Characterizing the severe turbulence environments associated with commercial aviation accidents. A real-time turbulence model (RTTM) designed for the operational prediction of hazardous aviation turbulence environments

Please download to get full document.

View again

of 36
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Characterizing the severe turbulence environments associated with commercial aviation accidents. A real-time turbulence model (RTTM) designed for the operational prediction of hazardous aviation turbulence environments
Document Share
Document Transcript
  Meteorol Atmos Phys 94, 235–270 (2006)DOI 10.1007/s00703-005-0181-4 1 Department of Marine, Earth, and Atmospheric Sciences, North Carolina State University, Raleigh, NC, USA 2 USDA = Forest Service, North Central Research Station, East Lansing, MI, USA 3 MESO  Inc., Raleigh, NC, USA Characterizing the severe turbulence environmentsassociated with commercial aviation accidents. A real-timeturbulence model ( RTTM ) designed for the operationalprediction of hazardous aviation turbulence environments M. L. Kaplan 1 , J. J. Charney 2 , K. T. Waight III 3 , K. M. Lux 1 , J. D. Cetola 1 ,A. W. Huffman 1 , A. J. Riordan 1 , S. D. Slusser 1 , M. T. Kiefer 1 ,P. S. Suffern 1 ,  and  Y.-L. Lin 1 With 19 FiguresReceived April 1, 2004; revised April 25, 2005; accepted November 24, 2005Published online: July 31, 2006 # Springer-Verlag 2006 Summary In this paper, we will focus on the real-time prediction of environments that are predisposed to producing moderate-severe (hazardous) aviation turbulence. We will describe thenumerical model and its postprocessing system that is de-signed for said prediction of environments predisposed tosevere aviation turbulence as well as presenting numerousexamples of its utility. The purpose of this paper is to dem-onstrate that simple hydrostatic precursor circulations orga-nize regions of preferred wave breaking and turbulence at thenonhydrostatic scales of motion. This will be demonstratedwith a hydrostatic numerical modeling system, which can berun in real time on a very inexpensive university computerworkstation employing simple forecast indices. The forecastsystem is designed to efficiently support forecasters who aredirecting research aircraft to measure the environmentimmediately surrounding turbulence.The numerical model is  MASS  version 5.13, which is in-tegrated over three different grid matrices in real-time on auniversity workstation in support of   NASA -Langley’s B-757turbulence research flight missions. The model horizontalresolutions are 60, 30, and 15km and the grids are centeredover the region of operational  NASA -Langley B-757 turbu-lence flight missions.The postprocessing system includes several turbulence-relatedproductsincludingfourturbulenceforecastingindices,winds,streamlines,turbulencekineticenergy,andRichardsonnumbers. Additionally there are convective products includ-ingprecipitation,cloudheight,cloudmassfluxes,liftedindex,and K-index. Furthermore, soundings, sounding parameters,and Froude number plots are also provided. The horizontalcross section plot products are provided from 16,000–46,000feet in 2,000 feet intervals. Products are available every threehoursatthe60and 30km grid interval and every 1.5 hours atthe 15km grid interval. The model is initialized from the NWS ETA  analyses and integrated two times a day. 1. Introduction The operational forecasting of turbulence poten-tial has been ongoing for several years. Manyindices aregenerated daily from operational num-erical weather prediction models. The NationalWeather Service has employed the Ellrod Index(e.g., Ellrod and Knapp, 1992), the  NOAA  Fore-casting Systems Laboratory has employed in-dices developed by Marroqin et al (1990) basedon turbulence kinetic energy and eddy dissipa-tion rate, and the Research Applications Programof the National Center for Atmospheric Researchhad developed the Integrated Turbulence Fore-casting Algorithm ( ITFA ) index as part of the  suite of products from the  NWS RUC  II-model(Sharman et al, 2000). This system is now knownas the Graphical Turbulence Guidance ( GTG )system. The Ellrod Index is by far the simplestbased solely on deformation and vertical windshear. The Marroquin index is based on a for-mulation of turbulence kinetic energy.  GTG  is,by far, the most comprehensive and sophisticatedindex, that is used operationally, employing aweighted set of nearly twenty individual compo-nent terms as well as contemporary observationsof turbulence pireps into a turbulence probabilityindex. In spite of the sophistication of   GTG  it isdesigned primarily for nonconvective turbulence,i.e., clear air turbulence ( CAT ) and mountainwave-induced turbulence. It is not exclusively de-signed to denote regions of severe turbulence buta broad cross-section of turbulence intensities in-cluding light, moderate, and severe. As noted inKaplan et al (2005a), most severe aviation turbu-lence encounters that result in damage to aircraftor onboard injuries are closely associated withmoist convection. Hence, a turbulence predictiveindex, to be employed on an operational basis,needs to have the capacity to predict not only  theenvironments that organize  light-moderate turbu-lence, which typically occurs in clear air, but themore severe turbulence most often found in asso-ciation with moist convection, i.e., convectively-induced turbulence ( CIT ) which is most oftenassociated with aviation accidents. Additionally,one of the negative aspects of   GTG , is that it isan amalgamation of so many different specificdynamical terms and empirical weighting factorsthat it is unclear what physical = dynamical pro-cesses are most relevant in organizing the envi-ronment that is favorable for turbulence in a givencase study, particularlysevere accident-producingturbulence. Therefore one is dependent upon sta-tistical validation rather than dynamical under-standing when the inevitable improvement of saidindex is undertaken. The use of the much simpler RTTM  indices facilitates a clearer understandingof what hydrostatic environments tend to orga-nize fine scale turbulence.In the remainder of this paper we will demon-strate the utility of the index described in Kaplanetal(2005a, b)indelimiting,inreal-time,areasof moderate-severe turbulence potential. Addition-ally we will demonstrate that said index has thecapacity to denote all three forms of turbulenceincluding  CAT ,  CIT  and mountain wave-inducedturbulence. We will do so with a hydrostatic nu-merical modeling system that is very inexpensiveto run on a university computer workstation. Inthe next section of the paper we will describe thenumerical modeling system and postprocessoremployed in the Real-Time Turbulence Model( RTTM ) run operationally at North Carolina StateUniversity in support of   NASA ’s B-757 turbu-lence research missions. We will focus on a de-scription of the key turbulence-forecasting indexin this section of the paper. The third section of this paper will focus on several real-time examples Table 1.  MASS  model (vers. 5.13) characteristicsModel numerics– Hydrostatic primitive equation model– 3-D equations for  u ,  v ,  T  ,  q , and  p – Cartesian grid on a polar stereographic map– Sigma-p terrain-following vertical coordinate– Vertical coverage from   10m to   16,000m– Energy-absorbing sponge layer near model top– Fourth-order horizontal space differencing on anunstaggered grid– Split–explicit time integration schemes: (a) forward–backward for the gravity mode and(b) Adams–Bashforth for the advective mode– Time-dependent lateral boundary conditions– Positive–definite advection scheme for the scalar variables– Massless tracer equations for ozone and aerosol transportInitialization– First guess from large-scale gridded analyses– Reanalysis with rawinsonde and surface data usinga 3D optimum interpolation scheme– High-resolution terrain data base derived fromobservations– High-resolution satellite or climatological sea surfacetemperature database– High-resolution land use classification scheme– High-resolution climatological subsoil moisture database derived from antecedent precipitation– High resolution normalized difference vegetation index PBL  specification– Blackadar  PBL  scheme– Surface energy budget– Soil hydrology scheme– Atmospheric radiation attenuation schemeMoisture physics– Grid-scale prognostic equations for cloud water and ice,rainwater, and snow– Kain–Fritsch convective parameterization236 M. L. Kaplan et al  of the model, its index, and simulated precipita-tion and wind fields associated with case studiesof observed turbulence as reported by pireps. Theability of the model to denote  CIT  will be di-agnosed, in particular, although its versatility indefining regions of both  CAT  and mountain wave-induced turbulence will also be described. Thefinal section will focus on summarizing the mod-eling system and its application to the problem of the real-time forecasting of turbulence potential.It must be reiterated that this system was devel-oped for its efficiency and its convenient use by NASA  weather forecasters who are familiar withits predictive indices and with whose  collabora-tion  these indices were developed. Additionally,its scientific utility is that it demonstrates the key hydrostatic  processes and environments that like-ly organize nonhydrostatic wave breaking eventsresulting in aviation turbulence. 2. The numerical model and postprocessor The numerical model is  MASS  version 5.13(Kaplan et al, 2000). The  MASS  model is a hy-drostatic terrain-following sigma coordinate sys-tem with comprehensive boundary layer andconvective parameterizations (note Table 1). Themodel is integrated over three different horizon-tal resolutions, i.e., 60km, 30km, and 15km forthe coarse, fine1 and fine2 grids with the finesttwo meshes employing one-way nested-grid lat-eral boundary conditions (note Fig. 1). The gridmatrix size is 70  60  50 points for the coarsemesh grid, 90  100  50 points for the 30kmgrid, and 90  100  50 points for the 15kmgrid. The vertical sigma spacing is such that thereare 15 levels below 850mb, between 850mb and400mb there are levels every 20mb, and above400mb they are spaced every 40mb. The modelis very efficiently programmed for operationalutility. The initial state for the coarse mesh em-ploys the  NWS ETA  analyses at 0000 UTC and1200 UTC as well as reanalyzed rawinsonde andaviation surface observations employing an op- Fig. 1.  RTTM  coarse, fine1, and fine2 gridlocations Table 2.  List of   RTTM  productsTurbulence products– Horizontal winds– Horizontal streamlines– Richardson number–  NASA  Turbulence parameter–  NCSU 1 Turbulence parameter–  NCSU 2 Turbulence parameter– Stone Turbulence parameter (Knox parameter)– Turbulent Kinetic EnergyConvective products– Convective precipitation– Total precipitation– Cloud Heights– Cloud Mass Fluxes– Lifted Index– K-index– Skew-t = log-  p  soundings– Froude Number ProfilesCharacterizing the severe turbulence environments associated with commercial aviation 237  timum interpolation scheme. Time-dependent lat-eral boundary conditions are derived from the NWS ETA  model 40km data set and are updatedevery 3 hours at the 60km scale. Hourly coarsermesh simulated fields serveas the time-dependentlateral boundary conditions for the 30 and 15kmgridmeshes.The30kmsimulationisinitializedat3 hours past 0000 UTC and 1200 UTC from thecoarse mesh simulation and the 15km simulationis initialized at 3 hours past 0300 UTC and 1500UTC from the fine1 simulation. The three gridsare located in a manner that enables them to pro-duce simulations covering the entire 24 hour pe-riod, i.e., 24 hours at the coarse mesh, 21 hours atthe fine1 mesh, and 18 hours at the fine2 meshfrom the  NWS ETA  analysis data cycles consis-tent with the range of typical operational  NASA -Langley B-757 turbulence research flights. Themodeling system is designed solely to supportthe  NASA  turbulence research flight missions. Fig. 2a.  Pireps from the  NOAA Aviation Digital Data source val-id from 1326 UTC–1508 UTC18 September 2001; ( b ) Watervapor infrared satellite imageryvalid at 1415 UTC 18 September2001238 M. L. Kaplan et al  The postprocessing system is designed to sup-port real-time forecasts of turbulence potential foruse in directing the  NASA  B-757 research air-craft to locations of turbulence. There are fourcomponents to the postprocessor (note Table 2).The most important component is the suite of turbulence products listed in Table 2. Theseinclude: winds, streamlines, Richardson num-bers, turbulence kinetic energy, and four turbu-lence prediction indices. They are depicted onhorizontal surfaces from 16,000 feet to 46,000feet in 2,000 feet intervals, as mid-upper tropo-spheric turbulence is the focus of the B-757 tur-bulence research flights. Fig. 3a.  Peoria, IL ( ILX ) rawinsonde soundingvalid at 1200 UTC 18 September 2001; ( b )  RUC II simulated 300mb wind isotachs (ms  1 ) validat 1500 UTC 18 September 2001 with super-imposed observed 1200 UTC rawinsonde windbarbs(ms  1 ),temperatures(  C),andheights(m)Characterizing the severe turbulence environments associated with commercial aviation 239
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks