Search Results
You are looking at 1 - 3 of 3 items for
- Author or Editor: Robert Barron x
- Refine by Access: All Content x
Abstract
The impact of the number of satellite altimeters providing sea surface height anomaly (SSHA) information for a data assimilation system is evaluated using two comparison frameworks and two statistical methodologies. The Naval Research Laboratory (NRL) Layered Ocean Model (NLOM) dynamically interpolates satellite SSHA track data measured from space to produce high-resolution (eddy resolving) fields. The Modular Ocean Data Assimilation System (MODAS) uses the NLOM SSHA to produce synthetic three-dimensional fields of temperature and salinity over the global ocean. A series of case studies is defined where NLOM assimilates different combinations of data streams from zero to three altimeters. The resulting NLOM SSHA fields and the MODAS synthetic profiles are evaluated relative to independently observed ocean temperature and salinity profiles for the years 2001–03. The NLOM SSHA values are compared with the difference of the observed dynamic height from the climatological dynamic height. The synthetics are compared with observations using a measure of thermocline depth. Comparisons are done point for point and for 1° radius regions that are linearly fit over 2-month periods. To evaluate the impact of data outliers, statistical evaluations are done with traditional Gaussian statistics and also with robust nonparametric statistics. Significant error reduction is obtained, particularly in high SSHA variability regions, by including at least one altimeter. Given the limitation of these methods, the overall differences between one and three altimeters are significant only in bias. Data outliers increase Gaussian statistical error and error uncertainty compared to the same computations using nonparametric statistical methods.
Abstract
The impact of the number of satellite altimeters providing sea surface height anomaly (SSHA) information for a data assimilation system is evaluated using two comparison frameworks and two statistical methodologies. The Naval Research Laboratory (NRL) Layered Ocean Model (NLOM) dynamically interpolates satellite SSHA track data measured from space to produce high-resolution (eddy resolving) fields. The Modular Ocean Data Assimilation System (MODAS) uses the NLOM SSHA to produce synthetic three-dimensional fields of temperature and salinity over the global ocean. A series of case studies is defined where NLOM assimilates different combinations of data streams from zero to three altimeters. The resulting NLOM SSHA fields and the MODAS synthetic profiles are evaluated relative to independently observed ocean temperature and salinity profiles for the years 2001–03. The NLOM SSHA values are compared with the difference of the observed dynamic height from the climatological dynamic height. The synthetics are compared with observations using a measure of thermocline depth. Comparisons are done point for point and for 1° radius regions that are linearly fit over 2-month periods. To evaluate the impact of data outliers, statistical evaluations are done with traditional Gaussian statistics and also with robust nonparametric statistics. Significant error reduction is obtained, particularly in high SSHA variability regions, by including at least one altimeter. Given the limitation of these methods, the overall differences between one and three altimeters are significant only in bias. Data outliers increase Gaussian statistical error and error uncertainty compared to the same computations using nonparametric statistical methods.
Abstract
Standard meteorological model performance evaluation (sMPE) can be insufficient in determining “fitness” for air quality modeling. An sMPE compares predictions of meteorological variables with community-based thresholds. Conceptually, these thresholds measure the model’s capability to represent mesoscale features that cause variability in air pollution. A method that instead examines features could provide a better estimate of fitness. This work compares measures of fitness from sMPE analysis with a feature-based MPE (fMPE). Meteorological simulations for Bogotá, Colombia, using the Weather Research and Forecasting (WRF) Model provide an ideal case study that highlights the importance of fMPE. Bogotá is particularly interesting because the complex topography presents challenges for WRF in sMPE. A cluster analysis identified four dominant meteorological features associated with air quality driven by wind patterns. The model predictions are able to pass several sMPE thresholds but show poor performance for wind direction. The base simulation can be improved with alternative surface characterization datasets for terrain, soil classification, and land use. Despite doubling the number of days with acceptable specific humidity, overall acceptability was never more than 10%. By comparison, an fMPE showed that predictions were able to reproduce the air-quality-relevant features on 38.4% of the days. The fMPE is based on features derived from an observational cluster analysis that have clear relationships with air quality, which suggests that reproducing those features will indicate better air quality model performance. An fMPE may be particularly useful for high-resolution modeling (1 km or less) when finescale variability can cause poor sMPE performance even when the general pattern that drives air pollution is well reproduced.
Abstract
Standard meteorological model performance evaluation (sMPE) can be insufficient in determining “fitness” for air quality modeling. An sMPE compares predictions of meteorological variables with community-based thresholds. Conceptually, these thresholds measure the model’s capability to represent mesoscale features that cause variability in air pollution. A method that instead examines features could provide a better estimate of fitness. This work compares measures of fitness from sMPE analysis with a feature-based MPE (fMPE). Meteorological simulations for Bogotá, Colombia, using the Weather Research and Forecasting (WRF) Model provide an ideal case study that highlights the importance of fMPE. Bogotá is particularly interesting because the complex topography presents challenges for WRF in sMPE. A cluster analysis identified four dominant meteorological features associated with air quality driven by wind patterns. The model predictions are able to pass several sMPE thresholds but show poor performance for wind direction. The base simulation can be improved with alternative surface characterization datasets for terrain, soil classification, and land use. Despite doubling the number of days with acceptable specific humidity, overall acceptability was never more than 10%. By comparison, an fMPE showed that predictions were able to reproduce the air-quality-relevant features on 38.4% of the days. The fMPE is based on features derived from an observational cluster analysis that have clear relationships with air quality, which suggests that reproducing those features will indicate better air quality model performance. An fMPE may be particularly useful for high-resolution modeling (1 km or less) when finescale variability can cause poor sMPE performance even when the general pattern that drives air pollution is well reproduced.
Abstract
The Juneau, Alaska, airport vicinity experiences frequent episodes of moderate and severe turbulence, which affect arriving and departing air traffic. The Federal Aviation Administration funded the National Center for Atmospheric Research to develop a warning system, consisting of carefully placed anemometers and wind profilers, along with data communications, an algorithm, and display, to warn pilots of potentially hazardous situations. The system uses regressions based on comparisons of research aircraft data with measurements from the ground-based sensors to estimate the turbulence intensity along selected flight paths. This paper describes the development of the turbulence warning system, from meteorological characteristics through sensor placement, algorithm construction and evaluation, and display design. The discussion includes how best estimates of winds were made in adverse meteorological and topographic conditions, how turbulence was calculated from aircraft conducting various flight maneuvers, how bad data were identified and removed from the system, how the regressors were selected, and the skill of the system.
Abstract
The Juneau, Alaska, airport vicinity experiences frequent episodes of moderate and severe turbulence, which affect arriving and departing air traffic. The Federal Aviation Administration funded the National Center for Atmospheric Research to develop a warning system, consisting of carefully placed anemometers and wind profilers, along with data communications, an algorithm, and display, to warn pilots of potentially hazardous situations. The system uses regressions based on comparisons of research aircraft data with measurements from the ground-based sensors to estimate the turbulence intensity along selected flight paths. This paper describes the development of the turbulence warning system, from meteorological characteristics through sensor placement, algorithm construction and evaluation, and display design. The discussion includes how best estimates of winds were made in adverse meteorological and topographic conditions, how turbulence was calculated from aircraft conducting various flight maneuvers, how bad data were identified and removed from the system, how the regressors were selected, and the skill of the system.
