Search Results
You are looking at 1 - 9 of 9 items for
- Author or Editor: B. E. Sheppard x
- Refine by Access: All Content x
Abstract
The Precipitation Occurrence Sensor System (POSS) is a small Doppler radar originally designed by the Meteorological Service of Canada (MSC) to report the occurrence, type, and intensity of precipitation in automated observing stations. It is also used for real-time estimation of raindrop size distributions (DSDs). From the DSD, various rainfall parameters can be calculated and relationships established, such as between the radar reflectivity factor (Z) and the rainfall rate (R). Earlier work presented first-order estimates of the sampling errors for some POSS rainfall parameter estimates. This work combines a Monte Carlo simulation and “inverse problem” analysis to better estimate errors due to the specific sampling problems of this disdrometer type. The uncertainties are necessary to determine the statistical significance of differences between DSD estimates by the POSS and other collocated disdrometers, or between POSS measurements in different climatologies. Additionally, confidence limits can be assigned to regression coefficients for rainfall parameter relationships determined from POSS estimates. An example is given of the uncertainties in the coefficients of measured Z–R relationships.
Abstract
The Precipitation Occurrence Sensor System (POSS) is a small Doppler radar originally designed by the Meteorological Service of Canada (MSC) to report the occurrence, type, and intensity of precipitation in automated observing stations. It is also used for real-time estimation of raindrop size distributions (DSDs). From the DSD, various rainfall parameters can be calculated and relationships established, such as between the radar reflectivity factor (Z) and the rainfall rate (R). Earlier work presented first-order estimates of the sampling errors for some POSS rainfall parameter estimates. This work combines a Monte Carlo simulation and “inverse problem” analysis to better estimate errors due to the specific sampling problems of this disdrometer type. The uncertainties are necessary to determine the statistical significance of differences between DSD estimates by the POSS and other collocated disdrometers, or between POSS measurements in different climatologies. Additionally, confidence limits can be assigned to regression coefficients for rainfall parameter relationships determined from POSS estimates. An example is given of the uncertainties in the coefficients of measured Z–R relationships.
Abstract
A discrete-ordinate radiative transfer algorithm for a multilayered plane-parallel atmosphere was used to evaluate the effect of raindrops on ground-based vertically pointing radiometric measurements in the 20-90- GHz range. The effect of rainfall rate and drop size distribution, the length of the rain column, the scattering phase function, and cloud liquid water content are examined. Errors in the estimation of columnar integrated vapor and liquid water are evaluated as functions of rainfall rate, length of rain column, and integrated cloud liquid water. Brightness temperatures calculated by the model, using a measured atmospheric state and measured raindrop size distributions, agreed with radiometric measurements at 20, 31, and 52 GHz with a standard error of estimate of 4.1, 7.4, and 6.8 K, respectively. Variances of model brightness temperatures calculated using 1-min-average raindrop size distributions over 10-min periods were also correlated with measured brightness temperature variances.
Abstract
A discrete-ordinate radiative transfer algorithm for a multilayered plane-parallel atmosphere was used to evaluate the effect of raindrops on ground-based vertically pointing radiometric measurements in the 20-90- GHz range. The effect of rainfall rate and drop size distribution, the length of the rain column, the scattering phase function, and cloud liquid water content are examined. Errors in the estimation of columnar integrated vapor and liquid water are evaluated as functions of rainfall rate, length of rain column, and integrated cloud liquid water. Brightness temperatures calculated by the model, using a measured atmospheric state and measured raindrop size distributions, agreed with radiometric measurements at 20, 31, and 52 GHz with a standard error of estimate of 4.1, 7.4, and 6.8 K, respectively. Variances of model brightness temperatures calculated using 1-min-average raindrop size distributions over 10-min periods were also correlated with measured brightness temperature variances.
Abstract
Precipitation detection and typing estimates from four sensors are evaluated using standard operational meteorological observations as a reference. All are active remote sensors radiating into a measurement volume near the sensor, and measuring some property of the scattered radiation. Three of the sensors use optical wavelengths and one uses microwave wavelengths.
A new analysis approach for comparison of the time series from the observer and instruments is presented. This approach reduces the mismatch of the different sampling intervals and response times of humans and instruments when observing precipitation events. The algorithm postprocesses the minutely output estimates from the sensors in a window of time around the nominal time of the human observation.
The paper examines the relationship between the probability of detection and false alarm ratio for these sensors. In order to represent the “trade-off” between these parameters, the Heidke skill score is used as a figure of merit in comparing performance. A statistical method is presented to test for significance in differences of this score between sensors.
Each of the sensors demonstrated skill in identifying and typing the precipitation. The window analysis showed improved scores compared to simultaneous comparisons. The difference is attributed to the analysis technique that was designed to approximate the observing instructions for the standard observations. The data processing algorithm that gave the highest Heidke skill score for each sensor resulted in identification scores of 79% for the microwave sensor but only 39%–40% for the three optical sensors when rain was reported by the observer. The identification score in snow was 63% for the microwave sensor and in the range of 53%–71% for the optical sensors. Use of a multiparameter algorithm with the microwave sensor improved the identification of both snow and drizzle over the report from the sensor alone.
Abstract
Precipitation detection and typing estimates from four sensors are evaluated using standard operational meteorological observations as a reference. All are active remote sensors radiating into a measurement volume near the sensor, and measuring some property of the scattered radiation. Three of the sensors use optical wavelengths and one uses microwave wavelengths.
A new analysis approach for comparison of the time series from the observer and instruments is presented. This approach reduces the mismatch of the different sampling intervals and response times of humans and instruments when observing precipitation events. The algorithm postprocesses the minutely output estimates from the sensors in a window of time around the nominal time of the human observation.
The paper examines the relationship between the probability of detection and false alarm ratio for these sensors. In order to represent the “trade-off” between these parameters, the Heidke skill score is used as a figure of merit in comparing performance. A statistical method is presented to test for significance in differences of this score between sensors.
Each of the sensors demonstrated skill in identifying and typing the precipitation. The window analysis showed improved scores compared to simultaneous comparisons. The difference is attributed to the analysis technique that was designed to approximate the observing instructions for the standard observations. The data processing algorithm that gave the highest Heidke skill score for each sensor resulted in identification scores of 79% for the microwave sensor but only 39%–40% for the three optical sensors when rain was reported by the observer. The identification score in snow was 63% for the microwave sensor and in the range of 53%–71% for the optical sensors. Use of a multiparameter algorithm with the microwave sensor improved the identification of both snow and drizzle over the report from the sensor alone.
Abstract
The Precipitation Occurrence Sensor System (POSS) is a small X-band Doppler radar originally developed by the Meteorological Service of Canada for reporting the occurrence, type, and intensity of precipitation from Automated Weather Observing Stations. This study evaluates POSS as a gauge for measuring amounts of both liquid and solid precipitation. Different precipitation rate estimation algorithms are described. The effect of different solid precipitation types on the Doppler velocity spectrum is discussed. Lacking any accepted reference for high temporal resolution rates, the POSS precipitation rate measurements are integrated over time periods ranging from 6 h to one day and validated against international and Canadian reference gauges. Data from a wide range of sites across Canada and for periods of several years are used. The statistical performance of POSS is described in terms of the distribution of ratios of POSS to reference gauge amounts (catch ratios). In liquid precipitation the median of the catch ratio distribution is 82% and the interquartile range was between −12% and 19% about the median. In solid precipitation the median is 90% and the interquartile range is between −17% and 24% about the median. The underestimation in both liquid and solid precipitation is shown to be a function of precipitation rate and phase. The effects of radome wetting, raindrop splashing, wind, and the radar “brightband” effect on the estimation of precipitation rates are discussed.
Abstract
The Precipitation Occurrence Sensor System (POSS) is a small X-band Doppler radar originally developed by the Meteorological Service of Canada for reporting the occurrence, type, and intensity of precipitation from Automated Weather Observing Stations. This study evaluates POSS as a gauge for measuring amounts of both liquid and solid precipitation. Different precipitation rate estimation algorithms are described. The effect of different solid precipitation types on the Doppler velocity spectrum is discussed. Lacking any accepted reference for high temporal resolution rates, the POSS precipitation rate measurements are integrated over time periods ranging from 6 h to one day and validated against international and Canadian reference gauges. Data from a wide range of sites across Canada and for periods of several years are used. The statistical performance of POSS is described in terms of the distribution of ratios of POSS to reference gauge amounts (catch ratios). In liquid precipitation the median of the catch ratio distribution is 82% and the interquartile range was between −12% and 19% about the median. In solid precipitation the median is 90% and the interquartile range is between −17% and 24% about the median. The underestimation in both liquid and solid precipitation is shown to be a function of precipitation rate and phase. The effects of radome wetting, raindrop splashing, wind, and the radar “brightband” effect on the estimation of precipitation rates are discussed.
Abstract
Three techniques for the measurement of raindrop size distributions are compared using data from a Joss-Waldvogel disdrometer (JWD), a Particle Measuring Systems 2DG spectrometer (PMS), and an Atmospheric Environment Service (AES) Precipitation Occurrence Sensor System (POSS). The techniques used are impact measurement by the JWD, optical imaging by the PMS, and Doppler velocity spectrum by the POSS. The sampling size errors arc compared. The effects of both vertical and horizontal winds on the measurements are evaluated. Accumulated rainfall amounts derived from the drop size distributions (DSDs) are compared to measurements by conventional gauges at the radar facility of AES at King City, Ontario, Canada. In general, DSDs and rain rates averaged for 1 min were in agreement between the three sensors While 1-min-averdged DSDs were multimodal, long-time-period averages followed the Marshall-Palmer distribution.
Abstract
Three techniques for the measurement of raindrop size distributions are compared using data from a Joss-Waldvogel disdrometer (JWD), a Particle Measuring Systems 2DG spectrometer (PMS), and an Atmospheric Environment Service (AES) Precipitation Occurrence Sensor System (POSS). The techniques used are impact measurement by the JWD, optical imaging by the PMS, and Doppler velocity spectrum by the POSS. The sampling size errors arc compared. The effects of both vertical and horizontal winds on the measurements are evaluated. Accumulated rainfall amounts derived from the drop size distributions (DSDs) are compared to measurements by conventional gauges at the radar facility of AES at King City, Ontario, Canada. In general, DSDs and rain rates averaged for 1 min were in agreement between the three sensors While 1-min-averdged DSDs were multimodal, long-time-period averages followed the Marshall-Palmer distribution.
Abstract
Precipitation and environmental conditions occurring during accretion in Canadian east coast winter storms are described and investigated. Accretion is generally associated with snow, freezing rain, and ice pellets within saturated conditions. Precipitation types are sometimes invariant but usually evolve during individual accretion events. Accretion events are also generally associated with moderate wind speeds (average of 7.5 m s−1) and warm temperatures (between −1° and 0°C are most common). Remote sensing of particle shapes and terminal velocities are capable of identifying some of the features of these precipitation types. Model calculations indicate that a detailed understanding of precipitation characteristics, such as the nature of wet snow, is needed to accurately simulate accretion.
Abstract
Precipitation and environmental conditions occurring during accretion in Canadian east coast winter storms are described and investigated. Accretion is generally associated with snow, freezing rain, and ice pellets within saturated conditions. Precipitation types are sometimes invariant but usually evolve during individual accretion events. Accretion events are also generally associated with moderate wind speeds (average of 7.5 m s−1) and warm temperatures (between −1° and 0°C are most common). Remote sensing of particle shapes and terminal velocities are capable of identifying some of the features of these precipitation types. Model calculations indicate that a detailed understanding of precipitation characteristics, such as the nature of wet snow, is needed to accurately simulate accretion.
Abstract
Atmospheric temperature profiles and integrated water vapor and liquid are retrieved from ground-based microwave radiometric measurements using both nonlinear optimal estimation (NLOE) and statistical inversion (SI) methods. The results obtained from both methods are compared with collocated radiosonde observations during the Canadian Atlantic Storms Program field project in 1986. In general, the NLOE was superior to the SI method when clouds with high liquid water contents or when precipitation was present. Under these conditions, temperature profiles derived using NLOE had smaller root-mean-square differences from radiosonde observations than those retrieved using SI. Also, the overestimation of integrated vapor retrieved using the SI method was eliminated using the NLOE method. The radiometric observations were used in two case studies of winter cyclonic storms striking Atlantic Canada.
Abstract
Atmospheric temperature profiles and integrated water vapor and liquid are retrieved from ground-based microwave radiometric measurements using both nonlinear optimal estimation (NLOE) and statistical inversion (SI) methods. The results obtained from both methods are compared with collocated radiosonde observations during the Canadian Atlantic Storms Program field project in 1986. In general, the NLOE was superior to the SI method when clouds with high liquid water contents or when precipitation was present. Under these conditions, temperature profiles derived using NLOE had smaller root-mean-square differences from radiosonde observations than those retrieved using SI. Also, the overestimation of integrated vapor retrieved using the SI method was eliminated using the NLOE method. The radiometric observations were used in two case studies of winter cyclonic storms striking Atlantic Canada.
Abstract
Radiometrically determined temperature profiles were compared with radiosonde observations during the Canadian Atlantic Storms Program (CASP). A total of 108 profiles were available for the comparison, taken during diverse weather conditions. The root-mean-square difference between the temperature determined by radiometer and by radiosonde generally increased with height, from approximately 1°K near the ground to about 3.5°K at the 400 mb level. The accuracy of the radiometer estimates rapidly deteriorated at altitudes above this level. These results are consistent with the accuracy expected from the retrieval method used and with the results reported by others.
The total sample was stratified to examine possible effects of precipitation type, fog, and temperature inversions on the accuracy of the retrieved temperature profiles. It was found that fog and rain cause a bias in the measurements, with the retrieved temperatures too warm by an average amount of up to 2°K, depending on altitude. Snow, however, was found to cause no degradation in the accuracy of radiometric temperature estimates. Temperature inversions near the ground were fairly accurately identified by the radiometer, although inversions aloft were imprecisely measured.
Thicknesses of standard layers from radiometer observations agreed closely with those determined by radiosonde. The root-mean-square percentage differences were less than 2%. Radiometric estimates of total water vapor were also accurate, agreeing within 13% of the radiosonde data and having a correlation coefficient of 0.97. These estimates were less accurate when rain or fog was present.
Abstract
Radiometrically determined temperature profiles were compared with radiosonde observations during the Canadian Atlantic Storms Program (CASP). A total of 108 profiles were available for the comparison, taken during diverse weather conditions. The root-mean-square difference between the temperature determined by radiometer and by radiosonde generally increased with height, from approximately 1°K near the ground to about 3.5°K at the 400 mb level. The accuracy of the radiometer estimates rapidly deteriorated at altitudes above this level. These results are consistent with the accuracy expected from the retrieval method used and with the results reported by others.
The total sample was stratified to examine possible effects of precipitation type, fog, and temperature inversions on the accuracy of the retrieved temperature profiles. It was found that fog and rain cause a bias in the measurements, with the retrieved temperatures too warm by an average amount of up to 2°K, depending on altitude. Snow, however, was found to cause no degradation in the accuracy of radiometric temperature estimates. Temperature inversions near the ground were fairly accurately identified by the radiometer, although inversions aloft were imprecisely measured.
Thicknesses of standard layers from radiometer observations agreed closely with those determined by radiosonde. The root-mean-square percentage differences were less than 2%. Radiometric estimates of total water vapor were also accurate, agreeing within 13% of the radiosonde data and having a correlation coefficient of 0.97. These estimates were less accurate when rain or fog was present.
