Search

Given the increasing use of climate projections and multi-model ensemble weighting for a diverse array of applications, this project assesses the sensitivities of climate model weighting strategies and their resulting ensemble means to multiple components, such as the weighting schemes, climate variables, or spatial domains of interest. The purpose of this study is to assess the sensitivities associated with multi-model weighting strategies. The analysis makes use of global climate models from the Coupled Model Intercomparison Project Phase 5 (CMIP5) and their statistically downscaled counterparts created with the localized constructed analogs (LOCA) method. This work focuses on historical and projected future mean precipitation and daily high temperatures of the south-central United States. Results suggest that the model weights and the corresponding weighted model means can be sensitive to the weighting strategy that is applied. For instance, when estimating model weights based on Louisiana precipitation, the weighted projections show a wetter and cooler south-central domain in the future compared to other weighting strategies. Alternatively, for example, when estimating model weights based on New Mexico temperature, the weighted projections show a drier and warmer south-central domain in the future. However, when considering the entire south-central domain in estimating the model weights, the weighted future projections show a compromise in the precipitation and temperature estimates. As for uncertainty, our matrix of results provided a more certain picture of future climate compared to the spread in the original model ensemble. If future impact assessments utilize weighting strategies, then our findings suggest that how the specific weighting strategy is used with climate projections may depend on the needs of an impact assessment or adaptation plan.

Observation-based climate model evaluation and future projections help policymakers in developing action plans for efficient management of water resources and mitigation of the impacts of hazardous extremes. Apart from this socioeconomic importance, the scientific value cannot be overstated, especially in light of the upcoming Fifth U.S. National Climate Assessment (NCA) report. In this study, we evaluate the realism of hydroclimate variability in the historical simulations of a suite of coupled general circulation models (CGCMs) participating in the fifth and sixth phases of the Coupled Model Intercomparison Project (CMIP5 and CMIP6). Our results demonstrate systematic biases in the simulated seasonal precipitation—most prominently, wet bias over the mountainous western United States in winter, and dry bias over the U.S. central plains in summer. A distinctive feature of this work is our focus on the examination of the atmospheric water budget, in particular the relative importance of remote and local contributions—convergence of moisture fluxes and local land surface processes (evapotranspiration) respectively—in helping produce precipitation. This diagnosis reveals that the leading contribution of the remote influence in winter is overestimated by the CMIP6 multimodel mean (MMM), whereas the local influence, which is more influential in summer, is underestimated. Our results aid in understanding the drivers of seasonal precipitation over the United States, where precipitation will likely increase by the end of the century but with significant model disagreement for the summer and fall. In support of ongoing NCA efforts, our study aims to contribute a comprehensive, regional-level analysis of the moisture budget and emphasizes the importance of realistically simulating its major components in CGCMs.

Recent extreme heat events, especially those that have occurred in the Western United States (WUS), are fueling wildfires and funneling smoke at an unprecedented level, impacting air/water quality and leading to an increase in respiratory hospitalizations. Under greenhouse warming, extreme weather conditions that favor wildfire ignition are expected to occur more frequently over the contiguous United States (CONUS). Therefore, predicting wildfire danger under a changing climate is essential in managing future wildfires and protecting the welfare of people and the environment. In response to mitigating wildfire risks, the Canadian Fire Weather Index (FWI) was developed to provide a numeric rating representing the intensity of a spreading fire. In this work, we utilized fine-scale (0.25° x 0.25°) daily meteorological inputs from thirty-five general circulation models in NASA Earth Exchange Global Daily Downscaled Projections Coupled Model Intercomparison Project Phase 6 (NEX-GDDP-CMIP6) data to calculate the FWI. Using the daily maximum temperature, relative humidity, wind speed, and precipitation from NEX-GDDP-CMIP6, we calculated the FWI of historical and future simulations from the periods of 1950 to 2100 under different emission scenarios (Shared Socioeconomic Pathways 2-4.5 and 5-8.5). We have analyzed the FWI for the GISS-E2-1-G model, which indicates a 2-3% increase per decade in future fire danger under both emission-pathway-driven climate scenarios during the dry season in the Southwestern US. We have found that the FWI climatology in the Southwestern US during the Summer presents high to extreme fire danger (> 50) and higher FWI values in the future compared to historical observations. Moreover, we have explored the uncertainties across multiple models using NEX-GDDP-CMIP6 statistically downscaled data and found a significant spread of the FWI across the models for historical observations and future simulations. To correlate the link between the FWI and actual fire occurrence, we will calculate the FWI using reanalysis data (MERRA-2) and validate the FWI with actual fire occurrence data from Global Fire Emissions Database (GFED) with a special emphasis on the WUS. While supporting the US NCA and NASA’s Climate Adaptation Service Investigator (CASI), we will also try to contribute FWI to NASA’s FireSense, an initiative to bring an Earth systems approach to improving wildfire and wildland fire management.

Given the increasing use of climate projections and multi-model ensemble weighting for a diverse array of applications, this project assesses the sensitivities of climate model weighting, and their resulting ensemble means, to multiple components, such as the weighting schemes, climate variables, or spatial domains of interest. The analysis makes use of global climate models from the Coupled Model Intercomparison Project Phase 5 (CMIP5), and their statistically downscaled counterparts created with the Localized Canonical Analogs (LOCA) method. This work focuses on historical and projected future mean precipitation and daily high temperatures of the south-central United States. Results suggest that model weights and corresponding weighted projections are highly sensitive to the weighting method as well as to the selected variables and spatial domains. For instance, when estimating model weights based on Louisiana precipitation, the weighted projections show a wetter and cooler south-central domain in the future compared to other weighting schemes. Alternatively, for example, when estimating model weights based on New Mexico temperature, the weighted projections show a drier and warmer south-central domain in the future. However, when considering the entire south-central domain in estimating the model weights, the weighted future projections show a compromise in the precipitation and temperature estimates. If future impact assessments utilize weighting schemes, then our findings suggest that how the weighting scheme is derived and applied to the projections may depend on the needs of an impact assessment or adaptation plan. From the results of our analysis, we summarize our recommendations concerning multi-model ensemble weighting as follows: Weighted ensemble means should be used not only for national and international assessments but also for regional impacts assessments and planning. Multiple strategies for model weighting are employed when feasible, to assure that uncertainties from various sources (e.g., weighting strategy used, domain or variable of interest applied, etc.) are considered. That weighting is derived for individual sub-regions (such as the NCA regions) in addition to what is derived for the continental United States. That domain-specific weighting be derived using both common (e.g. precipitation) and stakeholder-specific (e.g. streamflow) variables to produce relevant analysis for impact assessments and planning.

The current study investigates the diurnal cycle of dust and dust mixture loading across the global tropics, subtropics, and mid-latitudes by analyzing aerosol extinction and typing profiles observed by the Cloud-Aerosol Transport System (CATS) lidar aboard the International Space Station. According to the comparison with ground-based and other satellite observations, CATS aerosol and dust and dust mixture loading observations exhibit reasonable quality but significant day–night inconsistency. To account for this day–night inconsistency in CATS data quality, the diurnal variability in dust and dust mixture characteristics is currently examined separately for daytime and nighttime periods. Based on an analysis of variance (ANOVA) analytical framework, pronounced diurnal variations in dust and dust mixture loading are generally uncovered during daytime periods and over terrestrial areas. The current study identifies statistically significant diurnal variability in dust and dust mixture loading over key dust sources, including the Bodélé Depression, the West African El Djouf, Rub' al-Khali desert, and western and southern North America, confirming the previous observation-based findings regarding the diurnal cycle of dust emission and underlying meteorological processes in these regions. Significant seasonal and diurnal variability in dust and dust mixture is identified over the Iraqi and Thar deserts. The identified significant diurnal cycles in dust mixture loading over the vegetated regions in the Amazon and tropical southern Africa are hypothesized to be driven by enhanced dust emission due to wildfires.

This study utilizes Bayesian model averaging (BMA) as a framework to constrain the spread of uncertainty in climate projections of precipitation over the contiguous United States (CONUS). We use a subset of historical model simulations and future model projections (RCP8.5) from the Coupled Model Intercomparison Project phase 5 (CMIP5). We evaluate the representation of five precipitation summary metrics in the historical simulations using observations from the NASA Tropical Rainfall Measuring Mission (TRMM) satellites. The summary metrics include mean, annual and interannual variability, and maximum and minimum extremes of precipitation. The estimated model average produced with BMA is shown to have higher accuracy in simulating mean rainfall than the ensemble mean (RMSE of 0.49 for BMA versus 0.65 for ensemble mean), and a more constrained spread of uncertainty with roughly a third of the total uncertainty than is produced with the multimodel ensemble. The results show that, by the end of the century, the mean daily rainfall is projected to increase for most of the East Coast and the Northwest, may decrease in the southern United States, and with little change expected for the Southwest. For extremes, the wettest year on record is projected to become wetter for the majority of CONUS and the driest year to become drier. We show that BMA offers a framework to more accurately estimate and to constrain the spread of uncertainties of future climate, such as precipitation changes over CONUS.

This paper introduces a novel visual analytics approach, DCPViz, to enable climate scientists to explore massive climate data interactively without requiring the upfront movement of massive data. Thus, climate scientists are afforded more effective approaches to support the identification of potential trends and patterns in climate projections and their subsequent impacts. We designed the DCPViz pipeline to fetch and extract NEX-DCP30 data with minimal data transfer from their public sources. We implemented DCPViz to demonstrate its scalability and scientific value and to evaluate its utility under three use cases based on different models and through domain expert feedback.

A 30-yr climatology of lightning days and associated synoptic meteorological patterns are characterized across the interior western United States (WUS). Locally centered composite analyses show preferred synoptic meteorological patterns with positive 500-hPa geopotential height anomalies located to the northeast and negative sea level pressure anomalies to the northwest and collocated with local lightning days. Variations in preferred patterns for local lightning days are seen across the interior WUS. Areas not commonly affected by the North American monsoon system including the western Great Basin and northern Rocky Mountains show higher-amplitude anomalies of geopotential height, moisture, and midtropospheric instability patterns suggesting the importance of episodic midlatitude dynamics to lightning days in such locations. By contrast, locations closer to the core of the North American monsoon show weaker anomalies, likely reflecting the prevalence of favorable mesoscale dynamics key to lightning production during warm-season months in locations in the interior Southwest. Meteorological patterns for select locations are explored in more detail and two case studies of notably active lightning events are presented. Results from this observational analysis provide a foundation for evaluating meteorological conditions on lightning days in climate model simulations for the interior WUS.

Thermal infrared spectra can provide unique information on atmospheric ammonia enhancements, including those produced by agricultural, industrial, and biomass burning sources. In this work, we develop and test a quantitative ammonia retrieval algorithm using data from the airborne hyperspectral thermal emission spectrometer (HyTES) instrument over diverse targets, including persistent ammonia sources, such as cattle feedlots and a power plant, as well as a small, smoldering wildfire. Although no formal validation was possible due to lack of collocated observations, the HyTES ammonia values over feedlots in the California San Joaquin Valley are shown to be consistent within the estimated retrieval uncertainty of $\sim$ 20%–50% with in situ ammonia values collected over the same facilities during the 2013 DISCOVER-AQ field campaign. HyTES also observed meter-scale thermal hot spots associated with a small wildfire in Northern Arizona, which was determined to coincide with the emission of free ammonia, presumed to originate from the fire. This very localized emission source, not visible from satellite sensors, could be resolved in HyTES imagery. The ammonia enhancement immediately downwind from the smoldering fire, before the free ammonia reacted to form aerosols, was found to be distinguishable from the background with values three to ten times smaller than ammonia enhancements from the power plant and cattle feedlots, respectively. High-resolution detection and quantification of ammonia over nonpersistent sources like wildfires from HyTES shows that it is a powerful tool that can be used to improve regional scale fire emission inventories by accounting for the presence of temporally short ammonia processes.

Exposure to fine particulate matter (PM2.5) air pollution has been shown in numerous studies to be associated with detrimental health effects. However, the ability to conduct epidemiological assessments can be limited due to challenges in generating reliable PM2.5 estimates, particularly in parts of the world such as the Middle East where measurements are scarce and extreme meteorological events such as sandstorms are frequent. In order to supplement exposure modeling efforts under such conditions, satellite-retrieved aerosol optical depth (AOD) has proven to be useful due to its global coverage. By using AODs from the Multiangle Implementation of Atmospheric Correction (MAIAC) of the MODerate Resolution Imaging Spectroradiometer (MODIS) and the Multiangle Imaging Spectroradiometer (MISR) combined with meteorological and assimilated aerosol information from the Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2), we constructed machine learning models to predict PM2.5 in the area surrounding the Persian Gulf, including Kuwait, Bahrain, and the United Arab Emirates (U.A.E). Our models showed regional differences in predictive performance, with better results in the U.A.E. (median test R2 = 0.66) than Kuwait (median test R2 = 0.51). Variable importance also differed by region, where satellite-retrieved AOD variables were more important for predicting PM2.5 in Kuwait than in the U.A.E. Divergent trends in the temporal and spatial autocorrelations of PM2.5 and AOD in the two regions offered possible explanations for differences in predictive performance and variable importance. In a test of model transferability, we found that models trained in one region and applied to another did not predict PM2.5 well, even if the transferred model had better performance. Overall the results of our study suggest that models developed over large geographic areas could generate PM2.5 estimates with greater uncertainty than could be obtained by taking a regional modeling approach. Furthermore, development of methods to better incorporate spatial and temporal autocorrelations in machine learning models warrants further examination.

Given that persons with a prior history of respiratory diseases tend to demonstrate more severe illness from COVID-19 and, hence, are at higher risk of serious symptoms, ambient air quality data from NASA's satellite observations might provide a critical insight into which geographical areas may exhibit higher numbers of hospitalizations due to COVID-19, how the expected severity of COVID-19 and associated survival rates may vary across space in the future, and most importantly how given this information, health professionals can distribute vaccines in a more efficient, timely, and fair manner. Despite the utmost urgency of this problem, there yet exists no systematic analysis on linkages among COVID-19 clinical severity, air quality, and other atmospheric conditions, beyond relatively simplistic regression-based models. The goal of this project is to glean a deeper insight into sophisticated spatio-temporal dependencies among air quality, atmospheric conditions, and COVID-19 clinical severity using the machinery of Geometric Deep Learning (GDL), while providing quantitative uncertainty estimates. Our results based on the GDL model on a county level in three US states, California, Pennsylvania and Texas, indicate that AOD attributes to COVID-19 clinical severity in 39, 30, and 132 counties out of 58, 67, and 254 total counties, respectively. In turn, relative humidity is another important factor for understanding dynamics of clinical course and mortality risks due COVID-19, but predictive utility of temperature is noticeably lower. Our findings do not only contribute to understanding of latent factors behind COVID-19 progression but open new perspectives for innovative use of NASA's datasets for biosurveillance and social good.

The ability of the 5th phase of the Coupled Model Intercomparison Project (CMIP5) to reproduce twentieth-century climate trends over the seven CONUS regions of the National Climate Assessment is evaluated. This evaluation is carried out for summer and winter for three time periods, 1895–1939, 1940–1979, and 1980–2005. The evaluation includes all 206 CMIP5 historical simulations from 48 unique models and their multi-model ensemble (MME), as well as a gridded in situ dataset of surface air temperature and precipitation. Analysis is performed on both individual members and the MME, and considers reproducing the correct sign of the trends by the members as well as reproducing the trend values. While the MME exhibits some trend bias in most cases, it reproduces historical temperature trends with reasonable fidelity for summer for all time periods and all regions, including at the CONUS scale, except the Northern Great Plains from 1895 to 1939 and Southeast during 1980–2005. Likewise, for DJF, the MME reproduces historical temperature trends across all time periods over all regions, including at the CONUS scale, except the Southeast from 1895 to 1939 and the Midwest during 1940–1979. Model skill was highest across all of the seven regions during JJA and DJF for the 1980–2005 period. The quantitatively best result is seen during DJF in the Southwest region with at least 74% of the ensemble members correctly reproducing the observed trend across all of the time periods. No clear trends in MME precipitation were identified at these scales due to high model precipitation variability.

Climate model evaluation is complicated by the presence of observational uncertainty. In this study we analyze daily precipitation indices and compare multiple gridded observational and reanalysis products with regional climate models (RCMs) from the North American component of the Coordinated Regional Climate Downscaling Experiment (NA-CORDEX) multimodel ensemble. In the context of model evaluation, observational product differences across the contiguous United States (CONUS) are also deemed nontrivial for some indices, especially for annual counts of consecutive wet days and for heavy precipitation indices. Multidimensional scaling (MDS) is used to directly include this observational spread into the model evaluation procedure, enabling visualization and interpretation of model differences relative to a “cloud” of observational uncertainty. Applying MDS to the evaluation of NA-CORDEX RCMs reveals situations of added value from dynamical downscaling, situations of degraded performance from dynamical downscaling, and the sensitivity of model performance to model resolution. On precipitation days, higher-resolution RCMs typically simulate higher mean and extreme precipitation rates than their lower-resolution pairs, sometimes improving model fidelity with observations. These results document the model spread and biases in daily precipitation extremes across the full NA-CORDEX model ensemble. The often-large divergence between in situ observations, satellite data, and reanalysis, shown here for CONUS, is especially relevant for data-sparse regions of the globe where satellite and reanalysis products are extensively relied upon. This highlights the need to carefully consider multiple observational products when evaluating climate models.

The current study investigates the diurnal cycle of dust loading across the global tropics, sub-tropics, and mid-latitudes by analyzing aerosol extinction and typing profiles observed by the Cloud–Aerosol Transport System (CATS) lidar aboard the International Space Station. According to the comparison with ground-based and other satellite observations, CATS aerosol and dust loading observations exhibits reasonable quality but significant day–night inconsistency. To account for this day–night inconsistency in CATS data quality, the diurnal variability in dust characteristics are currently examined separately for daytime and nighttime periods. Based on an analysis of variance analytical framework, pronounced diurnal variations in dust loading are generally uncovered during daytime periods and over terrestrial areas. The current study identifies statistically significant diurnal variability in dust loading over key dust sources, including the Bodélé Depression, the West African El Djouf, Rub-al Khali Desert, and western and southern North America, confirming the previous observation-based findings regarding the diurnal cycle of dust emission and underlying meteorological processes in these regions. Significant seasonal dust diurnal variability is identified over the Iraqi and Thar deserts. The identified significant diurnal cycles in dust loading over the rainforests in Amazon and tropical southern Africa are hypothesized to be driven by enhanced dust emission due to wildfires.

A multivariate assessment of climate model projections over South America from the CMIP5 archive is presented. Change in near-surface temperature, precipitation, evapotranspiration, integrated water vapor transport (IVT), sea level pressure, and wind at multiple pressure levels is quantified across the multi-model suite and an assessment of model-to-model agreement on projected change performed. All models project warming by the mid- and late-21st century throughout the continent, with the highest magnitude projected over tropical regions. The CMIP5 models are in strong agreement that precipitation will decrease in all seasons over portions of Patagonia, especially along the northern portions of the current-climate mid-latitude storm track. This is consistent with a robustly projected poleward shift of the Pacific extratropical high and mid-latitude storm track indicated by a systematic increase in sea level pressure and decrease in westerly wind over Patagonia. Decreased precipitation for the months of September, October, and November is also projected, with strong model agreement, over portions of northern and northeastern Brazil, coincident with decreases in sea level pressure and increases in evapotranspiration. IVT is broadly projected to decrease over southern South America, coincident with the projected poleward shift of the mid-latitude storm track indicators, with increases projected in the vicinity of the South Atlantic Convergence Zone in austral spring and summer. Further decomposition of the thermodynamic and dynamic components to this change in IVT indicate that the projected decreases in the mid-latitudes are primarily driven by changes in circulation (i.e. dynamic) while the sub-tropical and tropical changes have a predominantly thermodynamic origin. Results provide a comprehensive picture of climate change across South America and highlight where projections should be interpreted with the most confidence.

The need to better understand climate change has driven model simulations to greater fidelity with improved spatiotemporal resolution (e.g., < 10 km at sub-hourly cadence). For example, the 7 km GE...

The current study investigates the diurnal cycle of dust loading across the global tropics, sub-tropics, and mid-latitudes by analyzing aerosol extinction and typing profiles observed by the Cloud-Aerosol Transport System (CATS) lidar aboard the International Space Station. According to the comparison with ground-based and other satellite observations, CATS aerosol and dust loading observations exhibits reasonable quality and insignificant day-night inconsistency, thereby supporting the current analysis of dust diurnal cycle using CATS data. Based on an analysis of variance analytical framework, statistically significant diurnal variability in dust loading is identified over key dust sources, including the Bodélé depression, West African El Djouf, Rub-al Khali desert, and western and southern North America, confirming the previous observation-based findings regarding the diurnal cycle of dust emission and underlying meteorological processes in these regions. Insignificant annual mean dust diurnal variability is identified over the Iraqi, Thar, and Taklamakan deserts. The currently identified significant diurnal cycles in dust loading over the rainforests in Amazon and tropical southern Africa, and drylands in South America and the central Australia, are hypothesized to be driven by enhanced dust emission due to wildfires and enhanced katabatic and frontal winds, respectively.