Search Results

You are looking at 21 - 30 of 62 items for

  • Author or Editor: Kenneth Howard x
  • Refine by Access: All Content x
Clear All Modify Search
Clinton E. Wallace
,
Robert A. Maddox
, and
Kenneth W. Howard

Abstract

The daily evolution of local surface conditions at Phoenix, Arizona, and the characteristics of the 1200 UTC sounding at Tucson, Arizona, have been examined to determine important meteorological features that lead to thunderstorm occurrence over the low deserts of central Arizona. Each day of July and August during the period 1990–95 has been stratified based upon daily mean, surface moisture conditions at Phoenix, Arizona, and the occurrence of afternoon and evening convective activity in the Phoenix metropolitan area. The nearest operational sounding, taken 160 km to the southeast at Tucson, is shown to be not representative of low-level thermodynamic conditions in central Arizona. Thus, Phoenix forecasters’ ability to identify precursor conditions for the development of thunderstorms is impaired. On days that convective storms occur in the Phoenix area, there is a decrease in the diurnal amplitude of surface dewpoint changes, signifying increased/deeper boundary layer moisture. This signal is very subtle and may not have much forecast utility. Additionally, it is found that surges of moist air from the Gulf of California do not occur frequently during the 36–48 h immediately prior to thunderstorm events in the Phoenix area. It is shown that the 1200 UTC Tucson wind profile has a significant northerly flow in low levels on moist days when storms do not occur in the Phoenix area. The forecaster needs information on the local temperature and moisture profile to assess the potential for thunderstorms in the Phoenix area. However, routine upper-air observations are unavailable. Steps are being taken to obtain morning soundings in Phoenix, and the improving capabilities of satellite-derived thermodynamic data and mesoscale models may also provide the forecaster critical information in the future. The findings, although specifically developed for the Phoenix area, may be relevant to thunderstorm forecasting in many regions of the interior West.

Full access
Steven V. Vasiloff
,
Kenneth W. Howard
, and
Jian Zhang

Abstract

The principal source of information for operational flash flood monitoring and warning issuance is weather radar–based quantitative estimates of precipitation. Rain gauges are considered truth for the purposes of validating and calibrating real-time radar-derived precipitation data, both in a real-time sense and climatologically. This paper examines various uncertainties and challenges involved with using radar and rain gauge data in a severe local storm environment. A series of severe thunderstorm systems that occurred across northeastern Montana illustrates various problems with comparing radar precipitation estimates and real-time gauge data, including extreme wind effects, hail, missing gauge data, and radar quality control. Ten radar–gauge time series pairs were analyzed with most found to be not useful for real-time radar calibration. These issues must be carefully considered within the context of ongoing efforts to develop robust real-time tools for evaluating radar–gauge uncertainties. Recommendations are made for radar and gauge data quality control efforts that would benefit the operational use of gauge data.

Full access
Kenneth R. Knapp
,
Michael C. Kruk
,
David H. Levinson
,
Howard J. Diamond
, and
Charles J. Neumann

The goal of the International Best Track Archive for Climate Stewardship (IBTrACS) project is to collect the historical tropical cyclone best-track data from all available Regional Specialized Meteorological Centers (RSMCs) and other agencies, combine the disparate datasets into one product, and disseminate in formats used by the tropical cyclone community. Each RSMC forecasts and monitors storms for a specific region and annually archives best-track data, which consist of information on a storm's position, intensity, and other related parameters. IBTrACS is a new dataset based on the best-track data from numerous sources. Moreover, rather than preferentially selecting one track and intensity for each storm, the mean position, the original intensities from the agencies, and summary statistics are provided. This article discusses the dataset construction, explores the tropical cyclone climatology from IBTrACS, and concludes with an analysis of uncertainty in the tropical cyclone intensity record.

Full access
Jonathan J. Gourley
,
Robert A. Maddox
,
Kenneth W. Howard
, and
Donald W. Burgess

Abstract

Implementation of the National Weather Service Weather Surveillance Radar-1988 Doppler (WSR-88D) radar network provides the potential to monitor rainfall and snowfall accumulations at fine spatial and temporal resolutions. An automated, operational algorithm called the Precipitation Processing System (PPS) uses reflectivity data to estimate precipitation accumulations. The utility of these estimates has yet to be quantified in the Intermountain West during winter months. The accuracy of precipitation estimates from the operational PPS during cool-season, stratiform-precipitation events in Arizona is examined. In addition, a method, with the potential for automation, is developed to improve estimates of precipitation by calibrating infrared data (10.7-μm band) from Geostationary Operational Environmental Satellite-9 using reflectivity-derived rainfall rates from WSR-88D radar. The “multisensor” approach provides more accurate estimates of rainfall across lower elevations during cool-season extratropical storms. After the melting layer has been manually identified using volumetric radar reflectivity data, reflectivity measured in or above it is discarded. Melting-layer heights also indicate the altitude of the rain–snow line. This information is used to delineate and map frozen versus liquid precipitation types. Rain gauges are used as an independent, ground-based source to assess the magnitude of improvements made over PPS rainfall products. Although the technique is designed and evaluated over a limited area in Arizona, it may be applicable to many mountainous regions.

Full access
Andrew P. Osborne
,
Jian Zhang
,
Micheal J. Simpson
,
Kenneth W. Howard
, and
Stephen B. Cocks

Abstract

The Multi-Radar Multi-Sensor (MRMS) system produces a suite of hydrometeorological products that are widely used for applications such as flash flood warning operations, water resource management, and climatological studies. The MRMS radar-based quantitative precipitation estimation (QPE) products have greater challenges in the western United States compared to the eastern two-thirds of the CONUS due to terrain-related blockages and gaps in radar coverage. Further, orographic enhancement of precipitation often occurs, which is highly variable in space and time and difficult to accurately capture with physically based approaches. A deep learning approach was applied in this study to understand the correlations between several interacting variables and to obtain a more accurate precipitation estimation in these scenarios. The model presented here is a convolutional neural network (CNN), which uses spatial information from small grids of several radar variables to predict an estimated precipitation value at the central grid point. Several case analyses are presented along with a yearlong statistical evaluation. The CNN model 24-h QPE shows higher accuracy than the MRMS radar QPE for several cool season atmospheric river events. Areas of consistent improvement from the CNN model are highlighted in the discussion along with areas where the model can be further improved. The initial findings from this work help set the foundation for further exploration of machine learning techniques and products for precipitation estimation as part of the MRMS operational system.

Significance Statement

This study explores the development and use of a deep learning model to generate precipitation fields in the complex terrain of the western United States. Generally, the model is able to improve on the statistical performance of existing radar-based precipitation estimation methods for several case studies and over a long-term period in 2021. We explore the patterns associated with certain areas of strong performance and suggest potential means of improving areas with weaker performance. These initial results indicate the potential of deep learning to supplement radar-based approaches in areas with observational limitations.

Open access
David J. Stensrud
,
Michael H. Jain
,
Kenneth W. Howard
, and
Robert A. Maddox

Abstract

A brief field project was conducted during July 1988 to assess the potential for Next Generation Weather Radar (NEXRAD), 404-MHz radar wind profilers, and digital sounding systems to monitor the low-level wind field during clear-air conditions. The low-level jet was chosen as the phenomenon of interest because it is neither well sampled nor resolved by the current upper-air network, yet it is a common feature of mesoscale convective system and severe thunderstorm environments. Data were collected under quiescent synoptic conditions during several low-level jet events using a 10-cm NEXRAD-like Doppler radar and a digital sounding system colocated in Norman, Oklahoma. These data suggest that the areal-averaged horizontal winds calculated from the Doppler radar data using the Velocity Azimuth Display (VAD) technique are comparable with the winds observed using a digital sounding system, except under weak wind conditions. However, the vertical spacing of 304 m (1000 ft) between levels of horizontal VAD calculated winds, as currently proposed for NEXRAD, may not be of sufficient resolution to document the detailed wind structure of these events. The height of the maximum wind speed of the low-level jet on all days studied was below the planned lowest observation range gate of the 404-MHz radar wind profiler, indicating that a combination of NEXRAD and profiler data might be needed to sample the important wind field structure of the lower atmosphere. Lastly, the National Weather Service rawinsonde data processing software affects the vertical resolution of the low-level wind field in operational, and therefore archived, upper-air soundings. The procedure used to calculate NWS 1000 ft winds actually damps the wind speed profile and artificially increases the height of the level of maximum wind speed associated with the low-level jet. The appropriateness of these highly smoothed 1000 ft winds for input into sophisticated mesoscale weather prediction models should be considered.

Full access
Andrew J. Negri
,
Robert F. Adler
,
Robert A. Maddox
,
Kenneth W. Howard
, and
Peter R. Keehn

Abstract

A three-year climatology of satellite-estimated rainfall for the warm season for the southwest United States and Mexico has been derived from data from the Special Sensor Microwave Imager (SSM/1). The microwave data have been stratified by month (June, July, August), yew (1988, 1989, 1990), and time of day (morning and evening orbits). A rain algorithm was employed that relates 86-GHz brightness temperatures to rain rate using a coupled cloud-radiative transfer model.

Results identify an early evening maximum in rainfall along the western slope of the Sierra Madre Occidental during all three months. A prominent morning rainfall maximum was found off the western Mexican coast near Mazatlan in July and August. Substantial differences between morning and evening estimates were noted. To the extent that three years constitute a climatology, results of interannual variability are presented. Results are compared and contrasted to high-resolution (8 km, hourly) infrared cloud climatologies, which consist of the frequency of occurrence of cloud colder than −38°C and −58°C. This comparison has broad implications for the estimation of rainfall by simple (cloud threshold) techniques.

By sampling the infrared data to approximate the time and space resolution of the microwave, we produce ratios (or adjustment factors) by which we can adjust the infrared rain estimation schemes. This produces a combined micro wave/infrared rain algorithm for monthly rainfall. Using a limited set of raingage data as ground truth, an improvement (lower bias and root-mean-square error) was demonstrated by this combined technique when compared to either method alone. The diurnal variability of convection during July 1990 was examined using hourly rain estimates from the GOES precipitation index and the convective stratiform technique, revealing a maximum in estimated rainfall from 1800 to 2100 local time. It is in this time period when the SSM/1 evening orbit occurs. A high-resolution topographic database was available to aid in interpreting the influence of topography on the rainfall patterns.

Full access
David J. Stensrud
,
Robert L. Gall
,
Steven L. Mullen
, and
Kenneth W. Howard

Abstract

The Mexican monsoon is a significant feature in the climate of the southwestern United States and Mexico during the summer months. Rainfall in northwestern Mexico during the months of July through September accounts for 60% to 80% of the total annual rainfall, while rainfall in Arizona for these same months accounts for over 40% of the total annual rainfall. Deep convection during the monsoon season produces frequent damaging surface winds, flash flooding, and hail and is a difficult forecast problem. Past numerical simulations frequently have been unable to reproduce the widespread, heavy rains over Mexico and the southwestern United States associated with the monsoon.

The Pennsylvania State University/National Center for Atmospheric Research mesoscale model is used to simulate 32 successive 24-h periods during the monsoon season. Mean fields produced by the model simulations are compared against observations to validate the ability of the model to reproduce many of the observed features, including the large-scale midtropospheric wind field, southerly low-level winds over the Gulf of California, and the heavy rains over western Mexico. Preliminary analysis of the mean model fields also suggest that the Gulf of California is the dominant moisture source for deep convection over Mexico and the southwestern United States, with upslope flow along the Sierra Madre Occidental advecting low-level gulf moisture into western Mexico during the daytime and southerly flow at the northern end of the gulf advecting gulf moisture into Arizona on most days. These results illustrate the usefulness of four-dimensional data assimilation techniques to create proxy datasets containing realistic mesoscale features that can be used for detailed diagnostic studies.

Full access
Stephen B. Cocks
,
Jian Zhang
,
Steven M. Martinaitis
,
Youcun Qi
,
Brian Kaney
, and
Kenneth Howard

Abstract

Multi-Radar Multi-Sensor (MRMS) quantitative precipitation estimation (QPE) radar only (Q3RAD), Q3RAD local gauge corrected (Q3gc), dual polarization (Dual Pol), legacy Precipitation Processing System (PPS), and National Centers for Environmental Prediction (NCEP) stage IV product performance were evaluated for data collected east of the Rockies during the 2014 warm season. For over 22 000 radar QPE–gauge data pairs, Q3RAD had a higher correlation coefficient (0.85) and a lower mean absolute error (9.4 mm) than the Dual Pol (0.83 and 10.5 mm, respectively) and PPS (0.79 and 10.8 mm, respectively). Q3RAD performed best when the radar beam sampled precipitation within or above the melting layer because of its use of a reflectivity mosaic corrected for brightband contamination. Both NCEP stage IV and Q3gc showed improvement over the radar-only QPEs; while stage IV exhibited the lower errors, the performance of Q3gc was remarkable considering the estimates were automatically generated in near–real time. Regional analysis indicated Q3RAD outperformed Dual Pol and PPS over the southern plains, Southeast/mid-Atlantic, and Northeast. Over the northern United States, Q3RAD had a higher wet bias below the melting layer than both Dual Pol and PPS; within and above the melting layer, Q3RAD exhibited the lowest errors. The Q3RAD wet bias was likely due to MRMS’s overestimation of tropical rain areas in continental regions and applying a high yield reflectivity–rain-rate relationship. An adjustment based on precipitation climatology reduced the wet bias errors by ~22% and will be implemented in the operational MRMS in the fall of 2016.

Full access
Steven M. Martinaitis
,
Heather M. Grams
,
Carrie Langston
,
Jian Zhang
, and
Kenneth Howard

Abstract

Precipitation values estimated by radar are assumed to be the amount of precipitation that occurred at the surface, yet this notion is inaccurate. Numerous atmospheric and microphysical processes can alter the precipitation rate between the radar beam elevation and the surface. One such process is evaporation. This study determines the applicability of integrating an evaporation correction scheme for real-time radar-derived mosaicked precipitation rates to reduce quantitative precipitation estimate (QPE) overestimation and to reduce the coverage of false surface precipitation. An evaporation technique previously developed for large-scale numerical modeling is applied to Multi-Radar Multi-Sensor (MRMS) precipitation rates through the use of 2D and 3D numerical weather prediction (NWP) atmospheric parameters as well as basic radar properties. Hourly accumulated QPE with evaporation adjustment compared against gauge observations saw an average reduction of the overestimation bias by 57%–76% for rain events and 42%–49% for primarily snow events. The removal of false surface precipitation also reduced the number of hourly gauge observations that were considered as “false zero” observations by 52.1% for rain and 38.2% for snow. Optimum computational efficiency was achieved through the use of simplified equations and hourly 10-km horizontal resolution NWP data. The run time for the evaporation correction algorithm is 6–7 s.

Full access