Gregory J. Hakim

Gregory J. Hakim, Sanjit Masanam

Global deep-learning weather prediction models have recently been shown to produce forecasts that rival those from physics-based models run at operational centers. It is unclear whether these models have encoded atmospheric dynamics, or simply pattern matching that produces the smallest forecast error. Answering this question is crucial to establishing the utility of these models as tools for basic science. Here we subject one such model, Pangu-weather, to a set of four classical dynamical experiments that do not resemble the model training data. Localized perturbations to the model output and the initial conditions are added to steady time-averaged conditions, to assess the propagation speed and structural evolution of signals away from the local source. Perturbing the model physics by adding a steady tropical heat source results in a classical Matsuno--Gill response near the heating, and planetary waves that radiate into the extratropics. A localized disturbance on the winter-averaged North Pacific jet stream produces realistic extratropical cyclones and fronts, including the spontaneous emergence of polar lows. Perturbing the 500hPa height field alone yields adjustment from a state of rest to one of wind--pressure balance over ~6 hours. Localized subtropical low pressure systems produce Atlantic hurricanes, provided the initial amplitude exceeds about 5 hPa, and setting the initial humidity to zero eliminates hurricane development. We conclude that the model encodes realistic physics in all experiments, and suggest it can be used as a tool for rapidly testing ideas before using expensive physics-based models.

Zilu Meng, Gregory J. Hakim, Wenchang Yang et al.

Deep learning (DL)-based general circulation models (GCMs) are emerging as fast simulators, yet their ability to replicate extreme events outside their training range remains unknown. Here, we evaluate two such models -- the hybrid Neural General Circulation Model (NGCM) and purely data-driven Deep Learning Earth System Model (DL\textit{ESy}M) -- against a conventional high-resolution land-atmosphere model (HiRAM) in simulating land heatwaves and coldwaves. All models are forced with observed sea surface temperatures and sea ice over 1900-2020, focusing on the out-of-sample early-20th-century period (1900-1960). Both DL models generalize successfully to unseen climate conditions, broadly reproducing the frequency and spatial patterns of heatwave and cold wave events during 1900-1960 with skill comparable to HiRAM. An exception is over portions of North Asia and North America, where all models perform poorly during 1940-1960. Due to excessive temperature autocorrelation, DL\textit{ESy}M tends to overestimate heatwave and cold wave frequencies, whereas the physics-DL hybrid NGCM exhibits persistence more similar to HiRAM.

P. Trent Vonich, Gregory J. Hakim

Atmospheric predictability research has long held that the limit of skillful deterministic weather forecasts is about 14 days. We challenge this limit using GraphCast, a machine-learning weather model, by optimizing forecast initial conditions using gradient-based techniques for twice-daily forecasts spanning 2020. This approach yields an average error reduction of 86% at 10 days, with skill lasting beyond 30 days. Mean optimal initial-condition perturbations reveal large-scale, spatially coherent corrections to ERA5, primarily reflecting an intensification of the Hadley circulation. Forecasts using GraphCast-optimal initial conditions in the Pangu-Weather model achieve a 21% error reduction, peaking at 4 days, indicating that analysis corrections reflect a combination of both model bias and a reduction in analysis error. These results demonstrate that, given accurate initial conditions, skillful deterministic forecasts are consistently achievable far beyond two weeks, challenging long-standing assumptions about the limits of atmospheric predictability.

Zilu Meng, Gregory J. Hakim, Wenchang Yang et al.

Atmospheric blocking events drive persistent weather extremes in midlatitudes, but isolating the influence of sea surface temperature (SST) from chaotic internal atmospheric variability on these events remains a challenge. We address this challenge using century-long (1900-2010), large-ensemble simulations with two computationally efficient deep-learning general circulation models. We find these models skillfully reproduce the observed blocking climatology, matching or exceeding the performance of a traditional high-resolution model and representative CMIP6 models. Averaging the large ensembles filters internal atmospheric noise to isolate the SST-forced component of blocking variability, yielding substantially higher correlations with reanalysis than for individual ensemble members. We identify robust teleconnections linking Greenland blocking frequency to North Atlantic SST and El Niño-like patterns. Furthermore, SST-forced trends in blocking frequency show a consistent decline in winter over Greenland, and an increase over Europe. These results demonstrate that SST variability exerts a significant and physically interpretable influence on blocking frequency and establishes large ensembles from deep learning models as a powerful tool for separating forced SST signals from internal noise.

Zilu Meng, Gregory J. Hakim

A deep learning (DL) model, based on a transformer architecture, is trained on a climate-model dataset and compared with a standard linear inverse model (LIM) in the tropical Pacific. We show that the DL model produces more accurate forecasts compared to the LIM when tested on a reanalysis dataset. We then assess the ability of an ensemble Kalman filter to reconstruct the monthly-averaged upper ocean from a noisy set of 24 sea-surface temperature observations designed to mimic existing coral proxy measurements, and compare results for the DL model and LIM. Due to signal damping in the DL model, we implement a novel inflation technique by adding noise from hindcast experiments. Results show that assimilating observations with the DL model yields better reconstructions than the LIM for observation averaging times ranging from one month to one year. The improved reconstruction is due to the enhanced predictive capabilities of the DL model, which map the memory of past observations to future assimilation times.