by using the analytical solutions with evenly discretized spatial or spatial-temporal variables. Then GP surrogate models are built using supervised machine learning regression. The error of the GP model is quantified by the estimated epistemic uncertainty. The results are compared with those from GPU-accelerated volumetric lattice Boltzmann simulations. The results indicate that surrogate models can produce accurate fluid dynamics (without CFD simulations) with quantified uncertainty when both epistemic and aleatory uncertainties exist.