|
Ceiling & Visibility
|
||||
J. Ceiling and Visibility
[Background]
[National
Ceiling and Visibility System]
[Feasibility
Study on Visibility Nowcasting]
[Northeast
Corridor Field Study] [Northeast
Terminal Ceiling and Visibility]
[San
Francisco Airport Ceiling and Visibility]
1. Background
Adverse ceiling and visibility (C&V) conditions can produce major negative impacts on aviation - as a contributing factor in over 35% of all weather-related accidents in the U.S. civil aviation sector and as a major cause of flight delays nationwide. RAP research funded by the FAA directly addresses the scientific and engineering challenges brought about the need for improved weather information to address these problems.
C&V effects have their most critical impact on the general aviation (GA) pilot, particularly during the in-route phase of flight where unexpected encounters with reduced ceiling and visibility conditions are most likely. In these situations, the GA pilot may face serious risk of disorientation, loss of control, and controlled flight into obstructed terrain. Approximately 168 pilots and passengers are killed each year as a result of adverse C&V conditions. Improved national-scale analyses and forecasts of ceiling and visibility and new tools to present this information to pilots, weather briefers and others are needed to improve the C&V-related safety record. Toward this end, RAP leads the FAA National Ceiling and Visibility (NCV) product development team {link to www.rap.ucar.edu/projects/cvis}, composed of researchers from RAP, MIT Lincoln Laboratory, the Naval Research Laboratory at Monterey (NRL), NOAA's Forecast Systems Laboratory, the University of Quebec at Montreal, and NOAA's Aviation Weather Center.
Adverse C&V conditions also impose critical restrictions upon rates of traffic flow into and out of major airports across the country, and these restrictions in turn strongly affect the capacity and efficiency of the U.S. air traffic system as a whole. C&V forecasts for the air terminal area are critical to timely airport decisions on the opening and closing of individual runways, and for air traffic control decisions on traffic routing and spacing in the terminal area. The FAA Terminal Ceiling and Visibility (TCV) product development team leads the research needed for improved terminal-area forecasts and operational tools linking forecast information to the airport decision-making process. This team joins the R&D capabilities of MIT Lincoln Laboratories and RAP.
Though their domains differ, the NCV and TCV programs collaborate closely in the use of observations, numerical and statistical modeling, computer automation, and targeted research into problems in C&V phenomenology.
2. National Ceiling and Visibility System
RAP outlined the updated schematic representation for the NCV automated forecast system shown in Figure 1. The strategy is to develop parallel automated forecast techniques (e.g., regional and local numerical models, observation-based statistical techniques, rule-based methods, etc.) and to merge those inputs using an expert system-based integration process. As shown in Figure 1, forecast results are represented graphically in terms of aviation-critical thresholds for ceiling, visibility and flight category, e.g., VFR (visual flight rules), and IFR (instrument flight rules), and are presented on the NCV website. Real-time verification of component forecast skill feeds back to optimize the weighting of forecast components in the automated merging process.
The major improvements to the NCV forecast system in FY02 included: (1) upgrades to 20 km resolution, improved model microphysics, and improved visibility forecast methodology brought by RUC model improvements and the increase in RUC resolution from 40 to 20 km; (2) incorporation of four-per-day Eta model forecast inputs; (3) development of improved web-based display code providing a prototype Java application which is independent of browser characteristics allowing the saving of user preferences; (4) simplified color coding of C&V parameter values for forecast display; and (5) improved incorporation of elevation effects in the interpolation of observed C&V values to produce real-time analyses.
Figure 1. Schematic representation
of NCV forecast system components in use today and planned for future
implementation. Forecast components flow to the expert system-based
automated merging process shown at center.
Several other important research and development efforts yielded significant progress during the year. Application of NRL methodology for satellite-based cloud classification can significanly improve the automated distinction between cloudy and cloud-free regions. RAP imported that capability and is implementing this processing for both GOES East and West. An exploratory daytime cloud classification capability for GOES East was recently implemented to support future evaluation of the methodology and its application to cloud discrimination. Similar work toward a nighttime capability will follow in FY03, leading to a next-generation day/night cloud mask procedure for the NCV system.
The NCV system currently uses a nearest-neighbor interpolation technique for handling of point observations used to produce a real-time analysis of C&V quantities. RAP has begun examination of the krieging method as an alternative scheme to determine whether improvements can be made in the relevance of interpolated fields in problematic regions.
3. Feasibility Study on Visibility
Nowcasting Using Radar
ASOS
installations at major airports include sensors which measure visibility.
Dixon and Rasmussen used those visibility measurements, along with
radar observations and other ASOS data, to investigate the relationship
between radar reflectivity and visibility and to determine whether
it is feasible to base a short-term visibility forecast on those observations. The
results showed that a visibility nowcast is possible under snow conditions
using the motion of the radar echo, the vertical wind profile from
the radar beam to the surface, and taking into account the time for
the particle to fall from the radar volume to the surface.
These considerations are schematically shown in
Figure 1. In addition, a real-time calibration of visibility and radar reflectivity
is made in order to correlate radar trends of reflectivity with visibility
changes.Figure 2 shows the correlation
between reflectivity and visibility for a storm occurring on 18 March
2002 in New York.The good correlation evident in this figure suggests
that a visibility nowcast is indeed possible.
Figure 2.
Schematic of the radar/ASOS geometry and a sample snow trajectory.
Figure
3. Example of correlation between visibility (extinction
coefficient) and radar reflectivity (dBZ) of snow.
4. Northeast Corridor Field Study
The northeastern U.S. air travel corridor centered on New York City handles an extraordinarily high concentration of domestic and international transport-class flights, and has a particularly problematic susceptibility to the degradation of traffic efficiency due to adverse C&V in the late fall, winter and spring months. Low ceilings due to stratus, cloud and visibility reduction due to fog and precipitation are key issues in the area. The Northeast Corridor Field Study is an emerging collaborative effort among NCV, TCV and the FAA Winter Weather program to study the phenomenology of low ceiling and visibility events in this corridor, to engage the practical expertise of operational forecasters in the public and private sector in the region, and, as a result, to build improved analysis and forecast systems and user tools to better meet operational needs for dealing with low ceiling and visibility in the northeast. The field study is centered in the New York City area. Collection of field data was initiated in the fall of 2001 at the Rutgers University field site in New Jersey.
During this first year, field site pros and cons across the region were assessed in terms of availability of existing instrumentation, proximity to the New York airport region, proximity to operational sounding data, and suitability as a domain for assessment of smaller-scale column modeling techniques. A location at Brookhaven National Laboratory was selected as the primary field site since this offers use of a 90-m instrumentation tower (Figure 4), close proximity (~1 km) to the NWS Forecast Office and rawinsonde launch site, a secure instrument location, and excellent cooperation from the Brookhaven staff. The Rutgers location continues as an important secondary site providing ongoing observations and excellent cooperation from Rutgers University staff.
Figure 4. The Brookhaven National Laboratory
90 m tower (seen with slight image distortion).
Plans for the installation of key instrumentation at the Brookhaven site, use of its 90 m tower, and coordination with the collocated NWS Forecast/Rawinsonde Office have been developed. Instrumentation supporting use of the 90 m tower as a tool to observe the structure and evolution of fog is outlined in Table J1. The lease/purchase of a Radiometrics Inc. 12-channel microwave radiometer was contracted, and operation at the site will yield continuous profiling for temperature, humidity and liquid water. Testing of a new 14-channel Radiometrics microwave radiometer is also planned. Analysis of radiometer-derived profiles and verification against NWS soundings from the Brookhaven site are key to the assessment and use of the radiometer data. Additional site equipment will include a laser ceilometer, lightning protection for the tower and its immediate vicinity, and a RAP rawinsonde for supplemental soundings.
Tentative design for NCAR sensor deployment at
the BNL site.
R. Tardif has begun a climatological study of fog events in the New York City airport region. Focusing on cases for which the observed visibility was below 1 statute mile, results confirm the importance of moisture sources at the mesoscale. Increased fog frequencies were detected when areas characterized by high surface moisture availability (wetlands northeast of the Newark airport, Long Island Sound and coastal waters of the Atlantic Ocean) were located upwind from the stations considered (see Fig. J4). Higher frequencies of fog associated with the presence of precipitation were observed for the more urban locations (Newark and LaGuardia), pointing toward a possible urban heat island effect. The overall results suggest differences in the main formation mechanism of fog at the various locations. Examination of data for the Brookhaven site indicate that fog is more frequent there than in the Newark/LaGuardia/Kennedy airport area, 55 miles to the west.
Figure 5. Distribution of fog frequency (visual range < 1 statute mile) relative to wind direction for John F. Kennedy Airport (1977-1990 period).
5. Northeast Terminal Ceiling and Visibility
The FAA's Terminal Ceiling and Visibility (TC&V) Product Development Team, led by RAP, is responsible for development of operational products specifically designed for high-volume airports, which have a substantial loss of operating capacity during Instrument Meteorological Conditions (IMC). Since the implementation and cancellation of ground delay programs are highly dependent upon anticipation of changes in the available operating capacity, an accurate prediction of the onset and cessation of IMC provides an opportunity for improved traffic management decisions. The Initial C&V product development focused on the development of Marine Stratus forecast products for San Francisco International Airport.
Development of C&V products for the Northeast will extend the San Francisco technical developments, with a focus on C&V events associated with extratropical storms. This product development initiative will serve to reduce the C&V impacts in the northeast corridor and the upper Midwest. The New York City Air Traffic Control Center will serve as the initial operational testbed for product development, since it is already supported by a cooperative user group community and product development team associated with the Integrated Terminal Weather System (ITWS). Additionally, its weather exposure typifies that associated with this class of phenomenology, and it is an excellent candidate to reap immediate short-term benefits from product development to reduce delays associated with its unusually dense traffic volume.
In March 2002, a User Group meeting was held on Long Island, New York, to garner feedback from the regional traffic management and aviation communities regarding the causes and impact of low ceiling and visibility conditions. Two primary C&V product needs were identified: a map of current and future ceilings in the airport region; and a map of current and future surface visibility at the airport. The underlying philosophy for these product developments is to provide a consistent unified product presentation in which the component technical developments provide the information content that becomes more accurate as the skill evolves. A prototype product display was presented to and accepted by the users as a good starting point for conveying information necessary for their operational needs. The development process will involve technologies to provide information for this display. The time-scales of the products fall into four categories:
1) Diagnosis of the current situation
2) Nowcast (0-2 hours)
3) Tactical forecast (2-6 hours)
4) Strategic forecast (6-12 hours)
6. San Francisco Airport Ceiling and Visibility
San Francisco International Airport (SFIA) has the greatest delay due to C&V in the United States by a factor of two. Most of this delay occurs during the summer months, and is caused by marine stratus over the San Francisco Bay. The goal of this product development is to develop accurate 0-4 hour forecasts of the time of clearing in the approach to SFIA. This product development was initiated in 1996 concluded with operational demonstrations in 2001 and 2002 and will continue into 2003.
The marine stratus forecast system will provide products and supporting data to meteorologists forecasters, who are charged with forecasting the time that SFIA can initiate side-by-side approaches at the conclusion of a marine stratus event. The system collects data from public and special operational sensor systems, builds derived products, and executes several forecast models. A browser-based display presents the most pertinent sensor data and forecast products.
The estimation of the height of the inversion base has been identified as a critical element in the forecasting of the behavior of marinestratus. The development of a reliable technology for the estimation of the height of the inversion base was taken as the top priority for the project. This was followed by the development of forecast algorithms, which could assimilate the inversion base information. Four forecast algorithms were developed: a local dynamical model (COBEL) and three statistical algorithms. The final step wsa the development of an automated consensus forecast, based on the results from the component forecasts. This product structure and the user-friendly, Web-based display are the basis for the operational demonstrations. The performance of each the product was reviewed after the 2001 season, and improvements were implemented and demonstrated for the 2002 season.
COBEL forecasts are obtained by numerically integrating the equations describing atmospheric boundary layer flows and thermodynamic state. During the integration, from observations of the lower atmosphere are assimilated into the model to retrieve non-measured parameters such as temperature advection, vertical motion and cloud water amount. Retrieved parameters can then be used in subsequent, updated, COBEL forecasts to improve the results.
Initial profiles are provided to the model by using a procedure specially adapted to the SFO problem. This procedure consists of adapting the 1200Z Oakland NWS sounding to the observed local conditions at SFO. This is done by using the observed cloud base and cloud top (inversion base), to determine the height and thickness of the cloud layer, and using the measured surface temperature and dew point to define the sub-cloud temperature and humidity. The incoming solar radiation at the surface, measured at 1400UTC, is used to deduce the initial total cloud water amount. A linear profile within the cloud is assumed for liquid water content.
The statistical forecast models (SFMs) were developed using an approach is based on multi-linear regression, where the predictors have been nonlinearly rescaled to enhance correlations. Three different SFMs were developed. The Regional SFM uses regional NWS data. The Local SFM depends primarily on data collected from project sensors. This model reflects concepts, identified by operational forecasters as critical in their forecast approach. The Satellite SFM correlates subtle changes in the brightness and patterns of the satellite-observed clouds with the time of clearing.
The Consensus Forecast assesses the component forecasts and expressions of confidence in their forecasts, to arrive at a statistical best-estimate of the time clearing.
[TOP]