Browse

You are looking at 91 - 100 of 9,660 items for :

  • Journal of Applied Meteorology and Climatology x
  • Refine by Access: All Content x
Clear All
Kyle A. Hilburn, Imme Ebert-Uphoff, and Steven D. Miller

Abstract

The objective of this research is to develop techniques for assimilating GOES-R series observations in precipitating scenes for the purpose of improving short-term convective-scale forecasts of high-impact weather hazards. Whereas one approach is radiance assimilation, the information content of GOES-R radiances from its Advanced Baseline Imager saturates in precipitating scenes, and radiance assimilation does not make use of lightning observations from the GOES Lightning Mapper. Here, a convolutional neural network (CNN) is developed to transform GOES-R radiances and lightning into synthetic radar reflectivity fields to make use of existing radar assimilation techniques. We find that the ability of CNNs to utilize spatial context is essential for this application and offers breakthrough improvement in skill compared to traditional pixel-by-pixel based approaches. To understand the improved performance, we use a novel analysis method that combines several techniques, each providing different insights into the network’s reasoning. Channel-withholding experiments and spatial information–withholding experiments are used to show that the CNN achieves skill at high reflectivity values from the information content in radiance gradients and the presence of lightning. The attribution method, layerwise relevance propagation, demonstrates that the CNN uses radiance and lightning information synergistically, where lightning helps the CNN focus on which neighboring locations are most important. Synthetic inputs are used to quantify the sensitivity to radiance gradients, showing that sharper gradients produce a stronger response in predicted reflectivity. Lightning observations are found to be uniquely valuable for their ability to pinpoint locations of strong radar echoes.

Open access
Jingjing Dou, Robert Bornstein, Shiguang Miao, Jianning Sun, and Yizhou Zhang

Abstract

The aim of this study was the analysis and simulation of the life cycle of a bifurcating thunderstorm that passed over Beijing, China, on 22 July 2015. Data from 150 surface weather sites and an S-band radar were used in conjunction with WRF simulations that used its multilevel Building Environment Parameterization (BEP) urbanization option. The Urban-case simulation used Beijing land-use information, and the NoUrban one replaced all urban areas by croplands. The Urban case correctly simulated both the observed weak 10-m winds over Beijing (<1.0 m s−1) and the weak 2-m urban heat island (<0.5°C). Observed radar and rain gauge data, as well as the Urban-case results, all showed precipitation bifurcation around Beijing, with maximum accumulations in convergent flow areas on either side of the city. The Urban case also reproduced the observed precipitation minima over the urban area and in a downwind rain shadow. The observations and Urban-case results both also showed bifurcated flow, even when the storm was still upwind of Beijing. The subsequent bifurcated precipitation areas thus each moved along a preexisting flow branch. Urban-case vertical sections showed downward motion in the divergence areas over the urban core and upward motions over the lateral convergence zones, both up to 6 km. Given that the NoUrban case showed none of these features, these differences demonstrate how the impact of cities can extend upward into deep local convection. Additional case-study simulations are needed to more fully understand urban storm bifurcation mechanisms in this and other storms for cities in a variety of climates.

Restricted access
Lea Hartl, Martin Stuefer, Tohru Saito, and Yoshitomi Okura

Abstract

We present the data records and station history of an automatic weather station (AWS) on Denali Pass (5715 m MSL), Alaska. The station was installed by a team of climbers from the Japanese Alpine Club after a fatal accident involving Japanese climbers in 1989 and was operational intermittently between 1990 and 2007, measuring primarily air temperature and wind speed. In later years, the AWS was operated by the International Arctic Research Center of the University of Alaska Fairbanks. Station history is reconstructed from available documentation as archived by the expedition teams. To extract and preserve data records, the original datalogger files were processed. We highlight numerous challenges and sources of uncertainty resulting from the location of the station and the circumstances of its operation. The data records exemplify the harsh meteorological conditions at the site: air temperatures down to approximately −60°C were recorded, and wind speeds reached values in excess of 60 m s−1. Measured temperatures correlate strongly with reanalysis data at the 500-hPa level. An approximation of critical wind speed thresholds and a reanalysis-based reconstruction of the meteorological conditions during the 1989 accident confirm that the climbers faced extremely hazardous wind speeds and very low temperatures. The data from the Denali Pass AWS represent a unique historical record that can, we hope, serve as a basis for further monitoring efforts in the summit region of Denali.

Open access
Dev Niyogi, Sajad Jamshidi, David Smith, and Olivia Kellner

Abstract

An intercomparison of multiresolution evapotranspiration (ET) datasets with reference to ground-based measurements for the development of regional reference (ETref) and actual (ETa) evapotranspiration maps over Indiana is presented. A representative ETref equation for the state is identified by evaluating 10 years of in situ measurements (2009–19). A statewide ETref climatology is developed using the ETref equation and high-resolution surface meteorological data from the gridded surface meteorological dataset (gridMET). For ETa analyses, MODIS, Simplified Surface Energy Balance Operational dataset (SSEBop), Global Land Evaporation Amsterdam Model (GLEAM) (versions 3.3a and 3.3b), and NLDAS (Noah and VIC) datasets are evaluated using AmeriFlux data. Thirty years of rainfall data from Climate Hazards Group Infrared Precipitation with Station Data Rainfall (CHIRPS) are used with the ET datasets to develop effective precipitation fields. Results show that the standardized Penman–Monteith equation performs as the best ETref equation with median symmetric accuracy (MSA) of 0.37, Taylor’s skill score (TSC) of 0.89, and r 2 = 0.83. The analysis shows that the gridMET dataset overestimates wind speed and requires adjustment before a series of statewide ETref climatology maps are generated (1990–2020). For ETa, the MODIS and GLEAM (3.3b) datasets outperform the rest, with MSA = 0.5, TSC = 0.8, and r 2 = 0.8. The state ETa dataset is generated using all MODIS data from 2003 and blending the MODIS data with GLEAM (3.3b) to cover data unavailability. Using the top-performing datasets, annual ETref for Indiana is computed as 1110 mm, ETa as 708 mm, and precipitation as 1091 mm. A marginal increasing climatological trend is found for Indiana’s ETref (0.013 mm yr−1) while ETa is found to be relatively stable. The state’s water availability, defined as rainfall minus ETa, has remained positive and stable at 0.99 mm day−1 (annual magnitude of +3820 mm).

Restricted access
Yingkai Sha, David John Gagne II, Gregory West, and Roland Stull

Abstract

Many statistical downscaling methods require observational inputs and expert knowledge and thus cannot be generalized well across different regions. Convolutional neural networks (CNNs) are deep-learning models that have generalization abilities for various applications. In this research, we modify UNet, a semantic-segmentation CNN, and apply it to the downscaling of daily maximum/minimum 2-m temperature (TMAX/TMIN) over the western continental United States from 0.25° to 4-km grid spacings. We select high-resolution (HR) elevation, low-resolution (LR) elevation, and LR TMAX/TMIN as inputs; train UNet using Parameter–Elevation Regressions on Independent Slopes Model (PRISM) data over the south- and central-western United States from 2015 to 2018; and test it independently over both the training domains and the northwestern United States from 2018 to 2019. We found that the original UNet cannot generate enough fine-grained spatial details when transferred to the new northwestern U.S. domain. In response, we modified the original UNet by assigning an extra HR elevation output branch/loss function and training the modified UNet to reproduce both the supervised HR TMAX/TMIN and the unsupervised HR elevation. This improvement is named “UNet-Autoencoder (AE).” UNet-AE supports semisupervised model fine-tuning for unseen domains and showed better gridpoint-level performance with more than 10% mean absolute error (MAE) reduction relative to the original UNet. On the basis of its performance relative to the 4-km PRISM, UNet-AE is a good option to provide generalizable downscaling for regions that are underrepresented by observations.

Open access
Yingkai Sha, David John Gagne II, Gregory West, and Roland Stull

Abstract

Statistical downscaling (SD) derives localized information from larger-scale numerical models. Convolutional neural networks (CNNs) have learning and generalization abilities that can enhance the downscaling of gridded data (Part I of this study experimented with 2-m temperature). In this research, we adapt a semantic-segmentation CNN, called UNet, to the downscaling of daily precipitation in western North America, from the low resolution (LR) of 0.25° to the high resolution (HR) of 4-km grid spacings. We select LR precipitation, HR precipitation climatology, and elevation as inputs; train UNet over the subset of the south- and central-western United States using Parameter–Elevation Regressions on Independent Slopes Model (PRISM) data from 2015 to 2018, and test it independently in all available domains from 2018 to 2019. We proposed an improved version of UNet, which we call Nest-UNet, by adding deep-layer aggregation and nested skip connections. Both the original UNet and Nest-UNet show generalization ability across different regions and outperform the SD baseline (bias-correction spatial disaggregation), with lower downscaling error and more accurate fine-grained textures. Nest-UNet also shares the highest amount of information with station observations and PRISM, indicating good ability to reduce the uncertainty of HR downscaling targets.

Open access
S. C. Pryor, F.W. Letson, and R. J. Barthelmie

Abstract

ERA5 provides high-resolution, high-quality hourly wind speeds at 100 m and is a unique resource for quantifying temporal variability in likely wind-derived power production across the United States. Gross capacity factors (CF) in seven independent system operators (ISOs) are estimated using the location and rated power of each wind turbine, a simplified power curve, and ERA5 output from 1979 to 2018. Excluding the California ISO, the marginal probability of a calm (zero power production) is less than 0.1 in any ERA5 grid cell. When a calm occurs, the mean co-occurrence across wind-turbine-containing grid cells ranges from 0.38 to 0.39 for ISOs in the Midwest and central plains [Midcontinent (or Midwest) ISO (MISO), Southwest Power Pool (SPP), and the Electric Reliability Council of Texas (ERCOT) region], increasing to 0.54–0.58 for ISOs in the eastern United States [Pennsylvania–New Jersey–Maryland interconnection (PJM), New York ISO (NYISO), and New England ISO (NEISO)]. Periods with low gross CF have a median duration of ≤6 h, except in California, and are most likely during summer. Gross CF exhibit highest variance at periods of 1 day in ERCOT and SPP; on synoptic scales in MISO, NEISO, and NYISO; and on interannual time scales in PJM. This implies differences in optimal strategies for ensuring resilience of supply. Theoretical scenarios show adding wind energy capacity near existing wind farms is advantageous even in areas with high existing installed capacity (IC), while expanding into areas with lower IC is more beneficial to reducing ramps and the probability of gross CF falling below 20%. These results emphasize the benefits of large balancing areas and aggregation in reducing wind power variability and the likelihood of wind droughts.

Restricted access
Niilo Siljamo, Otto Hyvärinen, Aku Riihelä, and Markku Suomalainen

Abstract

Snow cover plays a significant role in the weather and climate system by affecting the energy and mass transfer between the surface and the atmosphere. It also has far-reaching effects on ecosystems of snow-covered areas. Therefore, global snow-cover observations in a timely manner are needed. Satellite-based instruments can be utilized to produce snow-cover information that is suitable for these needs. Highly variable surface and snow-cover features suggest that operational snow extent algorithms may benefit from at least a partly empirical approach that is based on carefully analyzed training data. Here, a new two-phase snow-cover algorithm utilizing data from the Advanced Very High Resolution Radiometer (AVHRR) on board the MetOp satellites of the European Organization for the Exploitation of Meteorological Satellites (EUMETSAT) is introduced and evaluated. This algorithm is used to produce the MetOp/AVHRR H32 snow extent product for the Satellite Application Facility on Support to Operational Hydrology and Water Management (H SAF). The algorithm aims at direct detection of snow-covered and snow-free pixels without preceding cloud masking. Pixels that cannot be classified reliably to snow or snow-free, because of clouds or other reasons, are set as unclassified. This reduces the coverage but increases the accuracy of the algorithm. More than four years of snow-depth and state-of-the-ground observations from weather stations were used to validate the product. Validation results show that the algorithm produces high-quality snow coverage data that may be suitable for numerical weather prediction, hydrological modeling, and other applications.

Open access
Liang Wang, Dan Li, Ning Zhang, Jianning Sun, and Weidong Guo

Abstract

Urban heat islands (UHIs) are caused by a multitude of changes induced by urbanization. However, the relative importance of biophysical and atmospheric factors in controlling the UHI intensity remains elusive. In this study, we quantify the magnitude of surface UHIs (SUHIs), or surface urban cool islands (SUCIs), and elucidate their biophysical and atmospheric drivers on the basis of observational data collected from one urban site and two rural grassland sites in and near the city of Nanjing, China. Results show that during the daytime a strong SUCI effect is observed when the short grassland site is used as the reference site whereas a moderate SUHI effect is observed when the tall grassland is used as the reference site. We find that the former is mostly caused by the lower aerodynamic resistance for convective heat transfer at the urban site and the latter is primarily caused by the higher surface resistance for evapotranspiration at the urban site. At night, SUHIs are observed when either the short or the tall grassland site is used as the reference site and are predominantly caused by the stronger release of heat storage at the urban site. In general, the magnitude of SUHI is much weaker, and even becomes SUCI during daytime, with the short grassland site being the reference site because of its larger aerodynamic resistance. The study highlights that the magnitude of SUHIs and SUCIs is mostly controlled by urban–rural differences of biophysical factors, with urban–rural differences of atmospheric conditions playing a minor role.

Restricted access
Christina Kumler-Bonfanti, Jebb Stewart, David Hall, and Mark Govett

Abstract

Extracting valuable information from large sets of diverse meteorological data is a time-intensive process. Machine-learning methods can help to improve both speed and accuracy of this process. Specifically, deep-learning image-segmentation models using the U-Net structure perform faster and can identify areas that are missed by more restrictive approaches, such as expert hand-labeling and a priori heuristic methods. This paper discusses four different state-of-the-art U-Net models designed for detection of tropical and extratropical cyclone regions of interest (ROI) from two separate input sources: total precipitable water output from the Global Forecast System (GFS) model and water vapor radiance images from the Geostationary Operational Environmental Satellite (GOES). These models are referred to as International Best Track Archive for Climate Stewardship (IBTrACS)-GFS, Heuristic-GFS, IBTrACS-GOES, and Heuristic-GOES. All four U-Nets are fast information extraction tools and perform with an ROI detection accuracy ranging from 80% to 99%. These are additionally evaluated with the Dice and Tversky intersection-over-union (IoU) metrics, having Dice coefficient scores ranging from 0.51 to 0.76 and Tversky coefficients ranging from 0.56 to 0.74. The extratropical cyclone U-Net model performed 3 times as fast as the comparable heuristic model used to detect the same ROI. The U-Nets were specifically selected for their capabilities in detecting cyclone ROI beyond the scope of the training labels. These machine-learning models identified more ambiguous and active ROI missed by the heuristic model and hand-labeling methods that are commonly used in generating real-time weather alerts, having a potentially direct impact on public safety.

Restricted access