Search Results
You are looking at 1 - 6 of 6 items for
- Author or Editor: F. Roux x
- Refine by Access: All Content x
Abstract
No abstract available.
Abstract
No abstract available.
Abstract
One of the major problems in three-dimensional wind field analysis from dual (or multiple) Doppler radar data resides in the non-stationary of the observed air flow within the volume sampling time which ranges typically from 2 to 5 min. The present part II is focused on this problem. Most often, the storm moves horizontally at a speed of 5–25 m s−1.Therefore, the temporal variation for a fixed observer at ground level results from the superposition of two effects: 1) the intrinsic temporal variation (or variation seen in a frame moving with the storm) and 2) the effect of horizontal advection.
The first contribution of the paper concerns the development of an algorithm for correcting for the advection effect in the case of a dual-Doppler radar observation. This algorithm, which provides a mathematically exact solution to the problem of correcting for advection, can be very easily implemented in a computer program.
The second contribution deals with the errors that may arise from an accurate (or lack of) evaluation of the advective velocity, or from an “Intrinsic” temporal variation in the moving frame. A spectral decomposition of the 3D wind field is considered, allowing us to study the dependence of the error on the scale of the motion. Specific conclusions are drawn about the requirements necessary to achieve a given accuracy in the vertical velocity field. i.e., admissible uncertainty in the advective velocity, and characteristic time of intrinsic temporal variation.
Finally an example of application to actual Doppler radar data is presented. The results obtained from non-advected analyses are compared and discussed.
Abstract
One of the major problems in three-dimensional wind field analysis from dual (or multiple) Doppler radar data resides in the non-stationary of the observed air flow within the volume sampling time which ranges typically from 2 to 5 min. The present part II is focused on this problem. Most often, the storm moves horizontally at a speed of 5–25 m s−1.Therefore, the temporal variation for a fixed observer at ground level results from the superposition of two effects: 1) the intrinsic temporal variation (or variation seen in a frame moving with the storm) and 2) the effect of horizontal advection.
The first contribution of the paper concerns the development of an algorithm for correcting for the advection effect in the case of a dual-Doppler radar observation. This algorithm, which provides a mathematically exact solution to the problem of correcting for advection, can be very easily implemented in a computer program.
The second contribution deals with the errors that may arise from an accurate (or lack of) evaluation of the advective velocity, or from an “Intrinsic” temporal variation in the moving frame. A spectral decomposition of the 3D wind field is considered, allowing us to study the dependence of the error on the scale of the motion. Specific conclusions are drawn about the requirements necessary to achieve a given accuracy in the vertical velocity field. i.e., admissible uncertainty in the advective velocity, and characteristic time of intrinsic temporal variation.
Finally an example of application to actual Doppler radar data is presented. The results obtained from non-advected analyses are compared and discussed.
Abstract
Durirg the night of 23/24 June 1981, new Korhogo, Ivory Coast, a squall line passed over the instrumented area of the COPT 81 experiment. Observations were obtained with a dual-Doppler radar system, a sounding station and 22 automatic meteorological surface stations. Data from these instruments and from satellite pictures were analyzed to depict the kinematic and thermodynamic structure of the squall line. Composite analysis techniques were used to obtain a vertical cross section of the reflectivity structure and of the wind field relative to the line. The redistributions of air, moisture and thermodynamic energy by the convection wet calculated through averaged two-dimensional wind fields from a dual-Doppler radar system. The method also allowed the evaluation of the exchanges that were occurring between the convective and the stratiform regions.
This squall line had many similarities with tropical squall lines previously described by others. The leading convective part, composed of intense updrafts and downdrafts, and the trailing part, containing weak mesoscale updraft and downdraft, were separated by a reflectivity trough. A notable feature of this line was the presence of a leading anvil induced by intense easterly environmental winds in the upper troposphere. Observations of the evolution of the system at different scales indicated that the mesoalpha-scale (following the classification of Orlanski) and the mosobeta-scale patterns combined to allow the system to have optimum conditions for maximum strength and a maximum lifetime.
A rear-to-front flow was found at midlevels in the stratiform region. The flow sloped downward to the surface and took on the characteristics of a density current in the forward half of the squall lice. Entering the convective region, this flow was supplied with cold air by the convective downdrafts and played an important role in forcing upward the less dense monsoon flow entering at the leading edge.
Calculations of mass, moisture and energy transports showed the importance of the transfers between the convective and the stratiform regions. Particularly large quantities of condensate and energy were transferred from the convective region toward the anvils and made important contributions to the precipitation budget in the stratiform region, while large quantities of water vapor and latent heat energy were transferred from the stratiform region toward the convective region through the rear-to-front flow. Diabatic heating resulting from condensation in the convective region was also evaluated.
Abstract
Durirg the night of 23/24 June 1981, new Korhogo, Ivory Coast, a squall line passed over the instrumented area of the COPT 81 experiment. Observations were obtained with a dual-Doppler radar system, a sounding station and 22 automatic meteorological surface stations. Data from these instruments and from satellite pictures were analyzed to depict the kinematic and thermodynamic structure of the squall line. Composite analysis techniques were used to obtain a vertical cross section of the reflectivity structure and of the wind field relative to the line. The redistributions of air, moisture and thermodynamic energy by the convection wet calculated through averaged two-dimensional wind fields from a dual-Doppler radar system. The method also allowed the evaluation of the exchanges that were occurring between the convective and the stratiform regions.
This squall line had many similarities with tropical squall lines previously described by others. The leading convective part, composed of intense updrafts and downdrafts, and the trailing part, containing weak mesoscale updraft and downdraft, were separated by a reflectivity trough. A notable feature of this line was the presence of a leading anvil induced by intense easterly environmental winds in the upper troposphere. Observations of the evolution of the system at different scales indicated that the mesoalpha-scale (following the classification of Orlanski) and the mosobeta-scale patterns combined to allow the system to have optimum conditions for maximum strength and a maximum lifetime.
A rear-to-front flow was found at midlevels in the stratiform region. The flow sloped downward to the surface and took on the characteristics of a density current in the forward half of the squall lice. Entering the convective region, this flow was supplied with cold air by the convective downdrafts and played an important role in forcing upward the less dense monsoon flow entering at the leading edge.
Calculations of mass, moisture and energy transports showed the importance of the transfers between the convective and the stratiform regions. Particularly large quantities of condensate and energy were transferred from the convective region toward the anvils and made important contributions to the precipitation budget in the stratiform region, while large quantities of water vapor and latent heat energy were transferred from the stratiform region toward the convective region through the rear-to-front flow. Diabatic heating resulting from condensation in the convective region was also evaluated.
Abstract
Three different airborne Doppler radar sampling strategies were tested in Hurricane Gustav (1990) on 29 August 1990. The two new strategies were the fore-aft scanning technique (FAST) and airborne dual-platform Doppler sampling. FAST employs radar mans in cones pointing alternately fore and aft of the vertical plane that is perpendicular to the flight track. The airborne dual-platform sampling uses two Doppler radars, each aboard a separate aircraft. The Doppler radars scan strictly in the vertical plant normal to the flight track. The aircraft fly simultaneously along different, preferably perpendicular, tracks. The third strategy tested in Hurricane Gustav was single-platform sampling, which uses one Doppler radar on one aircraft that flies two consecutive, usually orthogonal, flight tracks. The antenna scans in the plane normal to the flight track. The third technique had been used previously in hurricanes and other disturbed weather.
The rms differences between the aircraft in situ winds and the Doppler winds derived near the aircraft by single-platform sampling, dual-platform sampling, and FAST are found to be 7.8, 5.1, and 2.5 m s−1, respectively. These results suggest that in hurricanes dual-platform flat-plane sampling and FAST both enable substantial improvements in the accuracy and temporal resolution of airborne Doppler wind fields over those obtained from single-platform, fiat-plane scanning. The FAST results should be applicable to dual-beam sampling, which began in 1991. The actual rms errors of Doppler winds far from the flight tracks, at levels well above flight level, and in highly sheared environments may be significantly higher than the above differences.
Abstract
Three different airborne Doppler radar sampling strategies were tested in Hurricane Gustav (1990) on 29 August 1990. The two new strategies were the fore-aft scanning technique (FAST) and airborne dual-platform Doppler sampling. FAST employs radar mans in cones pointing alternately fore and aft of the vertical plane that is perpendicular to the flight track. The airborne dual-platform sampling uses two Doppler radars, each aboard a separate aircraft. The Doppler radars scan strictly in the vertical plant normal to the flight track. The aircraft fly simultaneously along different, preferably perpendicular, tracks. The third strategy tested in Hurricane Gustav was single-platform sampling, which uses one Doppler radar on one aircraft that flies two consecutive, usually orthogonal, flight tracks. The antenna scans in the plane normal to the flight track. The third technique had been used previously in hurricanes and other disturbed weather.
The rms differences between the aircraft in situ winds and the Doppler winds derived near the aircraft by single-platform sampling, dual-platform sampling, and FAST are found to be 7.8, 5.1, and 2.5 m s−1, respectively. These results suggest that in hurricanes dual-platform flat-plane sampling and FAST both enable substantial improvements in the accuracy and temporal resolution of airborne Doppler wind fields over those obtained from single-platform, fiat-plane scanning. The FAST results should be applicable to dual-beam sampling, which began in 1991. The actual rms errors of Doppler winds far from the flight tracks, at levels well above flight level, and in highly sheared environments may be significantly higher than the above differences.
A real-time and automated multiple-Doppler analysis method for ground-based radar data, with an emphasis on observations conducted over complex terrain, is presented. It is the result of a joint effort of the radar groups of Centre National de Recherches Météorologiques and Laboratoire d'Aérologie with a view to converging toward a common optimized procedure to retrieve mass-conserved three-dimensional wind fields in the presence of complex topography. The multiple-Doppler synthesis and continuity adjustment technique initially proposed for airborne Doppler radar data, then extended to ground-based Doppler radars and nonflat orography, is combined with a variational approach aimed at improving the vertical velocity calculation over mountainous regions. This procedure was successfully applied in real time during the Mesoscale Alpine Programme Special Observing Period. The real-time processing and display of Doppler radar data were intended to assist nowcast and aircraft missions, and involved efforts of the United Sates, France, and Switzerland.
A real-time and automated multiple-Doppler analysis method for ground-based radar data, with an emphasis on observations conducted over complex terrain, is presented. It is the result of a joint effort of the radar groups of Centre National de Recherches Météorologiques and Laboratoire d'Aérologie with a view to converging toward a common optimized procedure to retrieve mass-conserved three-dimensional wind fields in the presence of complex topography. The multiple-Doppler synthesis and continuity adjustment technique initially proposed for airborne Doppler radar data, then extended to ground-based Doppler radars and nonflat orography, is combined with a variational approach aimed at improving the vertical velocity calculation over mountainous regions. This procedure was successfully applied in real time during the Mesoscale Alpine Programme Special Observing Period. The real-time processing and display of Doppler radar data were intended to assist nowcast and aircraft missions, and involved efforts of the United Sates, France, and Switzerland.
Abstract
Since 2007, meteorologists of the U.S. Army Test and Evaluation Command (ATEC) at Dugway Proving Ground (DPG), Utah, have relied on a mesoscale ensemble prediction system (EPS) known as the Ensemble Four-Dimensional Weather System (E-4DWX). This article describes E-4DWX and the innovative way in which it is calibrated, how it performs, why it was developed, and how meteorologists at DPG use it. E-4DWX has 30 operational members, each configured to produce forecasts of 48 h every 6 h on a 272-processor high performance computer (HPC) at DPG. The ensemble’s members differ from one another in initial-, lateral-, and lower-boundary conditions; in methods of data assimilation; and in physical parameterizations. The predictive core of all members is the Advanced Research core of the Weather Research and Forecasting (WRF) Model. Numerical predictions of the most useful near-surface variables are dynamically calibrated through algorithms that combine logistic regression and quantile regression, generating statistically realistic probabilistic depictions of the atmosphere’s future state at DPG’s observing sites. Army meteorologists view E-4DWX’s output via customized figures posted to a restricted website. Some of these figures summarize collective results—for example, through means, standard deviations, or fractions of the ensemble exceeding thresholds. Other figures show each forecast, individually or grouped—for example, through spaghetti diagrams and time series. This article presents examples of each type of figure.
Abstract
Since 2007, meteorologists of the U.S. Army Test and Evaluation Command (ATEC) at Dugway Proving Ground (DPG), Utah, have relied on a mesoscale ensemble prediction system (EPS) known as the Ensemble Four-Dimensional Weather System (E-4DWX). This article describes E-4DWX and the innovative way in which it is calibrated, how it performs, why it was developed, and how meteorologists at DPG use it. E-4DWX has 30 operational members, each configured to produce forecasts of 48 h every 6 h on a 272-processor high performance computer (HPC) at DPG. The ensemble’s members differ from one another in initial-, lateral-, and lower-boundary conditions; in methods of data assimilation; and in physical parameterizations. The predictive core of all members is the Advanced Research core of the Weather Research and Forecasting (WRF) Model. Numerical predictions of the most useful near-surface variables are dynamically calibrated through algorithms that combine logistic regression and quantile regression, generating statistically realistic probabilistic depictions of the atmosphere’s future state at DPG’s observing sites. Army meteorologists view E-4DWX’s output via customized figures posted to a restricted website. Some of these figures summarize collective results—for example, through means, standard deviations, or fractions of the ensemble exceeding thresholds. Other figures show each forecast, individually or grouped—for example, through spaghetti diagrams and time series. This article presents examples of each type of figure.
