We analyze the large-scale statistically meaningful patterns (LSMPs), also called large-scale meteorological patterns, that precede extreme precipitation (PEx) events over Northern California (NorCal). We find LSMPs by applying k-means clustering to the two leading principal components of daily 500hPa geopotential height anomalies persisting two days before the onset. A statistical significance test based on the Monte Carlo simulations suggests the existence of a minimum of four statistically distinguished LSMP clusters. The four LSMP clusters are characterized as the NW continental negative height anomaly, the Eastward positive “PNA”, the Westward negative “PNA”, and the Prominent Alaskan ridge. These four clusters, shown in multiple atmospheric and oceanic variables, evolve very differently and have distant links to the Arctic and tropical Pacific regions. Using binary forecast skill measures and a new copula-based framework for predicting PEx events, we show that the LSMP indices are useful predictors of NorCal PEx events, with the moisture-based variables being the best predictors of PEx events at least six days before the onset, and the lower atmospheric variables being better than their upper atmospheric counterparts any day in advance.

We propose the first unified objective framework (SyCLoPS) for detecting and classifying all types of low-pressure systems (LPSs) in a given dataset. We use the state-of-the-art automated feature tracking software TempestExtremes (TE) to detect and track LPS features globally in ERA5 and compute 16 parameters from commonly-found atmospheric variables for classification. A Python classifier is implemented to classify all LPSs at once. The framework assigns 16 different labels (classes) to each LPS data point (node) and designates four different types of high-impact LPS tracks, including tropical cyclone (TC) tracks, Monsoon System (MS) tracks, subtropical tropical-like cyclone (STLC) tracks, and polar low (PL) tracks. The classification process involves disentangling high-altitude and drier LPSs, differentiating tropical and non-tropical LPSs using novel criteria, and optimizing for the detection of the four types of high-impact LPS. We compare our labels to those in the International Best Track Archive for Climate Stewardship (IBTrACS) and find that they are in good agreement. TC detection using SyCLoPS produces better tropical cyclone detection skill compared to the previous algorithms. Finally, we demonstrate that the output of SyCLoPS is valuable for investigating various aspects of LPSs, such as the evolution of a single LPS track, patterns and trends in LPS activities, and precipitation or wind influence associated with a particular LPS class.

A document by Shiheng Duan. Click on the document to view its contents.

The new generation of heterogeneous CPU/GPU computer systems offer much greater computational performance but are not yet widely used for climate modeling. One reason for this is that traditional climate models were written before GPUs were available and would require an extensive overhaul to run on these new machines. In addition, even conventional “high–resolution’ simulations don’t provide enough parallel work to keep GPUs busy, so the benefits of such overhaul would be limited for the types of simulations climate scientists are accustomed to. The vision of the Simple Cloud-Resolving Energy Exascale Earth System (E3SM) Atmosphere Model (SCREAM) project is to create a global atmospheric model with the architecture to efficiently use GPUs and horizontal resolution sufficient to fully take advantage of GPU parallelism. After 5 yrs of model development, SCREAM is finally ready for use. In this paper, we describe the design of this new code, its performance on both CPU and heterogeneous machines, and its ability to simulate real-world climate via a set of four 40 day simulations covering all 4 seasons of the year.

The 1997 New Year’s flood event was the most costly in California’s history. This compound extreme event was driven by a category 5 atmospheric river that led to widespread snowmelt. Extreme precipitation, snowmelt, and saturated soils produced heavy runoff causing widespread inundation in the Sacramento Valley. This study recreates the 1997 flood using the Regionally Refined Mesh capabilities of the Energy Exascale Earth System Model (RRM-E3SM) under prescribed ocean conditions. Understanding the processes causing extreme events inform practical efforts to anticipate and prepare for such events in the future, and also provides a rich context to evaluate model skill in representing extremes. Three California-focused RRM grids, with horizontal resolution refinement of 14km down to 3.5km, and six forecast lead times, 28 December 1996 at 00Z through 30 December 1996 at 12Z, are assessed for their ability to recreate the 1997 flood. Planetary to synoptic scale atmospheric circulations and integrated vapor transport are weakly influenced by horizontal resolution refinement over California. Topography and mesoscale circulations, such as the Sierra barrier jet, are prominently influenced by horizontal resolution. The finest resolution RRM-E3SM simulation best represents storm total precipitation and storm duration snowpack changes. Traditional time-series and causal analysis frameworks are used to examine runoff sensitivities state-wide and above major reservoirs. These frameworks show that horizontal resolution plays a more prominent role in shaping reservoir inflows, namely the magnitude and time-series shape, than forecast lead time, 2-to-4 days prior to the 1997 flood onset.

It has been proposed that conservation laws might not be beneficial for accurate hydrological modeling due to errors in input (precipitation) and target (streamflow) data (particularly at the event time scale), and this might explain why deep learning models (which are not based on enforcing closure) can out-perform catchment-scale conceptual and process-based models at predicting streamflow. We test this hypothesis at the event and multi-year time scale using physics-informed (mass conserving) machine learning and find that: (1) enforcing closure in the rainfall-runoff mass balance does appear to harm the overall skill of hydrological models, (2) deep learning models learn to account for spatiotemporally variable biases in data (3) however this “closure” effect accounts for only a small fraction of the difference in predictive skill between deep learning and conceptual models.

The water cycle is an important component of the earth system and it plays a key role in many facets of society, including energy production, agriculture, and human health and safety. In this study, the Energy Exascale Earth System Model version 1 (E3SMv1) is run with low-resolution (roughly 110 km) and high-resolution (roughly 25 km) configurations — as established by the High Resolution Model Intercomparison Project protocol — to evaluate the atmospheric and terrestrial water budgets over the conterminous United States (CONUS) at the large watershed scale. The water cycle slows down in the HR experiment relative to the LR, with decreasing fluxes of precipitation, evapotranspiration, atmospheric moisture convergence, and runoff. The reductions in these terms exacerbate biases for some watersheds, while reducing them in others. For example, precipitation biases are exacerbated at HR over the Eastern and Central CONUS watersheds, while precipitation biases are reduced at HR over the Western CONUS watersheds. The most pronounced changes to the water cycle come from reductions in precipitation and evapotranspiration, the latter of which results from decreases in evaporative fraction. While the HR simulation is warmer than the LR, moisture convergence decreases despite the increased atmospheric water vapor, suggesting circulation biases are an important factor. Additional exploratory metrics show improvements to water cycle extremes (both in precipitation and streamflow), fractional contributions of different storm types to total precipitation, and mountain snowpack.

In this paper the meteorological drivers of North American Monsoon (NAM) extreme precipitation events (EPEs) are identified and analyzed. First, the NAM area and its subregions are distinguished using self-organizing maps (SOM) applied to the Climate Prediction Center (CPC) global precipitation dataset. This delineation emphasizes the distinct extreme precipitation character and drivers in each subregion, and we subsequently argue these subregions are more suitable for regional analysis given the inhomogeneous geographical features in the NAM area. For each EPE, defined as daily precipitation exceeding the 95th precipitation percentile, five synoptic features and one mesoscale feature are investigated and assigned as potential drivers. Essentially all EPEs can be associated with at least one selected driver, with only one event remaining as unclassified. The attribution result demonstrates the dominant role of Gulf of California moisture surges, followed by mesoscale convective systems. Finally, a frequency and probability analysis is conducted to contrast precipitation distributions conditioned on the associated meteorological drivers. Interactions and influences among candidate features are revealed by the precipitation probability density functions.

This paper describes the first implementation of the d x=3.25 km version of the Energy Exascale Earth System Model (E3SM) global atmosphere model and its behavior in a 40 day prescribed-sea-surface-temperature simulation (Jan 20-Feb 28, 2020). This simulation was performed as part of the DYnamics of the Atmospheric general circulation Modeled On Non-hydrostatic Domains (DYAMOND) phase 2 model intercomparison. Effective resolution is found to be $\sim 6x the horizontal grid resolution despite using a coarser grid for physical parameterizations. Despite this new model being in an immature and untuned state, moving to 3.25 km grid spacing solves several long-standing problems with the E3SM model. In particular, Amazon precipitation is much more realistic, the frequency of light and heavy precipitation is improved, agreement between the simulated and observed diurnal cycle of tropical precipitation is excellent, and the vertical structure of tropical convection and coastal stratocumulus look good. In addition, the new model is able to capture the frequency and structure of important weather events (e.g. hurricanes, midlatitude storms including atmospheric rivers, and cold air outbreaks). Interestingly, this model does not get rid of the erroneous southern branch of the intertropical convergence zone nor the tendency for strongest convection to occur over the Maritime Continent rather than the West Pacific, both of which are classic climate model biases. Several other problems with the simulation are identified, underscoring the fact that this model is a work in progress.

The Atmospheric River (AR) Tracking Method Intercomparison Project (ARTMIP) is a community effort to systematically assess how the uncertainties from AR detectors (ARDTs) impact our scientific understanding of ARs. This study describes the ARTMIP Tier 2 experimental design and initial results using the Coupled Model Intercomparison Project (CMIP) Phases 5 and 6 multi-model ensembles. We show that AR statistics from a given ARDT in CMIP5/6 historical simulations compare remarkably well with the MERRA-2 reanalysis. In CMIP5/6 future simulations, most ARDTs project a global increase in AR frequency, counts, and sizes, especially along the western coastlines of the Pacific and Atlantic oceans. We find that the choice of ARDT is the dominant contributor to the uncertainty in projected AR frequency when compared with model choice. These results imply that new projects investigating future changes in ARs should explicitly consider ARDT uncertainty as a core part of the experimental design.

As the most severe drought over the Northeastern United States (NEUS) in the past century, the 1960s drought had pronounced socioeconomic and natural impacts. Although it was followed by a persisting wet period, the conditions leading to the 1960s extreme drought could return in the future, along with its challenges to water management. To project the characteristics and potential consequences of such a future drought, pseudo-global warming simulations using the Weather Research and Forecasting Model are performed to simulate the dynamical conditions of the historical 1960s drought, but with modified thermodynamic conditions under the RCP8.5 scenario in the early (2021-2027), middle (2041-2047) and late (2091-2097) 21st century. Our analysis focuses on essential hydroclimatic variables including temperature, precipitation, evapotranspiration, soil moisture, snowpack and surface runoff. In contrast to the historical 1960s drought, similar dynamical conditions will generally produce more precipitation, increased soil moisture and evapotranspiration, and reduced snowpack. However, we also find that although wet months get much wetter, dry months may become drier, meaning that wetting trends are most significant in wet months but are essentially negligible for extremely dry months with negative monthly mean net precipitation. For these months, the trend towards wetting conditions provides little relief from the effects of extreme dry months. These conditions may even aggravate water shortages due to an increasingly rapid transition from wet to dry conditions. Other challenges emerge for residents and stakeholders in this region, including more extreme hot days, record-low snow pack, frozen ground degradation and subsequent decreases in surface runoff.

In this paper, we introduce a testbed for evaluating and comparing climate modeling systems at cloud resolving scales using hindcasts of the June 2012 North American derecho. The testbed is applied to two models: the regionally-refined Simple Cloud-Resolving E3SM Atmosphere Model (SCREAM) at horizontal resolutions ranging from 6.5 to 1.625 km and the Weather Research and Forecasting (WRF) model with 4 km grid spacing. We find the simulation results to be highly sensitive to the initial conditions, initialization time, and model configurations, with initial conditions from the Rapid Refresh (RAP) producing the best simulation. Significant improvement is identified in the SCREAM simulations as horizontal grid spacing is refined. While a propagation delay of approximately 2 hours is found in both models, SCREAM at 1.625 km simulates the observed bow echo structure of the derecho well and predicts strong surface gusts that exceed 30 m/s. In comparison, WRF hardly produces surface wind over 25 m/s, and the derecho wind gust in WRF is 42-46% lower than in SCREAM. Moreover, WRF has a lower bias in simulating cold clouds but overestimates the precipitation intensity. Both models well reproduce the observed outgoing longwave radiation spatial patterns (Pearson correlation > 0.88) while they simulate larger areas of composite radar reflectivity > 40 dBZ by up to 4 times and underestimate the precipitating area by ~ 70\% in WRF and 47\% in SCREAM compared to observations.